Each oscillator is a silica sphere of 3μm radius, capped on one side with 80nm of carbon and suspended in a water-lutidine mixture held at 28°C. A 532nm laser heats the cap and drives the particle toward an assigned target point, but the delay between imaging a particle and repositioning the beam means it overshoots and settles into a small orbit instead. Flow fields in the liquid couple neighboring orbits, and data enters the system as displacements of the target points.

Lattice spacing sets coupling strength, since hydrodynamic forces fall off with distance, and a damping threshold sets how far each particle swings. Both are adjustable while the experiment runs, with a forecasting error that varies by more than a factor of three across that parameter space. Accuracy also held up when the input reached only 20% of the oscillators, and when individual particles stopped responding to the laser or clumped together.

The colloidal array reached a normalized root-mean-squared error of about 0.1 on the one-step Mackey-Glass prediction. Memristor devices now reach 0.01 or better on the same benchmark, the paper notes, adding that those results depend on time-multiplexing and follow nearly a decade of concentrated work.

The authors write that their reservoir doesn't outperform established physical implementations. An arXiv preprint from January, however, framed it differently, arguing that avoiding time-multiplexing set the platform apart from nearly all existing physical reservoirs, photonic, memristive, and spintronic ones included.

Running the reservoir takes a 532nm laser, a two-axis acousto-optical deflector scanning at 100 kHz, real-time microscopy with particle tracking, a temperature-controlled quartz cell, and a conventional computer for the 1,000 Gaussian kernels and ridge regression that produce the output. No energy stats appear anywhere in the paper, despite energy efficiency being the stated motivation, and the authors concede that the laser-driven setup might not be practically applicable and instead point toward simpler actuation schemes, such as electrode-driven colloids.

Clemens Bechinger, professor of soft condensed matter at the University of Konstanz, said in the university's announcement that the dynamics don't need to be fully understood, only to respond reliably, at which point "its physics can be directly harnessed for computation."

A separate team synchronized 105,000 nano-oscillators in 45 nanoseconds this month on a platform projected to run at tens of gigahertz.

Unlike conventional zippers that connect two flat surfaces in 2D, the Y-Zipper joins three flexible arms into a rigid 3D triangular tube. When open or unzipped, the structure behaves like soft plastic strips or floppy tentacles, with each arm flexing and twisting independently. Once zipped shut with a custom slider, however, the arms interlock to form a stiff, beam-like structure capable of supporting loads.

The concept originated in 1985 with MIT professor William Freeman, who proposed a triangular zipper system intended to rapidly assemble objects such as tents, furniture, and containers. At the time, however, manufacturing limitations made the design impractical. Freeman patented the design with the hope that fabrication technology would eventually catch up. Nearly four decades later, modern 3D printers and computational design tools finally enabled researchers to revisit the idea.

The CSAIL team developed software that allows users to customize how the zipper behaves once assembled. Depending on the design of the arms, the mechanism can form straight rods, arches, coils, or twisted screw-like structures. The system, including the three arms and the slider, was fabricated entirely by 3D printing using common polymer materials.

The engineering principle behind the system is relatively straightforward: triangles are inherently rigid. Structural engineering has relied on triangular geometry for decades in bridges, cranes, towers, and trusses because triangles resist deformation far better than flat or rectangular structures. The Y-Zipper exploits that same principle by forcing three flexible arms into a triangular configuration during closure, essentially assembling a lightweight structural beam on demand.

That ability to switch between soft and rigid states is particularly relevant for robotics and deployable systems. Engineers often struggle to combine flexibility and structural stiffness within the same mechanism. Soft robotic systems adapt well to unpredictable environments but often lack strength, while rigid systems provide stability at the cost of flexibility. MIT’s design attempts to combine both.

The researchers demonstrated a robotic quadruped with legs capable of changing height and stiffness by actuating the zipper mechanism with motors. Such systems could help robots navigate uneven terrain by dynamically adjusting limb geometry in response to the environment.

The team also tested the system in deployable structures. In one demonstration, they used the Y-Zipper to rapidly assemble a tent-like structure, with the three-sided mechanism serving as both the structural support frame and the joining system. According to the team, setup time dropped from roughly six minutes to one minute and 20 seconds because the zipper effectively snaps the structure into place.

Medical applications are another possible target. The researchers created a wrist-cast prototype that wrapped the mechanism around a wrist cast, allowing users to loosen it during the day for comfort before tightening it again at night for support.

Beyond engineering applications, the system can also produce dynamic moving structures for art and design. One prototype resembled a mechanical flower that “bloomed” as a motor zipped the structure upward.

Durability testing showed the mechanism surviving roughly 18,000 zip-and-unzip cycles before failure. According to the researchers, the structure’s elastic behavior helps distribute stress across the assembly instead of concentrating it in a single area.

The team evaluated versions of the structure made from popular 3D-printing materials, polylactic acid (PLA) and thermoplastic polyurethane (TPU). PLA handled heavier loads more effectively, while TPU provided greater flexibility. Future versions could use stronger materials such as metal and scale to much larger sizes. Researchers also suggested possible aerospace applications, including deployable spacecraft structures and robotic systems capable of grabbing rock samples during exploration missions.

The work was presented at the ACM Conference on Human Factors in Computing Systems (CHI) in April and detailed in a paper titled "Y-Zipper: 3D Printing Flexible–Rigid Transition Mechanism for Rapid and Reversible Assembly."

“*** Nothing to see here – carry on ***”

Before looking more closely at fast16, it is interesting to ponder who may be behind it and the origin of the name. Sentinel Labs notes that the name ‘fast16’ can be found referenced in an infamous NSA ‘territorial dispute’ file leak. Specifically, it was mentioned in the strongest terms in a do-not-disturb list provided to operators. The line “fast16 *** Nothing to see here – carry on ***” singles out fast16 as being one of - if not the - most important NSA hack tools.

The security researchers, including Vitaly Kamluk & Juan Andrés Guerrero-Saade, found fast16 based on an architectural hunch. As a number of high-tier threats in this category were built on an embedded Lua virtual machine, they decided to see if there were traces of earlier Lua VM tools.

A file called svcmgmt.exe, which was uploaded to VirusTotal nearly a decade ago, would be a key link. This ‘unremarkable’ file was a 2005 file that was indeed a “Lua-powered service binary.” However, “it still receives almost no detections: one engine classifies it as generally malicious, and even that with limited confidence,” note the security researchers.

How fast16 was delivered

The aforementioned svcmgmt.exe acts as a carrier worm for delivering the fast16.sys kernel driver. It is surprisingly stealthy for a tool of its age. For example, it would check the machine registry for signs of malware monitoring tools from companies like Symantec, TrendMicro, McAfee, etc., to decide whether to abort or to deploy.

Spreading of fast16 would occur via wormlets propagating through Windows service control and file-sharing APIs. This version of fast16 targeted Windows 2000 and Windows XP environments and preyed on default and weak admin passwords on file shares.

The prime targets of fast16

Fast16 was designed to corrupt floating-point calculations in a subtle, predictable, reproducible way. It would seek out executable files, and in particular, EXEs that had been compiled with the Intel C/C++ compiler.

The corruption of output from targeted executables was controlled in such a way that fast16 would introduce “small but systematic errors into physical‑world calculations.” In effect, engineering projects based on these calculations may degrade more quickly than expected “or even contribute to catastrophic damage,” note the researchers.

In the Sentinel Labs blog, three era-appropriate software packages were specifically named as targets of fast16.

LS‑DYNA 970 (crash/explosion simulations; typically used in nuclear-related modeling)

PKPM (Chinese structural engineering suite, used to design expansive infrastructure projects)

MOHID (Portuguese hydrodynamic environmental modeling software)

Other infected machines using the same software, doing the same calculations, would get the same subtly erroneous results.

What else is out there?

Fast16 is a rather momentous discovery that indicates state-grade cyber sabotage existed in the mid-noughties, predating the discovery of Stuxnet by at least five years.

The lineage of fast16 may be much longer and deeper in history, though. Some strings in the malware files have fingerprints from Cold War-era Unix systems. These are basically fossilized traces of software revision control systems dating back to the 1970s and 80s.

The team, led by Dr. Babak Bakhit from Cambridge's Department of Materials Science and Metallurgy, engineered a multicomponent thin film that forms an internal p-n junction, enabling the device to switch states smoothly at currents below 10 nanoamps while producing hundreds of distinct conductance levels.

Memristors are two-terminal devices that can store and process data in the same physical location, eliminating the energy-intensive data shuttling between separate memory and processing units in conventional computer architectures. Neuromorphic systems built from memristors could reduce computing power consumption by more than 70%, according to the paper.

Most existing HfO2-based memristors rely on filamentary resistive switching, where conductive paths grow and rupture inside the oxide. These filaments exhibit stochastic behavior, resulting in poor device-to-device and cycle-to-cycle uniformity that limits computational accuracy.

A different approach - adding strontium and titanium

The Cambridge team took a different approach by adding strontium and titanium to hafnium oxide and depositing the film in a two-step process, thereby creating a p-type Hf(Sr,Ti)O2 layer that self-assembles a p-n heterointerface with an underlying n-type titanium oxynitride layer. Resistance changes occur by shifting the energy barrier height at this interface rather than by growing or breaking filaments.

"Filamentary devices suffer from random behavior," Bakhit said in a Cambridge press release announcing the work. "But because our devices switch at the interface, they show outstanding uniformity from cycle to cycle and from device to device."

The devices demonstrated switching currents at or below 10^-8 amps, retention exceeding 10⁵ seconds, and endurance beyond 50,000 pulse-switching cycles. Using identical 1.0 V spikes comparable to biological neural signaling, the researchers achieved a conductance-modulation range exceeding 50 times across hundreds of distinct levels without saturation.

Synaptic update energy ranged from approximately 2.5 picojoules down to around 45 femtojoules. The devices also reproduced spike timing-dependent plasticity and maintained stable synaptic operation across roughly 40,000 electronic spikes.

One significant hurdle remains

The current deposition process requires temperatures of around 700°C, which exceeds standard CMOS manufacturing tolerances. "This is currently the main challenge in our device fabrication process," Bakhit said. "But we're now working on ways to bring the temperature down to make it more compatible with standard industry processes."

All materials used in the device stack are fully CMOS-compatible, and a patent application has been filed through Cambridge Enterprise.

The FlyTrap attack was demonstrated in the field, literally, by one of the researchers in the video above. A specially formulated AI-generated pattern is displayed by opening an umbrella after a drone locks on to track the person in the field. Specifically, the visual pattern performs a next-gen physical distance pulling (PDP) attack that works across multiple angles, even in motion, and in the real-world. This demo works on three commercial drones that were tested, the DJI Mini 4 Pro, the DJI Neo and the HoverAir X1.

The umbrella-printed visual physically draws victim drones closer as its neural network tracking systems interpret the pattern to be the subject moving further away. As the drone approaches the umbrella, the pattern causes the targeting bounding box to continue shrinking - so the drone moves to get closer. Autonomous drones lured by the pattern can then easily be ensnared using a net gun, or further induced to crash to Earth.

The research shows that FlyTrap is significantly more effective than prior adversarial‑ML techniques like older PDP tech and Targeted Gradient Transfer (TGT).

“Autonomous target tracking represents both tremendous potential and significant risk,” said paper co-author Alfred Chen, UC Irvine assistant professor of computer science. Chen reminds us that autonomous drones are used in areas like border patrol and public safety, but also by malicious actors.

Lead author, Shaoyuan Xie, a UC Irvine graduate student researcher in computer science added that “Our findings highlight urgent needs for security improvements in [autonomous target-tracking] systems before wider deployment in critical infrastructure."

It would be fascinating to see or hear about FlyTrap being used in the real-world. However, both DJI and HoverAir have been responsibly notified about the neural processing vulnerabilities in their autonomous tracking systems.

DART is the first-ever mission that looks at the feasibility of conducting a kinetic strike to deflect any celestial threat to the Earth, and it seems that it was quite successful. The probe’s impact ejected a significant amount of debris from the asteroid, essentially doubling the DART spacecraft’s effectiveness. Even though only the smaller Dimorphos was targeted, the changes to its momentum and orbit also affected the larger Didymos.

The numbers mentioned might seem minuscule, especially when compared to the size of these celestial bodies, but NASA is banking on detecting any collision hazards to our home planet early on. By affecting its trajectory as soon as it's detected, even a small change can make a huge impact over the massive distances involved in space.

DART can make the difference

“The change in the binary system’s orbital speed was about 11.7 microns per second, or 1.7 inches per hour,” Rahil Makadia of the University of Illinois Urbanas-Champaign, the lead author of the study on the DART mission, told NASA. “Over time, such a small change in an asteroid’s motion can make the difference between a hazardous object hitting or missing our planet.” Although the DART’s target is pretty small and won’t destroy the entire planet, it still has the potential to wipe out an area the size of London and its surrounding areas. However, its larger companion is a bigger threat, with the capability of destroying the entirety of Southern England.

Thankfully, these two asteroids would’ve never impacted Earth, even after the DART mission. Still, NASA is concerned that there might be several other smaller rocks floating in space that could pose a danger to the planet. Because of this, the U.S. space agency is now building an orbiting telescope designed to detect incoming threats like asteroids and comets — the first such satellite designed for planetary defense. That way, it can find these early, and if there’s one that could potentially impact our home, NASA could send a kinetic impact vehicle early on to change its direction.

No one seems to have a hard figure for the biggest bit flip contributor. However, systems sent to space will use specialized components hardened to resist interference from cosmic radiation, extremes of temperature, etc, and to include aggressive error checking.

The Mozilla team received nearly half a million auto-submitted crash reports last week (opt-in feature). Data from a recently introduced “memory tester that runs on user machines after the browser crashes,” guided senior engineer Gabriele Svelto towards his eyebrow-raising bit flips, causing 15% of crashes figure, which he admits “dwarfs all the previous estimates I saw regarding this problem.”

Mozilla employed, self-confessed ‘old school nerd’ Svelto says that an initial 10% estimate was revised up because “If I subtract crashes that are caused by resource exhaustion (such as out-of-memory crashes) this number goes up to around 15%.” Moreover, it was determined that one in two bit flip crashes was due to a “genuine hardware issue.” Svelto notes this could be undershooting the real figure as Mozilla’s memory test on crash feature “only checks up to 1 GiB of memory and runs for no longer than 3 seconds.”

Thus, it is hard to drill down to probably the most fascinating statistic of how many Firefox crashes are precipitated by an errant cosmic ray passing though…

As a parting shot, Svelto doesn’t want fancy Arm-based MacBook owners, or any other niche device owner, thinking this is just about PCs with shaky RAM. Every device with memory can be affected by bit flips, asserts the engineer. That doesn’t matter if it’s a Mac, smartphone, or even a printer or router. From that perspective, PC desktop DIYers are probably in a better position, as they can replace any faulty component without throwing away the whole caboodle. But please, great cosmic ray caster in the sky, please don’t make me think I have RAM issues in 2026…

Yaghi’s company, Atoco, also sees a market in “personalized water,” much like where households generate their own off-grid power from wind or solar. Prototypes have been successfully tested in places as arid as Death Valley. The 1,000 liters a day machine is far bigger than the social media prototype machine image we see alongside the Professor in the desert, at around 20ft in length, or the size of a shipping container.

Metal-Organic Frameworks (MOFs)

The science behind this new clean water harvesting machine is based on reticular chemistry, one of Yaghi’s specialist areas. In particular, the device is packed with Metal-Organic Frameworks (MOFs), which are synthetic porous materials engineered at the molecular level to have huge surface areas. A few grams of an MOF can have a surface area equivalent to a football arena, according to the source.

With this incredibly large surface area in such a compact form, the MOF is adept at collecting water from the air and condensing it to a liquid. Yaghi’s mechanism can do this without a power source. It uses the wind and air for water input, then the sun to drive condensation and evaporative action. It is worth repeating – the invention can operate as a self-contained, entirely off-grid device.

Atoco also touts the availability of on-grid solutions, though.

Inspired by hardship

The MOF Water Harvester device was inspired by Yaghi’s personal history, growing up in a refugee community in Jordan, where his family would listen out for the arrival of the water truck, hoping to beat the rush to fill their life-sustaining containers. The UN has been warning about water insecurity, or even water bankruptcy, for years, so inventions like this could have sizable positive impacts for humanity.

In the video embedded above, we see that eFlesh is quick and easy to print. 3D printer watchers will notice that the scientists used a Bambu Lab X1E 3D printer for the square sensor-grip example. Other videos show that the TPU filament 3D printing process was paused at about half an inch from the bed to insert a grid of magnets – key to its sensing functionality.

The scientists behind this innovation say that “Building an eFlesh sensor requires only four components: a hobbyist 3D printer, off-the-shelf magnets (costing less than $5), a simple CAD model of the desired shape, and a magnetometer circuit board.” They have decided to promote broad accessibility of this technology by releasing an open source design, including the 3D-printable STLs, and supporting software code. There’s even a CAD-to-eFlesh STL conversion tool available.

Robot sensing, moving, and manipulation of irregular or fragile objects has long been a conundrum for designers. That this solution is accessible, and purportedly cheap and effective, making it all the more welcome. Specifically, the New York University scientists say that eFlesh has “contact localization accuracy of 0.5 mm, with force prediction errors of 0.27 N along the z-axis and 0.12 N in the x/y-plane.”

Some canny coding is behind the “learning-based slip detection model that generalizes to unseen objects with 95% accuracy.” And the developers of eFlesh say that their “visuotactile control policies that improve manipulation performance by 40% over vision-only baselines.” This smart functionality is demonstrated in precise tasks like robotic arm plug insertion and credit card swiping.

So, these scientists may have just democratized tactile robotic sensing and manipulation. Shame about the name, though.

It may sound like an exercise in silliness, unless you are a believer, but this GitHub project does a great job demonstrating the power and flexibility of sched_ext. In brief, sched_ext is a relatively new pluggable framework that allows for custom Linux schedulers to be implemented without kernel patching. We are sure the feature wasn’t designed to assign CPU time and other system resources based on the movements of heavenly bodies – but it can.

Zampieri, who is a software engineer at Red Hat specializing in RISC-V architecture projects, discusses the features and implementation of his scx_horoscope on the afore-linked GitHub repository. However, I must confess that the astrological scheduling rules, which weigh the complex mix of planetary domains, the zodiac, element effects, retrograde effects, and more, to balance the priorities of your CPU, networking, system, and memory tasks, are hard to fathom. That complexity makes sched_ext all the more impressive.

The dizzying arrays of mumbo jumbo behind this scheduler almost encourage blind acceptance in its prioritizing decisions. After reading the GitHub, head spinning as I thought about the ever-moving boosts and debuffs of elemental effects, I almost felt hypnotized into believing. I was ready to accept, as Zampieri puts it, “if the universe can influence our lives, why not our CPU scheduling too.”

To sum up, it is a rather extraordinary piece of work to interweave astrology and a fully functional operating system scheduler. It actually draws on resources like accurate geocentric planetary positions, lunar phases, cosmic weather reports, and dynamic time slicing to come up with its scheduling adjustments, too.

Zampieri is clear that this GPL-2.0 licensed project is a “scientifically dubious, cosmically hilarious” work. It definitely isn’t recommended for use in production systems - not because of bugs, but because it works as intended… The dev is still looking to add “more cosmic chaos” to scx_horoscope, so contributors are welcome.

Neuromorphic computing seeks to mimic brain-like processing to take advantage of billions of years of animalian evolutionary advances. The theory is that this style of computing architecture inherits brain-like efficiency, adaptability, and parallelism.

Memristors are thought to be a good match for brain-like neuromorphic architectures. The Ohio researchers praise them for their unique electrical properties and “ability to mimic neural functions.” They can be “efficient and self-adaptive in situ learning.” Such abilities make memristors a good fit for applications like robots and autonomous vehicles.

Adding to the appeal of memristors is their characteristically low power consumption, and the “advantage of integrating memory and processing capabilities into a single device,” say the Ohio team.

Moving along to how fungal electronics melds with memristor technology, and the scientists highlight that shiitake mushroom mycelium is a sustainable, low-cost, and biodegradable alternative to rare-earth packed and expensive to produce memristors. This variety of mushroom, in particular, is also claimed to deliver adaptive electrical signaling “akin to neuronal spiking.”

Methods and testing

After cultivating these mushrooms, scientists dried and rehydrated shiitake mushrooms to create a consistent level of conductivity without introducing bulk moisture. Then, the memristive qualities of these mushrooms was tested and verified in the Ohio labs.

In the tests across a range of voltages, waveforms, and frequencies, the value of shiitake mycelium memristors became clear. The key finding was that “when used as RAM, our mushroom memristor was able to operate at up to 5,850 Hz at an accuracy of 90 ± 1%.”

Potential applications

The Ohio researchers mention that shiitake mycelium memristors offer potential in edge computing, aerospace, and embedded firmware applications. There, the low energy consumption, light weight, and radiation resistance of these memristors can all be potentially compelling advantages. However, don’t overlook the green computing credentials, with these memristors sustainably “bridging bioelectronics and unconventional computing.”

Whilst you become accustomed to these commands it’s good to work with example test files and directories and you should take extra care to ensure you are carefully following the instructions.

All the commands in this how-to will work on most Linux machines. We’ve used an Ubuntu LTS install but you could run this how-to on a Raspberry Pi. All of the how-to is performed via the Terminal. You can open a terminal window on most Linux machines by pressing ctrl, alt and t.

Working with TAR archives

When using Linux systems you are likely to come across tar archives as well as ZIP archives. It’s useful to practice creating .tar archives and also how to extract from a tar archive. Also of note is that sometimes TAR archives are referred to as tarballs. Tar archives in their standard form have the .tar suffix but these archives are not compressed. Compression is added using different tar compression tools which is why you will see tar archives with extra suffixes such as .tar.xz or .tar.gz.

To create or extract from these compressed archives you need additional arguments added to the tar command. We’re going to use two of the most popular compression methods, gzip and bzip2 along with a tar archive that has no compression.

Using a TAR archive

1. Open a new terminal window. This will open to our home directory.

2. Create a .tar file. Using test_directory as a target we’ll make a standard uncompressed .tar archive. This kind of archive is useful to group small files, such as logs into a single archive.

$tar cf test_archive.tar test_directory

Extracting TAR archives is straightforward. Instead of adding the c argument to create an archive we replace it with the x argument. We again need to add arguments that respond to the type of compression tool the archive was created with. When we extract from our 3 archives we would be creating duplicates of the test_directory contents so we will delete this directory each time we extract the next archive.

1. Delete the test_directory folder.

rm -r test_directory

2. Extract the standard .tar archive. After extracting the archive use ls to check the archive has been extracted.

tar xf test_archive.tarls

Using Gzip Compressed TAR Files

A gzip compressed tar archive (tar.gz) is one of many popular compression tools for TAR archives and it is common to find a .

1. Create a gzip archive by adding the z argument to the tar command.

tar czf test_archive.tar.gz test_directory

To extract a gzip compressed tar archive we need to add the z argument to identify that the archive uses the gzip compression method.

2. Delete the test_directory folder.

rm -r test_directory

3. Extract the gzip compressed tar archive using the z argument. Then list the directory contents to show that test_directory has been created.

tar xzf test_archive.tar.gzls

Using Bzip2 Compressed TAR Files

Another alternative compression method for a tar archive is bzip2, which is invoked using the j argument. Archives typically end with either a tar.bz2 or tbz suffix.

1. Create a bzip2 archive by adding the j argument to the tar command.

tar cjf test_archive.tar.bzip2 test_directory

To extract a bzip2 compressed tar archive we need to add the z argument to identify that the archive uses the gzip compression method.

2. Delete the test_directory folder.

rm -r test_directory

3. Extract the bzip2 compressed tar archive.

tar xf test_archive.tar.bzip2lsrm -r test_directory

Extract TAR Archives to a Specific Location

If we wish to extract an archive to a specific location we can use the C argument with the tar command and then specify the location. The location can be a relative or absolute path. So we can extract the archive to a sub directory inside a current directory, or we can specify the full path to another location in the file system.

Extract test_archive.tar into the Music directory. Here we are in the Home directory, and most Linux distributions feature a Music directory which we can extract the archive to.

tar xfC test_archive.tar Musiccd Musicls

To extract to another location in the file system it is best practice to use an absolute path.

Extract test_archive.tar to your desktop directory. Specify the full path, tab completion can be used to auto-complete directory names. Remember to change to match your own.

tar xfC test_archive.tar /home//Desktop/cd /home//Desktop/ls

Armed with these few examples you are now capable of extracting most common archives on the command line. Whilst GUI tool options exist for some archives, often when dealing with a compressed .tar archives, these terminal commands are much quicker and easier to perform.

While you become accustomed to these commands, it’s important to take care that you don’t kill a process which results in you losing work. Wherever possible use an application that isn’t currently in use on your machine as an example of a process to kill.

All the commands in this how-to will work on most Linux machines. We’ve used a Ubuntu 20.04 install but you could run this how-to on a Raspberry Pi. All of the how-to is performed via the Terminal. You can open a terminal window on most Linux machines by pressing ctrl, alt and t.

How to Kill a Known Application in Linux 

To begin, let's open a software application and then, in the terminal, use the killall command to kill the application. We have used Inkscape, the free and open source vector graphics package as an example, but it doesn’t really matter what you use, just so long as you aren’t depending on it for anything at the current time.

Using the killall command kills all processes linked to the named application. Later we will look at using the kill command to kill a specific process.

1. Launch Inkscape, or another test application, from the system gui.

2. In the terminal, use the killall command to kill all Inkscape processes. You should see that Inkscape closes immediately when this command is issued.

killall inkscape

Identifying Different Processes in Linux

Often, software applications have numerous processes running and it can be useful to try and work out which one is likely causing the problem. For example, it could be one browser window rather than all of Chromium or Firefox.Then you can close the problematic process while leaving the rest of the app running. We can use the top tool in the terminal to look at, interact and then kill the rogue processes.

1. Launch the top tool in the terminal. When it’s launched, you will see the terminal populate with information regarding all the processes and their status. The second line of the top screen shows the total number of tasks, which are currently running and which are sleeping, stopped or zombie status. Below this you will see a list of processes and this information will refresh at a default of once every three seconds.

top

2. Press i to constrain the current processes list to those currently active. This can be very useful if you are only interested in current running processes and makes the list easier to read.

3. Press z to add color to the top interface to make it easier to read. Because this is a text based interface, adding color can help distinguish between labels and data.

4. Launch Inkscape or another non essential application.

5. Find the listing for Inkscape and note the PID. The PID is the number assigned to identify the process and they are listed in the first column of the top output. In our example the PID for inkscape was 4582 but it will be different each time Inkscape is running.

6. Press k and then type the PID number to kill the process.

7. Press ENTER and you should see a message “Send pid 4582 signal [15/sigterm]” which is confirming you want to kill this process. Press “enter” once more to confirm and kill the process.

8. Press q to close top and return to the terminal emulator. Often, you will want to look at top for a quick glimpse at the processes and then return to the terminal.

Killing an Individual Process Outside of top

Occasionally you might want to kill an individual process outside of top. We can use the PID number found via top in combination with the kill command to do this. Note that the killall command we used earlier won’t work with an individual PID.

1. Launch Inkscape or another application. The choice of application doesn’t really matter, just make sure you aren’t working in the application.

2. Launch top in the terminal. Press i and z to constrain the results and add color to the top theme.

top

3. Identify and note the PID for the application to kill. In our example the PID for Inkscape was 4582.

4. Exit top by pressing q.

5. Use the kill command and the PID number to kill the Inkscape process from the terminal emulator. Replace the example PID value below with your own identified PID.

kill 4582

Using top provides an excellent set of detective tools to identify misbehaving processes and can often allow you to clinically close an unresponsive process whilst possibly saving other data in an application.

Killing Processes Using the ps Process Report

There is an alternative to top. We can use the ps command, designed to produce a report of all the running processes.

1. Launch Inkscape or another application. The choice of application doesn’t really matter, just make sure you aren’t working in the application.

2. Run the ps command with two arguments. The -e argument shows all the processes and the -f argument sets the format to the system standard syntax.

ps -ef

3. Use grep to filter the report. searching just for “inkscape”. The typical ps -ef output is extremely dense, packed with dozens of running processes. Using a filter we can search for the application name. The PID is the second column, just after the username. You may see multiple PIDs for the application.The one that we are interested in just shows the application name.

ps -ef | grep inkscape

4. Use the kill command along with the PID to kill the process. In our case Inkscape had a PID of 19166.

kill 19166

Process management is a key Linux skill to learn and it’s worth taking the time to practice these approaches before you need to deploy them in a real situation.

Bash scripts behave like those of other programming languages. The code runs line by line in a sequence. We can use conditional logic (selection) to alter the path that our code takes and the code can iterate and repeat the execution of the code until a condition is met or run the code indefinitely.

Bash, the Bourne Again SHell is one of many different shells available to Unix and Linux users (Zsh, Ksh, Tcsh, Fish are others) and is commonly installed with many of the main distributions. We ran Kubuntu when writing this tutorial, but the same code will also work on a Raspberry Pi (with Raspberry Pi OS) or almost any other Linux distribution.

In this how-to we shall take our first steps into creating Bash scripts with three example projects. The first is the venerable “Hello World”, then we learn how to create interactive scripts, before finally creating a real script to check the status of a site / server. We then go the extra mile and learn how to make our scripts executable and available system-wide.

How to Write a “Hello World” Bash Script

We’ll start simple with “Hello World”. This may seem trivial, but a hello world test is useful to understand the workflow of creating a Bash script, and to test the basic functionality. We’ll be using the nano text editor, but you can also write your scripts in another terminal or GUI based text editor.

1. Create a new file, hello.sh and open it with nano.

nano hello.sh

2. On the first line specify the interpreter to be used in the code. In this case it is Bash.

#!/bin/bash

3. On a new line use echo to print a string of text to the screen. Those familiar with Windows batch scripts will spot that Windows and Linux both use "echo" to print to the screen.

echo “Hello World”

4. Save the code by pressing CTRL + X, then press Y and Enter.

5. Run the code from the terminal.

bash hello.sh

The output of the command is a single line of “Hello World.” This proves that our script works and we can move on to something a little more interesting.

Capturing User Input in Bash Scripts

Bash is a fully formed language, and with it we can create interactive tools and applications. In this example we will create a greeting generator which will capture the username and then choose a random greeting from an array of greetings.

1. Create a new file, random-greetings.sh using nano.

nano random-greetings.sh

2. On the first line specify the interpreter to be used in the code. In this case it is Bash.

#!/bin/bash

3. Add an echo to print a message to the user.

echo "Welcome to the Tom's Hardware Greeting Generator"

4. Read the user’s keyboard input, using a prompt to ask a question to the user. The user’s input is saved to a variable name.

read -p "What is your name?: " name

5. Create an array greeting which will store five greetings. An array is a list of items stored with an index number. Using the name of the array and the index number we can pull items (greetings) from the array.

greeting[0]="Hola"greeting[1]="Greetings"greeting[2]="How do"greeting[3]="Hey there"greeting[4]="Howdy"

6. Create a variable size which will store the number of items in the array.

size=${#greeting[@]}

7. Create a variable index and store a randomly chosen number from zero to the size of the array, in this case five items.

index=$(($RANDOM % $size))

8. Use echo to print the randomly chosen greeting along with the user’s name.

echo ${greeting[$index]} $name

9. Save the code by pressing CTRL + X, then press Y and Enter.

10. Run the code from the terminal. Follow the instructions to see the greeting.

bash random-greetings.sh

Using Arguments with Bash Scripts

The beauty of the terminal is that we can pass arguments, extra instructions or parameters, to a command. In this example we will write a bash script to check the status of a URL, passed as an argument.

1. Create a new file site-check.sh and open nano.

nano site-check.sh

2. On the first line specify the interpreter to be used in the code. In this case it is Bash.

#!/bin/bash

3. Using echo, write a message to the user to advise them on the current HTTP response code for the URL. The value of $1 is the URL that we specified as an argument.

echo "The current HTTP response for " $1 " is..."

4. Create a variable, response to store the output of the curl command. Using the –write-out argument we specify that we want to see the HTTP response code. Another argument, –silent will ensure that we do not see all of the output from the command spill onto the screen. We then send any residual output such as errors to /dev/null, essentially a black hole for data. Lastly we use the $1 argument (our URL to check) to complete the command.

response=$(curl --write-out "%{http_code}\n" --silent --output /dev/null "$1")

5. Use echo to print the HTTP response code to the terminal.

echo "${response}"

6. Save the code by pressing CTRL + X, then press Y and Enter.

7. Run the code from the terminal; remember to include a full domain name to check.

bash site-check.sh https://tomshardware.com

Creating a system-wide executable Bash Script

Right now our scripts work well, but they are not available system-wide as an executable file. Let's change that. For this final part we shall convert the site-check.sh script into an application that we can use from any location on the drive.

First we need to make the file executable, then we can copy the file to /usr/bin to create an executable available anywhere.

1. Open a terminal window and navigate to site-check.sh.

2. Use chmod to set the file as executable. The +x argument specifies that the file should be executable.

chmod +x site-check.sh

3. Run the script. Using ./ we instruct the terminal to run an executable file in the current directory.

./site-check.sh https://bigl.es

4. Copy site-cehck.sh to the /usr/bin/ directory, and change the target filename to remove the .sh extension. Note that we will need to use sudo as the /usr/bin directory is not owned by our user. Take a look at our tutorial on user, file and folder permissions for more information.

sudo cp site-check.sh /usr/bin/site-check

5. Run site-check to test that the command works as expected. Remember to include a URL to check.

In this how-to we will use the grep command and commonly added arguments to search for specific strings of data within files. We’ll begin by setting up a small directory of test files as searching an entire filesystem can take some time and create a lot of results.

All the commands in this how-to will work on most Linux machines. We’ve used an Ubuntu LTS install but you could run this how-to on a Raspberry Pi. All of the how-to is performed via the Terminal. You can open a terminal window on most Linux machines by pressing ctrl, alt and t.

Set Up a Test Environment for grep 

As grep can be used in lots of different ways we need to set up a directory and some content that allow us to explore its uses. In this section we will create this test environment.

1. Set up a test directory and change directory so that you are inside it.

mkdir testcd test

2.Create 4 files, test1, test2, test3 and test4.

touch test1 test2 test3 test4ls

3. Edit test1 using nano to contain the following names on separate lines. Note that in test1 none of the names contain capitals. After editing in nano press control x to exit, press y to confirm the save and then press enter.

nano test1

4. Add the following text to the file. Then press CTRL + X, then Y and Enter to save and exit.

alimohamedclairealedsteve

5. Edit test2 using nano. In test2 we will add a single longer line of test containing the name steve.

nano test2

6. Add the following text to the file. Then press CTRL + X, then Y and Enter to save and exit.

this is a long line of test that contains the name steve

7. Edit test3 in nano. Similar to test1 we will add a list of names on separate lines but this list will include the name steven.

nano test3

8. Add the following text to the file. Then press CTRL + X, then Y and Enter to save and exit.

alicegeoffmurbarakmohamedsteven

9. Finally edit test4 to complete our test environment. Note that in this file we are using a capital letter at the start of Steve.

nano test4

10. Add the following text to the file. Then press CTRL + X, then Y and Enter to save and exit.

Steve ?

Simple Searches with grep 

Searching a file for a specific string is extremely useful. For example when debugging an error in a log file. Let’s start using grep in its most basic form.
The grep command syntax is simply grep followed by any arguments, then the string we wish to search for and then finally the location in which to search.

1. Search test1 for the string steve using grep. The search criteria is case sensitive so ensure that you're searching correctly. Notice that the word steve is simply returned in red to indicate it has been found in the file.

grep steve test1

2. Search for the same string in multiple files. We can simply add a list of files to the grep command for it to search. Notice that, with multiple search areas, the returned results are tagged with each filename from which the result is found. Also notice that grep returns the complete line of text that contains the searched string.

grep steve test1 test2

3. Search all files within the directory. Adding an asterisk forces grep to search all files within the current directory. Notice that the returned results include the result from test3 where the search string steve is contained within Steven. Note also that these results don’t contain the result from test4 as in its basic form the grep command is case sensitive.

grep steve *

4. Add the argument -i to make grep case insensitive. This will return results from all four of the test files in the directory.

grep -i steve *

Piping output to grep

The strongest use case for grep is when it is paired with another command. Using pipes we send the output of a command to grep and use it to search for patterns / keywords.

1. Open a new terminal window.

2. Use lsusb to list all of the USB devices attached to your machine. This will also list internal USB devices, such as laptop webcams. The output will differ depending on your machine, but you should be met with a wall of text.

lsusb

3. Use the lsusb command again, but this time use grep to search for Linux. By adding a pipe between lsusb and grep the output of the first command is used as the input of the second.

lsusb | grep Linux

Using dmesg and grep to Inspect the Kernel Ring Buffer

Let's try something a little more complex. This time we’ll use dmesg and grep to inspect the Kernel Ring Buffer (essentially the kernel’s log file). We are going to search for a keyword in dmesg, “secureboot”, and confirm that it is enabled.

1. Open a terminal and run the dmesg command using sudo. This will print a wall of console output to the terminal, something that we can search using grep.

sudo dmesg

2. Use the grep command to search for “secureboot” in the dmesg output. Use the -i argument to turn off case-sensitivity so that we catch every occurrence of secureboot. The output will show the lines at which secureboot appears in dmesg.

sudo dmesg | less | grep -i secureboot

Other Uses of grep 

Like many Linux commands there are many useful additions and variants for the grep command. Let’s look at a couple of interesting examples.

1. Perform an inverted search using the -v argument. This will return a list of every line from the test environment files that doesn’t contain the search string steve. This argument is useful to dismiss occurrences of strings in logs or files when debugging an issue. Note that again the results are case sensitive and therefore they include the line containing the capitalized Steve ? from test4.

grep -v steve *

2. Combine the -v and -i arguments to exclude all matching strings regardless of case.

grep -vi steve *

3. Search for a string that contains non alphanumeric text or spaces. If you include a search string with a space or other non alphanumeric text this can break the grep command syntax, to create a search string containing these you need to use quotes to contain the string. In this step we are searching for “Steve ?” which is contained in the test4 file.

grep “Steve ?” *

Searching Subdirectories with grep

Like many Linux commands there are many useful additions and variants for the grep command. Let’s look at a couple of interesting examples.

1. Perform an inverted search using the -v argument. This will return a list of every line from the test environment files that doesn’t contain the search string steve. This argument is useful to dismiss occurrences of strings in logs or files when debugging an issue. Note that again the results are case sensitive and therefore they include the line containing the capitalized Steve ? from test4.

grep -v steve *

2. Combine the -v and -i arguments to exclude all matching strings regardless of case.

grep -vi steve *

3. Search for a string that contains non alphanumeric text or spaces. If you include a search string with a space or other non alphanumeric text this can break the grep command syntax, to create a search string containing these you need to use quotes to contain the string. In this step we are searching for “Steve ?” which is contained in the test4 file.

grep “Steve ?” *

Searching Subdirectories with grep

Often we will want to search for a string in the files contained in sub directories. We can do this simply by adding the -r recursive argument to the grep command.

1. Create a subdirectory containing a test file within the test directory.

mkdir sub_directorycd sub_directorytouch test5

2. Open test5 using the nano text editor and add the text “steve in a sub directory” to the file. Then press CTRL + X, then Y and Enter to save and exit.

nano test5.

3. Return to the test directory and perform a search adding the -r option. Notice that the result for test5 includes the location of the file listed in the output.

cd ..grep -r steve *

With grep in Linux, you have a good collection of approaches for searching out file contents across your system. As with many Linux commands it can be worthwhile looking at the help menu to see all the varied arguments you can add to grep. Run grep --h in a terminal emulator to check out all the options.

MORE: How To Check Disk Usage in Linux

MORE: How To Kill a Process in Linux

MORE: How To Find Files in Linux

Managing passwords via the terminal is relatively simple. Users can change their own passwords and users belonging to the sudo (super user) account can administrate the passwords of other users.

In this how-to we’ll learn the basics of passwd, a command designed to manage passwords. We will also learn a few advanced arguments for this command, arguments which will enhance its use and make our lives easier.

These commands will work on most Linux machines. Our test PC ran Kubuntu, a derivative of Ubuntu, but you can also run through this how-to on a Raspberry Pi.
All of the how-to is conducted via the Terminal.

How to Change your Password in Linux

Any user can change their password at any time. It is good practice to frequently change your password, and to not reuse or rotate passwords.

1. Open a terminal.

2. Type in the passwd command to start the password change process. Passwd may look like a spelling mistake, but it is the command to work with passwords on the terminal.

$ passwd

3. Type in your current password and press Enter.

4. Type in your new password, press Enter. Then type it in again to confirm that it is correct. Note that Linux will not echo (print) the password to the screen, nor will it show any asterisk indicating password length.

Change Another User's Password

If you are an administrator of many users, there will come a time where a user forgets their password and it will fall to you, or someone with sudo privileges to reset their password.

To change another user's password we use the same command, but specify the user’s name.

1. Open a terminal.

2. Type in the passwd command along with the user name. To use this command you will either need to be root, or be part of the “sudo” group. In the code example we assume that you are in the sudo group.

3. Change the user’s password, and confirm the change.

Force a Password Reset

You’ve heard that one of your users has been reusing their passwords; this is bad. So let’s give them a chance to change their password. Using the -e argument we can enforce a password reset by expiring their password.

1. Open a terminal.

2. Type in the command and pass the -e argument to expire Tom’s password.

$ sudo passwd -e tom

When Tom next logs in, the system will force Tom to change their password. Here we have simulated Tom logging in by using the “su” command to switch our user to Tom.

There may come a time when a user account will need to be temporarily locked, preventing them from logging in. For this we can use the -l argument to lock the account. Here we are locking Tom out of their account.

1. Open a terminal.

2. Use the command with the -l argument to lock Tom’s account.

$ sudo passwd -l tom

When Tom next tries to login, their password is rejected and they receive an authentication error. Here we again simulate it using su to switch users.

3. To unlock Tom’s account we use the -u switch.

$ sudo passwd -u tom

Exploring the passwd Command

The passwd command has a number of useful arguments (parameters) that we can pass when using the command. Here are some examples.

Checking the status of a user’s password is a useful tool for system administrators when it comes to audits and housekeeping. Here our test account, Tom, is audited.

Open a terminal and run the command with the -S switch. This will show the status of the account.

$ sudo passwd -S tom

The output is formatted to show

The username

Password status Locked (L), No Password (NP), Password (P)

Date of last password change

Minimum password age

Maximum password age

Warning period (the number of days given to the user to change their password before it expires)

Inactivity period (number of days after a password expires before it is locked)

If we wanted to set the number of days for Tom’s warning period to 14 days, we would use this command.

sudo passwd -w 14 tom

If you want to list the password status for all users, then the -a switch is just the thing. This switch is used with -S to list the status of every user, even the users that you never knew existed. These extra users are used for specific tasks, such as printers and networking.

Here we can see all the users on our test machine, but the two “real” users are Les and Tom.

Whilst you become accustomed to these commands it’s good to work with example test files and directories and you should take extra care to ensure you are carefully following the instructions.

All the commands in this how-to will work on most Linux machines. We’ve used an Ubuntu LTS install but you could run this how-to on a Raspberry Pi. All of the how-to is performed via the Terminal. You can open a terminal window on most Linux machines by pressing ctrl, alt and t.

Creating a Zip Archive in Linux

To begin, let's create some empty test files and then add them to an archive. We’ll use the touch command to create files, the ls command to check that the files have been created and then the zip command to create a basic archive.

1. Create a test directory where we can safely try out the command. Change directory so that you are inside the new directory.

mkdir test_directorycd test_directory

2. Create some test files to add to our archive. Note the 3 different file suffixes.

$ touch test1.txt test2.h test3.c

3. List the files to check that they have been created correctly.

ls

4. Using the zip command, create a ZIP archive of the newly created files. To create a ZIP archive, we simply use the command zip followed by the name we are giving the archive and then a list of the files we wish to include in the archive. The following command will create a ZIP archive called test_archive.zip and the archive will include the three test files. Note that we don’t need to add the archive suffix .zip to the name of the archive when using the zip command.

zip test_archive test1.txt test2.h test3.c

5. Check that the ZIP archive has been created.

ls

Creating a ZIP Archive Selecting Only Specific File Types

It’s possible to supply the zip command with arguments that only add specific file types to the archive. This is a powerful feature that is efficiently achieved on the command line.

1. Delete any previous test archives and check you still have the original three test files. Note that, unlike when creating the ZIP archive, you have to include the .zip file extension when referencing an existing archive.

rm test_archive.zipls

2. Create a new ZIP archive which contains only the test files ending .h and .c . After creating the archive, if you check its contents it will only contain the files test2.h and test3.c and won’t contain test1.txt.

zip test_archive *.h *.c

3. To check the contents use less.

less test_archive.zip

Create a ZIP archive Which Contains a Directory

Often we need to make a ZIP archive containing directories which contain files, we can do this by adding the recursive argument -r to the zip command.

1. Delete the archives inside the test_directory but keep the other files. To avoid confusion it would be good practice to delete the archives leaving only the original files that we created.

rm test_archive.zip

2. Go up one directory and check that you are not inside the directory to archive. We can’t be inside the directory that we wish to add to the archive.

cdpwd

3. Create a ZIP archive containing the test_directory directory and its contents. Notice when you create the archive using the -r argument that you see a verbose output detailing each stage of the command as it descends into the directory and archives the files and folder.

zip -r new_archive test_directory

4. Make another similar ZIP archive using the -q (quiet) argument to disable the verbose output. You should see that this time that no steps are reported to the terminal in the creation of this archive. Then use ls to check the new archive has been created.

zip -r -q new_archive2 test_directoryls

Change the Level of Data Compression for a ZIP Archive

By default, the zip command uses a compression value of 6 taken from a range of 0-9 with 0 being uncompressed and 9 being the maximum allowable level of compression. We can simply add a numerical argument to change the level of compression.

1. Add random data to the test files. Right now our test files test1.txt, test2.h and test3.c are empty. They have zero bytes of data but we can add random “garbage data” to the files using a quick terminal command that will dump 64MB into each file.

cd test_directoryhead -c 64M  test1.txthead -c 64M  test2.hhead -c 64M  test3.c

2. Return to the parent directory and create a new ZIP archive with the default compression level of 6.

cdzip -r -q archive_default_compression test_directory

3. Create another archive with the same contents but increase the compression level to 9. Check the details of the new archive and the previous archive using ls -l , you should see that the size of the new archive is smaller than the previous archive as the data compression has been increased.

$ zip -r -q -9 archive_compressed test_directory$ ls -l

Extracting Zip Archives

We will create a ZIP archive called test_archive.zip which contains a directory test_directory which in turn contains 3 empty test files, test1.txt, test2.h, test3.c . To extract this archive we will use the unzip command.

1. Create a test folder containing the test files. After creating the test directory and files check the files have been created using ls before returning to the parent directory.

mkdir test_directorycd test_directorytouch test1.txt test2.h test3.clscd

2. Create a ZIP archive called test_archive.zip that contains the test_directory. Check that the archive has been successfully created using ls. The zip command has two arguments, the name of the archive that we wish to create, and the source of the files to be put into the archive.

zip -r test_archive test_directoryls

3. Delete the un-archived directory. If we extract the test_archive.zip in the parent directory which contains the test_directory we would be asked if we want to replace or rename the archive.

rm -r test_directory

4. Unzip the test_archive.zip file. Using unzip with no extra arguments we see a verbose output report of each step of extracting the archive.

unzip test_archive.zip

Extracting a ZIP Archive to a Specified Location

Often we will want to extract an archive into a different specified location. We can add the-d argument to the unzip command to achieve this.

1. Extract the test_archive.zip to a specified directory. We can specify a relative or absolute location. In this example we extract the contents to the Music directory using a relative path. We could also pass an absolute path which provides the full path to the target location.After extracting the archive, move into the specified directory and use ls to check the archive has been extracted.

unzip test_archive.zip -d ./Music

2. Change directory and list the contents to verify the files have been extracted correctly.

cd Musicls

With these few basic uses of the zip command you now have lots of options when creating ZIP archives in the linux terminal. Being able to select specific file types and being able to set the compression level in the terminal emulator gives quick access to these powerful tools that are often hard to find in a GUI application.

We are going to look at a number of different ways to get RAM and CPU data in the terminal, before we finally look at two applications which can provide a basic level of assurance, at a glance.

These commands will work on most Linux machines. Our test PC is running Kubuntu, the KDE flavor of Ubuntu, but we also certain that you can also run through this how-to on a Raspberry Pi or any other Linux device. All of the how-to is performed via the Terminal.

How to Check Your CPU in Linux

1. Open a terminal.

2. Use the cat command to display the data held in /proc/cpuinfo.

cat /proc/cpuinfo

This command will produce a lot of text, typically it will repeat the same information for the number of cores present in your CPU.

A more concise means to get most of this information is via lscpu, a command that lists the CPU details.

1. Open a terminal.

2. Use lscpu to display the CPU details. The command is quite verbose and we can easily see the number of CPU cores, minimum and maximum CPU speed and the CPU architecture.

lscpu

Using a little grep magic we can pull out just the information that we need.

1. Open a terminal.

2. Using a vertical pipe, send the output of the lscpu command to grep and search for “max”. This will give us the maximum possible CPU speed. Pipes are a way to send the output of one command as the input for another. Classic examples are piping the output of a command to a text file for later review.

lscpu | grep max

Another means to get CPU information is using dmidecode, a command that dumps the DMI (SMBIOS) contents into something we can understand.

1. Open a terminal.

2. Type in the dmidecode command using sudo, and the argument -t 4. There are many DMI types, with a numerical reference used to pull information for that component. In this case -t 4 is for the CPU. We can replace the 4 with processor for the same effect.

sudo dmidecode -t 4

How to Check Your RAM in Linux

1. Open a terminal window.

2. Use the free command. This will show the available memory, and how the memory has been allocated, in Kilobytes.

free

3. Use the -m switch to show the available memory in Megabytes, or -g for Gigabytes.

free -m

Alternatively use the -h switch to show the memory information scaled to the three shortest digits. Giving us at a glance information.

free -h

There is a similar way to /proc/cpuinfo for memory information.

1. Open a terminal window.

2. Use the cat command to print the contents of /proc/meminfo. The output is incredibly verbose and can prove useful for debug. In general use we would stick with free -m.

cat /proc/meminfo

If we need to know the timings and breakdown of memory modules in a machine then we can use lshw.

1. Open a terminal window.

2. Using sudo, issue the lshw command with the -short switch (the device tree) and -C with the class memory.

sudo lshw -short -C memory

We can also use dmidecode for a more detailed query.

sudo dmidecode -t memory

Information at a Glance

Should we need a more general level of information, our current CPU speed, RAM usage, network bandwidth then we have two commands which can be installed.

How to Install and Use Htop

The standard top command is great, but htop is a much better alternative. Htop provides us with an interactive list of running processes.

To install htop.

1. Open a terminal window.

2. Update your list of repositories.

sudo apt update

3. Install htop using apt.

sudo apt install htop

4. Run htop.

htop

At the top of the window we can see the utilization of our CPU cores, under that is our RAM, and finally is the swap.

We can scroll through the list with our cursor keys or search using F3 followed by the name of a process or application. Here we have searched for the GNU Image Manipulation Program (GIMP). We can kill the app by pressing F9.

To close htop, press F10.

How to Install and Use Bpytop

Our personal preference is bpytop, a Python implementation of htop but with much more to offer.

To install bpytop.

1. Open a terminal window.

2. Install bpytop using the Python package manager pip.

sudo pip3 install bpytop

3. Run bpytop from the terminal.

bpytop

Bpytop’s interface is split into a series of areas.

1. CPU Status
CPU Speed
Core utilization
Overall utilization

2. RAM Usage
Total RAM
Used RAM
Available RAM
Cache

3. Disk / Storage Usage
Used and free space on all mounted devices

4. Network Interface Usage
Up and down speeds

5. Processes (just like htop)
Process management

We can show or hide an area by pressing the corresponding number (1..4). If we just wanted CPU, RAM, storage and Network information then we press 4. To bring it back, press 4 again.

To filter the processes for a particular application.

1. Press f to filter. It has to be lower case f.

2. Type in the application / process name / Pid. The search will adapt to show the process, in our case GIMP.

3. Press T to terminate the process, or K to kill. These have to be upper case T and K.

To exit bpytop press q at any time.

Picking the Best SSD for You

The era of PCIe 5.0 SSDs is also upon us, propelling storage performance to new heights. Blazing-fast PCIe 5.0 M.2 SSDs, which offer up to twice the sequential speeds of the older PCIe 4.0 standard, are now supported with Intel and AMD's current platforms, like Zen 5 Ryzen 9000 and the Intel Core 200S series.

It's great if your desktop system can handle a PCIe 5.0 drive, but they are still more expensive and certainly aren't a requirement. For example, the PCIe 4.0 Samsung 990 Pro is our current choice for the best SSD overall, and the best SSD for gaming. This drive is rated for 7,450 / 6,900 MBps of sequential read/write throughput and 1.2 / 1.55 million read/write IOPS. That means less time waiting for game levels to load or videos to transcode, not to mention a snappier experience in Windows.

PCIe 5.0 SSDs still have plenty to offer. The Crucial T705 ranks as the fastest consumer SSD in the world that you can actually buy, alongside similar SSDs like the Sabrent Rocket 5, delivering up to a blistering 14.5 GB/s of sequential throughput and 1.8 million random IOPS over the PCIe 5.0 interface. That's an amazing level of performance from a surprisingly compact device.

While the PCIe 5.0 drives are the fastest SSDs money can buy right now, believe it or not, raw speed isn't everything. In regular desktop tasks such as web browsing or light desktop work, you may not even notice the difference between a PCIe 3.0 SSD and one with a 4.0 interface, let alone a new bleeding-edge PCIe 5.0 model. The latest PCIe 5.0 SSDs also carry a heavy price premium for now, so you're probably best suited with a PCIe 4.0 model — unless you're after the fastest possible performance money can buy, of course. If that's the case and your system supports it, go for a new PCIe 5.0 SSD.

Ultimately, the best SSD for you is one that provides enough capacity to hold your data at a price you can afford. Consider that a high-end, AAA game can use more than 100GB of data, and Windows 11 all by itself may need 60GB. These days, we feel 2TB drives represent the sweet spot, with 4TB models becoming increasingly common.

Best SSDs in 2026 at a glance (more info below):

Here's the shortlist of our rankings, but we have deeper breakdowns for these drives below, along with far more picks for other categories, like PS5 SSDs, RGB SSDs, workstation SSDs, and SATA SSDs, among other categories.

Quick Shopping Tips

Below, you'll find our list of the best SSDs. For even more information, check out our SSD Buyer's Guide. Iif you're looking for an external SSD, you can check out our Best External Hard Drives and SSD page, or learn how to save some money by building your own external SSD.

Best SSDs You Can Buy Today

Samsung hit back at its competitors with this impressive update to the 980 Pro. New hardware and new options, including a heatsink with RGB and a 4TB variant, have allowed Samsung to retake the M.2 SSD crown. Performance is excellent across the board, setting a few new performance records, such as with 4K random read performance. In our testing, the drive was consistent, power-efficient, and cool. Samsung has also updated its software for this drive, giving it the best SSD toolbox available, and the drive is backed by a competent warranty and decent support.

$20 extra for a heatsink and RGB is a good deal, and Samsung will likely discount this drive over time. Competing PCIe 5.0 drives on the market offer faster performance, but they still carry a premium.

Read: Samsung 990 Pro review

WD has taken its popular Black SN850 SSD and turned it up to 11. The Black SN850X leverages an improved controller and newer flash to get the most out of the PCIe 4.0 interface. Performance is improved across the board, and the drive rivals most of the top contenders in the PCIe 4.0 market. There's also a heatsink option that comes with RGB at 1TB and 2TB. WD also supports the SSD with its decent Dashboard application and a respectable five-year warranty.

The M.2 Black SN850X was a bit pricey at launch, however, with a daunting MSRP, but those prices have largely come down. The touted Game Mode 2.0 feature felt incomplete in our testing, although WD ensures us that this will improve with future firmware updates. All-in-all, this is a good compromise if you can’t find the Samsung 990 Pro.

Read: WD Black SN850X review

Fastest Best SSDs

The Corsair MP700 Pro XT delivers shocking levels of performance with excellent power efficiency. It’s the first Phison E28-based drive we reviewed, and it’s also the fastest drive, period, we’ve ever reviewed. If you want a no-compromises storage solution, this is it. The only exception would be if you’re gunning for 8TB in a single drive, in which case we recommend the Samsung 9100 Pro or, when it arrives, the 8TB model of the SanDisk WD Black SN8100. The Black SN8100 was our old pick for fastest SSD and still remains relevant at lower capacities if it’s priced lower than the MP700 Pro XT.

Read: Corsair MP700 Pro XT review

Not content with the release of the WD Black SN7100, our pick for the best laptop M.2 SSD, Sandisk followed up quickly with the WD Black SN8100. This drive is a response to the Samsung 9100 Pro – a good drive in its own right – with the added desire to destroy all existing high-end PCIe 5.0 SSDs. The SN8100 achieves this by using a new, more efficient controller from SMI along with very efficient BiCS8 flash. Through this, the drive is not only highly power-efficient for an SSD of this caliber, but it’s also able to hit higher performance levels with less overhead. For the time being, it’s the fastest all-around drive out there.

Sandisk could have skimped on the capacity end, but there’s 1TB to 4TB at launch, with 8TB promised in 2025. Historicall,y it has been difficult to get even 4TB drives out the door at this performance level and while it might take some time for Sandisk to improve availability there, the upfront promise of 8TB has us more hopeful. An 8TB drive in this class is a dream come true for many enthusiasts. That said, you’ll pay for the privilege. The SN8100 is not and will not be an inexpensive drive..

Read: Sandisk WD Black SN8100 review

Best Budget PCIe 5.0 High-End SSD

“Budget” and “PCIe 5.0” are two terms that are difficult to reconcile, but the modern consumer SSD landscape is such that you actually might want to save some money when getting a high-end drive. The Lexar NM1090 Pro is positioned precisely to take advantage of that, offering full PCIe 5.0 bandwidth but at a lower cost by using older flash. The cost differential is small, but so is the real-world performance impact. This drive will max out your PCIe 5.0 slot and won’t leave much performance on the table, which means the drive can be a slightly more affordable way to future-proof.

You’re still getting 14 GB/s or more of theoretical bandwidth with up to over 2 million 4K random read IOPS. These are startling numbers that beat any PCIe 4.0 drive and also all lower-end Gen 5 drives. Earlier high-end Gen 5 drives will fall behind in other areas, such as being much less power-efficient than the NM1090 Pro. This is a drive that could functionally work in some laptops – especially in Gen 4 mode – and it never needs any active cooling. You’re trading one generation of NAND flash for the ability to get a top-tier drive for a little less money. A worthwhile trade-off, even if the drive doesn’t stand out in any way.

Read: Lexar NM1090 Pro review

Best Laptop SSDs

If you’re looking for a new SSD for your laptop or other portable device, the Sandisk WD Black SN7100 should be at the top of your list. Battery life is usually of utmost importance, and there is no drive on the market more efficient than this one. In the grand scheme of things, your SSD probably won’t make a huge dent in your power budget unless you’re pushing things a little harder, but on the other hand, every bit matters. That selling point alone might not be enough for everyone, but luckily, the Black SN7100 also has fantastic random read performance, which makes it exceptionally responsive.

While the drive did not have a 4TB model at the time of launch and review, it is now part of the lineup, which removes one of the negatives we had for the drive. You no longer have to compromise on capacity. However, the drive still suffers from middling all-around performance, and its pricing remains a sticking point. Sandisk has been more prone to sales as of late, though, and we think that even if this drive isn’t the best pick for desktop, it’s pretty hard to beat for laptops.

Read: Sandisk WD Black SN7100 Review

The Crucial T500 combines cutting-edge flash with a customized controller that manages to be power-efficient with just four channels but also squeezes in the coveted performance-boosting DRAM cache. The T500 is also a single-sided drive with TCG Opal support, making it perfect for professional laptop use.

Many laptops are still stuck with PCIe 3.0 slots, and that’s fine. The T500 will be even more efficient when run at 3.0, and its benefits, aside from bandwidth potential, do not disappear. While the T500 does offer a heatsinked version, which we have in our all-around best SSD category, you’ll be going bare for a laptop. In this respect, it can even be better than DRAM-less drives, as the T500’s controller has more surface area and a metal IHS to prevent controller overheating. It’s simply the finest drive for laptops at this time unless you really want more horsepower. That’s on the menu, too, especially once the 4TB version arrives.

Read: Crucial T500 Review

Best 4TB SSD

4TB has become a more attractive capacity point for SSDs as time has gone on. While there are now many options available, most come with compromises of one sort or another. You may have to settle for QLC, a weaker controller, no DRAM, unreliable hardware, etc. This is not always a big deal, especially if the drive is intended to be a secondary gaming drive. In the PlayStation 5, however, extra cooling is beneficial, so it’s convenient to have a heatsink option available. At the same time, laptops favor bare drives and especially single-sided drives, the latter of which have been very rare with TLC until recently.

Samsung has managed all of this with its high-performing 990 Pro SSD. You have a powerful controller with DRAM, cutting-edge TLC flash, and a single-sided drive with or without heatsink even at 4TB. WD’s SN850X has been out a while at 4TB but has no heatsink option and is double-sided, with the SN850P being a latter heatsinked version for the PS5. There has been an increasing amount of 4TB TLC drives, including the Lexar NM790 and Addlink A93, but these cannot compare to the power and brand power of Samsung’s 990 Pro. You do have to pay for that privilege given the high MSRP, but at this time there is no substitute.

Read: Samsung 990 Pro Review

Now that Crucial has finally brought out the 4TB SKU for the T500, it can replace the T700 on our best SSDs list for the best 4TB SSD alternative. The T700 is still a good choice for this, but the T500 is better for a few reasons. While both drives have a heatsink option, the T700 requires one, while the T500 can work bare in a laptop. The T500 is also more power-efficient but doesn’t skimp on performance by omitting DRAM. And while the T700 is PCIe 5.0 capable, many machines — including laptops and the PS5 — won’t benefit from that extra bandwidth.

The 4TB T500 is not without its faults, though. Its pricing is a little high for what you get, matching other high-end drives, which makes more sense on desktops. This is partly because the T500 has inconsistent sustained performance while those like the 990 Pro and SN850X do not. The 4TB T500 is also double-sided, which potentially reduces its compatibility. There are already single-sided, 4TB DRAM-less drives for less, such as the Lexar NM790, and there may be more in the future, although in general, this fact shouldn’t reduce the T500’s appeal.

Read: Crucial T500 4TB review

Best High-Capacity (8TB) SSD

We love the WD Black SN8100 and we love the SanDisk Optimus GX Pro 8100 now, too. Okay, it’s the same drive by a different name, but we finally got to see what this hardware can do at 8TB with the latter. While the WD Black SN850X remains the best bargain for an 8TB SSD, the Optimus GX Pro 8100 / Black SN8100 offers the best performance for an 8TB drive. It achieves all of this while remaining power-efficient, which is no mean feat. It even offers an optional heatsink with RGB LED lighting if that suits your fancy.

Most importantly and especially for a high-capacity drive, the drive’s random read latency is incredibly low. This ensures the most responsive experience possible with fast game and app loading times. The drive even has a Game Mode feature to improve this even further. It’s maybe a little expensive to get just for games, but if you want the very best this is the way to go. It’s the best all-around 8TB drive out there if you need one drive to rule them all. The Samsung 9100 Pro is not the worst consolation prize but it’s just not as good. We’ll have to wait for more high-end, Gen 5 8TB drives to see if SanDisk’s masterpiece can be unseated.

Read: SanDisk Optimus GX Pro 8100 Review

The Samsung 9100 Pro is not a record-setting drive now that the second generation of Gen 5 SSD controllers are coming out in retail products. However, it is the first high-end drive to offer an 8TB SKU. This is a massive amount of storage, especially for a drive at this performance level. The 9100 Pro delivers good results in all of our tests, and the overall package is perfect for enthusiasts who want this much space in one drive for their HEDTs and workstations. It can also work in laptops, but the double-sided nature of the drive makes it more suitable for desktops, especially with the optional and affordable heatsink. Behind it all is Samsung’s name, so you know you’re getting a high-quality product with support to back your expensive investment.

Read: Samsung 9100 Pro 8TB SSD review

Best Budget M.2 SSD

Everybody loves a good budget drive and, luckily, there’s almost always something to fill that gap. Right now, it’s the Biwin NV7400, taking over from our long-time favorite TeamGroup MP44. There’s nothing wrong with the MP44; it’s just getting harder to find. It also has variable hardware – a reality for almost all drives – while our more recent NV7400 sample had very good hardware, indeed. Biwin did a great job with this drive, and we can recommend it for pretty much any system, so if its price is at or below similar contenders like the MP44, then it’s an easy choice. Assuming things stay that way.

There are many things to like. It has good all-around performance, capped by good power efficiency. The single-sided design, that makes it a good pick for laptops and the PS5. You can always add your own heatsink, though. It also has high TBW if that’s something you look at, although we think a drive with DRAM is better for high writes. Also, there is no 8TB SKU here – the MP44 does have that – but we really don’t think that matters for a budget drive. There are less expensive drives, sure, but this category is for the best budget drive, and the NV7400’s bandwidth and 232-Layer flash take the prize.

Read: Biwin Black Opal NV7400 2TB SSD Review

We recently replaced the TeamGroup MP44L on our Best PS5 SSDs list for a good reason: the MP44L has had its hardware changed over its life, but it’s now showing up with decidedly bad controllers and QLC flash. It was our budget champion for quite some time, but it feels like Team has introduced the MP44Q to fill that gap.

Drives in the middle of the PCIe 4.0 range – usually around 5 GB/s peak, give or take – are becoming less sensible by the day. What this means for you is that it’s more appropriate to opt for budget-friendly 7 GB/s drives, such as the MP44 and MP44Q, if you’re budget-conscious. Which drive to pick depends on the pricing and availability at the time of purchase. The MP44Q can end up competing with the MP44 with comparable pricing at times, and even at its best, it’s up against other QLC-based drives.

That doesn’t mean the MP44L and its class of drives should be ignored completely, as sometimes it really does come down to a few $ in either direction. However, we would strongly recommend going up to the MP44Q at the least if you value hardware longevity. Cheap drives are cheap for a reason. The MP44L once stood out with good controllers and TLC flash, but now that the market squeeze has left it as a poor man’s Kingston NV3 – Kingston’s NV series being the poster child for hardware that changes in some budget SSDs – we think you can do better. The MP44Q offers a good, power-efficient alternative.

Read: TeamGroup MP44Q review

The WD Green SN3000 is not a drive for everyone. It’s on the lower end for budget Gen 4 drives in terms of peak bandwidth, and it inherits a mixed reputation. WD’s Green line has always been associated with dirt-cheap drives that are very much the last resort. This isn’t helped by the fact that the SN3000 has a shorter warranty than usual at three years, and also has some of the lowest TBW, or warrantied writes, of any drive we’ve tested in recent memory. That paints a nasty picture, but we believe this drive is actually a diamond in the rough.

The drive has good performance thanks to its BiCS8 QLC flash, so it feels more responsive than expected. This flash also helps the drive stay efficient enough for mobile devices. On top of this, you can get the drive at smaller capacities that have been hard to find. If you’re buying on a budget, it’s a veritable godsend. We can overlook its weak sustained write performance, as that’s not a factor in this segment. As for the TBW, we can’t deny that it is dismal, but in its defense, a read-heavy drive won’t have any problem here over three years.

Read: WD Green SN3000 Review

Sandisk’s newest Blue NVMe SSD is an achievement despite its complete transition from TLC to QLC flash. This drive has something for everyone, offering decent all-around performance, low 4K random read latency, and high power efficiency, all within a nice capacity range. In fact, it’s difficult to tell it apart from the TLC-based Black SN7100 in everyday use. Given that the Blue SN5100 is using less-expensive QLC flash, it should end up priced better, too, which would be quite something given the great value the Black SN7100 provides.

The Blue SN5100’s excellent balance makes it an all-around winner for use in mobile systems, the PS5, HTPCs, and as a secondary or gaming desktop drive. It’s a drive for everyone and is our pick for the best QLC drive on the market, although the Crucial P310 comes close. QLC usually means capacity, and thankfully, the Blue SN5100 is available up to 4T,B which leaves us with few criticisms. Yes, it still has weaker sustained write performance than a TLC drive, but this is less concerning than the launch price. This drive, when settled at its proper price point, will be a great pick.

Read: Sandisk WD Blue SN5100 Review

WD took its popular Black SN850 SSD and turned it up to 11, but luckily for value seekers, the price isn't nearly as extreme. The current $156 price on Amazon for the 2TB model is a great deal, even if it's now $25 more than it cost last year. The Black SN850X uses an improved controller and newer flash to get the most out of the PCIe 4.0 interface, thus delivering excellent performance with the Sony PlayStation 5. WD improved performance across the board, and the drive comes with a heatsink option at 1TB and 2TB capacity points.

WD also supports the SSD with a solid five-year warranty that will let you game with peace of mind. This drive is made for the PlayStation 5, and while it can be a bit pricier than budget options, overall, it's still our top pick for the PS5. It's also fast for gaming on a PC, particularly with DirectStorage starting to become useful, so this drive is plenty attractive.

WD has taken the course of releasing an officially licensed SN850P SSD. That drive is a glorified heatsinked SN850X and you should only pick it if you want the heatsink at 4TB. Even then, it's far cheaper to get a bare SN850X and add your own heatsink.

Read: WD Black SN850X Review

The Addlink A93 is a perfect example of a great PlayStation 5 expansion drive – it delivers top performance at up to 4TB without breaking the bank. The PS5 doesn’t need anything special, but you might as well get a drive that will last a long time. The A93 checks this box with its TLC flash, and it’s also designed for the PS5 with a compliant heatsink and a single-sided drive design. It’s DRAM-less with a four-channel controller, which means it’s power-efficient and, with its heatsink, it’s therefore unlikely to overheat, which is always good news for console lovers.

That said, enthusiasts would still prefer to have DRAM, which is why we have the WD Black SN850X as our first pick for a PS5 drive. If you’re trying to save some money, though, the A93 is a decent substitute. There are competing drives that would work as well, such as the Lexar NM790, but currently, the A93 offers the best package for the best price. It has all the performance you need with no real drawbacks, unless you really have your heart set on 8TB.

Read: Addlink A93 SSD Review

The Crucial P310 came as a bit of a surprise, but a welcome one. M.2 2230 SSDs have ratcheted up in popularity ever since Valve’s Steam Deck launched, and now there are more portable gaming systems than ever. There’s also Microsoft’s Surface Pro line and some laptops that take M.2 2230 or M.2 2242 - this drive can be extended up to M.2 2280 if needed - which used to mean going to eBay for OEM options like the WD SN740. This hasn’t been the case in a while, but finding a decent 2TB drive has remained difficult. The P310 handles that challenge like a champ.

Sure, it’s QLC-based, which means it’s not quite as fast or consistent as it could be, but it’s more power-efficient than the TLC-based WD Black SN770M and has more throughput. In fact, it’s the fastest 2TB M.2 2230 SSD we’ve ever tested. We expect the updated Corsair MP600 Mini would beat it, but the P310 has better availability and should be less expensive. It’s fast enough where it matters, which makes it the best option if you’re looking purely for capacity, but your host system should be able to take PCIe 4.0 drives to fully benefit.

Read: Crucial P310 SSD review

Kingston’s NV3 (2230) is a good example of a budget drive done right. Our reviewed version has good hardware, including superior QLC flash. However, the specifications are lower than normal which gives Kingston the ability to change the hardware down the road. This isn’t a big trade-off considering the drive is priced right at 1TB and 2TB, the most popular drive capacities for this form factor. The good power efficiency also means it won’t drain your battery or overheat.

When we’re talking about the Steam Deck this drive is more than fast enough as the device is limited to PCIe 3.0. Other and future devices will take PCIe 4.0 and even 5.0 SSDs, however. For the most part you don’t need a drive with TLC flash even in that case if your primary workload is gaming. The NV3 would have more competition there, certainly on the performance front but also possibly with capacity down the line. Aside from the excellent Crucial P310, we may begin to see 4TB drives in this and the M.2 2242 form factor with double-sided solutions, 2TB dies, or both. Right now the NV3 is a solid deal, though.

Read: Kingston NV3 (2230) SSD review

The Sandisk Optimus GX7100M is, in essence, a short-form-factor version of the WD Black SN7100. The hardware, at least, is the same – Sandisk’s well-regarded, proprietary controller technology combined with BiCS8 TLC flash. The controller is known for its reliability, and the flash, well, it’s power-efficient with excellent random read performance. This makes it great for a gaming drive where you want the fastest load times possible. This form factor is designed for portable gaming devices, so the drive needs to be efficient, too. Less battery drain and less heat is the name of the game.

There is only so much you can do about heat in this form factor, though, but it shouldn’t be an issue in the Steam Deck. On the other hand, you’re not getting the most out of this drive on a PCIe 3.0 host device. Considering the premium you pay for the drive – it’s particularly expensive at 2TB, which is the go-to capacity as an upgrade – you might want to go a different way with the Deck on a budget. On the other hand, Sandisk has strong software support, including Game Mode if you’re on or multi-booting to Windows, which provides a premium experience for non-Deck devices, especially.

Read: Sandisk Optimus GX 7100M Review

The Corsair MP700 Micro is the best M.2 2242 SSD on the market, and it’s not even close. It’s not only the fastest drive, it also has the highest capacity possible at up to 4TB in a single-sided design. This makes it an effective, if niche, storage solution. If you happen to need an M.2 2242 SSD with this much space or want the highest performance and power efficiency possible, the MP700 Micro is your drive. Although a bit overkill, it would also be great in PCIe 3.0 and 4.0 slots if that’s all you have available.

The overkill factor is also a drawback to this SSD. You can get by with slower drives like the Sabrent Rocket Nano 2242, especially as most M.2 2242 devices are not PCIe 5.0 capable. While it’s impossible to match the capacity of the 4TB MP700 Micro, it’s a very expensive drive, and the current memory market may end up encouraging smaller drives for affordability purposes. If all you need is 1TB or 2TB, you can get by with many M.2 2242 or 2240 drives instead. The MP700 Micro still has its place, though, and it could be a niche solution for specific HTPC builds as well.

Read: Corsair MP700 Micro review

With the growing popularity of M.2 2230 SSDs, it was only a matter of time before we saw retail 2242 options. The Sabrent Rocket Nano 2242 is one of these, alongside the Corsair MP600 Micro. Alternatives include OEM and last-gen drives, like Sabrent’s original Rocket 2242, but some are double-sided. Not so with the Rocket Nano 2242, which will fit in the Lenovo Legion Go and many laptops with at least one M.2 2242 slot. It’s an easy drop-in solution with good performance and power efficiency.

The drive is only currently available at 1TB. However, with dual NAND packages, we expect larger capacity options in the future. M.2 2230 SSDs can also be extended for M.2 2242, but the 2TB options currently on the market all have their own drawbacks, except perhaps for the imminent Corsair MP600 Mini (E27T). However, the Rocket Nano 2242 gives plenty of performance for portable devices as it stands and is an easy pickup for M.2 2242.

Read: Sabrent Rocket Nano 2242 SSD review

PNY had its heart set on producing a very fast RGB-capable SSD, and with the CS3150 XLR8, or CS3150, it succeeded. This PCIe 5.0 SSD also has a heatsink with dual fans to ensure it never overheats. PNY’s software allows control over the RGB and fans, with synchronization possible for the former if you have other PNY RGB products. The warranty is standard, but the drive does support hardware encryption via the TCG Opal 2.0 specification, which may be a selling point for some.

The CS3150 isn’t perfect, though. It’s expensive and can be difficult to find. It’s only available at 1TB and 2TB capacities, needing 2TB to hit its maximum performance numbers. There are also other drives equal or faster to it, although for many workloads this isn’t particularly relevant. If RGB isn’t your thing, this drive also comes without the RGB in both white and black variants. Regardless of the model you go for, the drive can operate without throttling, and its performance is good across the board.

Read: PNY CS3150 Review

Best Workstation SSDs

The Inland Performance Plus has been around for a while, so long in fact that it hasn’t been using its original hardware in a long time. It was updated when the Gaming Performance Plus came out with a newer type of flash. It also saw an expanded capacity range over time, up to 8TB. The performance specifications have not changed, though, and remain quite good for a Gen 4 drive. The warranty is also excellent at six years – more than the typical five – with up to a whopping 6,000TBW at 8TB.

This is a solid all-around drive that’s good for anything, but it’s especially good for NAS. Aside from the high write endurance and warranty period, the drive also has a balanced pSLC cache scheme that makes it more ideal for expected workloads. It doesn’t hurt that the drive is also actually available – well, at the time of writing, anyway – for prices that are not outside the competitive range. It’s Inland so there’s no real software support, but that’s our only ding.

Read: Inland Performance Plus Review

Truth be told, the WD Black SN850X is a great all-around drive that could fit almost any category. Previously, though, there were a plethora of DRAM-equipped drives that were great for workstations at a lower price. Right now, with greater scarcity, the SN850X is a solid choice because it scales to 8TB and has a heatsink option. In fact, we would specifically recommend it with a heatsink for workstation usage, but you may be able to provide your own at a lower premium than the specific SKUs. For that reason, we note that the bare drive is single-sided at 1TB and 2TB.

The SN850X has excellent all-around performance, but for some workstation workloads, you might want good sustained write performance, too. While the SN850X isn’t the best of its generation for that, it’s surprisingly good, and for most writes, it’s not going to have any problems thanks to nCache technology. It’s also ideal to have good software support on a high-end machine, and WD has that covered with an excellent toolbox as well as Acronis True Image for file handling. Our only gripe is that the drive remains expensive, but that’s probably acceptable for the higher-end space.

Read: WD Black SN850X Review

Best Workstation SSD Alternate

The Kingston KC3000 is a long-standing favorite, a popular Gen 4 SSD that has been around for a while. There are many drives out there with the same hardware, including the previous holders of our Best Workstation SSD and SSD Alternate slots. The primary reason for the change is that drives with the Phison E18 controller have been identified with issues that could impact performance or reliability. However, only a few manufacturers have actually addressed this with a firmware update. Kingston is the primary one, offering the update for both the KC3000 and the Fury Renegade.

This makes the KC3000 a safer pick for workstation use, given that it also has decent all-around performance and is readily available. It also has a higher-than-standard TBW endurance rating and the Kingston name, which means it can be serviced more widely around the world. The only downside is that there are competing products, including some that are less expensive. However, if the reported issues are something you take seriously, then that brings things into focus. The KC3000 – or its sibling, the Fury Renegade – checks the right boxes for workstation use if you want a reliable solution.

Read: Kingston KC3000 Review

Best SATA SSDs

You can get a SATA drive in the M.2 form factor, but most SATA drives are 2.5-inch models, which allows them to drop into the same bays that hold laptop hard drives. SATA drives are the cheapest.

If you’re still buying SATA SSDs, your options are vanishingly few if you’re looking for a quality drive. Crucial has stopped making its popular MX500 – our previous Best SATA SSD – and other manufacturers have swapped to DRAM-less hardware in some cases, the WD Blue SA510 being the most prominent example. While having DRAM isn’t make-or-break with newer M.2 NVMe SSDs, it’s more useful for maintaining good performance with SATA SSDs. TLC flash is also preferable to QLC if you want to maintain high write speeds and performance.

This leaves the Samsung 870 EVO as the likely best remaining SATA SSD available. You’re going to get quality TLC flash and DRAM. Samsung also offers a decent capacity range, which means you should be able to find the drive you need when you need it. The vast majority of SATA SSDs out there are DRAM-less nowadays, but they’re also cheaper and might be good enough for basic upgrades. The 870 EVO, as a result, is an expensive proposition, but if you want a drive to last, it’s the best option.

Read: Samsung 870 EVO Review

How We Test SSDs and Benchmark Charts

We use the same test system for all our SSD benchmarks. You can find the specifications in the boxout, and the short summary is that it's an Intel Alder Lake platform — chosen because it was the first platform to support PCIe 5.0 for expansion cards and M.2 slots. We have periodically looked at newer platforms, but Raptor Lake didn't change the results much if at all, and AMD's PCIe 5.0 platforms tend to be slightly slower than Intel's platforms.

We have a battery of benchmarks, each of which gets run multiple times. We use the best result from each test. Here are the charts of all currently tested SSDs (from the past three years, give or take). We froze Windows 11 at version 22H2 in order to keep the test results consistent — various security updates have had an impact on certain benchmarks over the years.

We've grouped the SSDs by capacity, beginning with the 4TB and larger drives, then the 2TB drives (which are easily the most popular and well-represented class in our testing), then the 1TB drives, and finally all the 2230 drives (in both Gen3 and Gen4 modes). We haven't tested any new 500GB-class or smaller SSDs in several years as that market is mostly dead for DIY upgrades these days.

4TB SSD and larger testing results

2TB SSD testing results

1TB SSD testing results

M.2 2230 SSD testing results

Finding Discounts on the Best SSDs in 2026

Whether you're shopping for one of the best SSDs or one that didn't quite make our list, you may find savings by checking out the latest Crucial promo codes, Newegg promo codes, Amazon promo codes, Corsair coupon codes, Samsung promo codes or Micro Center coupons.

MORE: Best External SSDs and Hard Drives

MORE: All SSD Content

Before the election, Trump criticized the CHIPS Act, labeling it highly flawed and proposing tariffs as an alternative. But Lutnick, his chosen appointee for the position who will inherit program management from Raimondo, has indicated that he plans to continue with its implementation.

In a recent meeting with Raimondo, Lutnick expressed his commitment to the CHIPS and Science initiative, according to remarks Raimondo shared during a staff gathering last week, published by Bloomberg. Neither the Commerce Department nor Trump's transition team, responded when asked for clarification.

The $52 billion CHIPS and Science Act, launched under President Joe Biden to revitalize U.S. semiconductor manufacturing, is now transitioning to Donald Trump's administration. With $39 billion in grants already allocated and over $450 billion in private investments spurred, the program does not seem like a failure. Indeed, its progress is evident: Intel, GlobalFoundries, TSMC, Texas Instruments, and Samsung Foundry are building massive fabs in the U.S. While both Intel and Samsung faced some setbacks, both are committed to invest in America.

But while the new Commerce Secretary reportedly plans to continue the CHIPS and Science program, other people in Trump's team are not very fond of the initiative.

Vivek Ramaswamy, who is going to co-lead Donald Trump's Department of Government Efficiency (DOGE) presidential advisory commission with Elon Musk next year, criticized the Biden administration's approach to distributing multi-billion grants to semiconductor manufacturers weeks before Biden's tenancy is over, reports Politico. He called the action 'inappropriate' ahead of the power transition and threatened DOGE review of the funding contracts as well as other initiatives of the current administration.

"Wasteful subsidies under the IRA and CHIPS Act are being rapidly pushed out before January 20," Ramaswamy wrote in an X post a couple of months ago. "DOGE will review every one of these 11th-hour gambits and recommend that Inspectors General scrutinize these last-minute contracts. Political appointees who go on to work for beneficiaries of this midnight spending spree should be exposed unsparingly."

In the weeks following the election, over $16 billion in funding contracts were awarded to five semiconductor companies, including BAE Systems, GlobalFoundries, Intel, Rocket Lab, and TSMC as chipmakers raced to ink their deals with the U.S. government and get their funding under the CHIPS Act. Before the election, the administration had awarded just one major contract, totaling $123 million, to Polar Semiconductor in Minnesota.

Commerce Secretary Gina Raimondo has defended the timeline, asserting it was part of the original plan under the CHIPS and Science Act. Moreover, she has denied that the timeline has been influenced by any potential policy reversals under the next administration.

The Chinese Ministry of Commerce alleges that subsidies under the CHIPS and Science Act, which allocated $52.7 billion for U.S. semiconductor production, research, and workforce development, give American companies an unfair edge. These subsidies reportedly allow U.S. chipmakers to export mature node chips to China at reduced prices, undermining Chinese competitors. 

While the aim of the CHIPS and Science Act was to bring back production of logic and memory chips on leading-edge nodes back to the U.S., the U.S. government awarded billions of dollars to companies that produce chips on specialty or mature process technologies. The list of such companies is not too long and includes contract chipmakers GlobalFoundries and SkyWater Technologies, as well as integrated device manufacturers Microchip, Polar, and Texas Instruments. While there are makers of GaN, SiC, InP semiconductors which received subsidies, the objections cited by GlobalTimes and Reuters only mention 'mature node chips.' 

To get subsidies from the U.S. government, companies must invest millions or billions from their own pocket, therefore  they don't really cut down their CapEx budgets (which could allow them to reduce costs and lower prices). So it's unclear whether the allegations of MOFCOM have merits. However, the U.S. government's subsidies are aimed to make American chipmakers bigger and more competitive in general. 

This is not something that China likes as the country has built and continues to build dozens of fabs that are focused on making chips on 28nm-class process technologies and less-advanced nodes. Most, if not all, of these fabs are constructed with significant subsidies from the Chinese government. Chips made on mature nodes tend to be very cheap, but they are essential for everyday products like home appliances and communication devices, so their addressable market is vast. 

Last year, Barclays predicted that China could expand its chip production capacity by 60% over the following three years, and the vast majority of this capacity will be dedicated to mature nodes. China is certainly interested in ensuring that those fabs have enough customers as it aims to flood the market with cheap chips designed to replace equivalents made in Europe and the U.S.

Corning makes various glass materials to build photomasks, such as high-purity fused silica, ultra-low expansion (ULE), and extreme ULE glass. A major advantage of ULE and Extreme ULE materials is their exceptionally low thermal expansion to ensure maximum consistency in harsh EUV environments (which are getting harsher as EUV tools gain more powerful light sources to increase performance) and exceptional uniformity to reduce photomask ‘waviness’ to minimize circuit variability (i.e., lower performance and power variability). 

Corning’s ULE glass is used to make photomasks for DUV and EUV lithography, whereas Extreme ULE material is projected for next-generation High-NA EUV lithography.

Ensuring that companies in America that use advanced DUV, Low-NA EUV, and High-NA EUV tools have access to materials used to make photomasks is crucial for the whole industry in general and companies like Intel, GlobalFoundries, TSMC, Texas Instruments, and Samsung Foundry.

Corning’s $32 million funding will support the expansion of its manufacturing facility in Canton, New York, to increase the production of specialized materials like high-purity fused silica (HPFS) and Extreme ULE glass. This project will add 130 manufacturing jobs and over 175 construction jobs.

“Corning is proud to be an industry leader of specialty glass and precision optical lenses for the semiconductor industry, and our Canton facility plays a vital role in the nation’s mission to strengthen the domestic semiconductor supply chain,” said Corning Incorporated Chairman and CEO Wendell Weeks. “This proposed investment and the continued commitment in Canton will enable us to establish the next-generation manufacturing capabilities needed to produce vital materials for lithography tools that will develop the world’s most complex microchips, strengthening our commitment to advancing semiconductor technology and job creation in the United States.”

Soldering is one of, if not the most important maker skill and using the best soldering iron is the smart way to produce your best work. The best GPUs, CPUs and even 3D printers have components soldered to printed circuit boards (PCBs). Whether you are a seasoned pro, or new to soldering, you need the right iron for the task at hand, be it robotics, data science projects or arcade cabinets. But what is the best soldering iron?

On my bench I have a full hot air rework and temperature controlled soldering station, but I will often reach for a USB type C powered, smart soldering iron for quick fixes — the cheap and practical Fnirsi HS-01 has been my go to soldering iron since 2023. I've been soldering projects for well over 20 years, 15 years of that in a professional capacity and I know that a good soldering station is essential for larger soldering projects. But smart soldering irons are ideal for smaller projects, or for when you are on the move. They are quick to heat up and often have precision tips that we can use in surface mount projects.

I’ve personally tested more than 15 different soldering irons, measuring both their ease-of-use and the time it takes for them to get to an acceptable working temperature. Below, I’ve listed the best soldering irons, along with a guide to help you choose the right kind of soldering iron for your needs. This list is always changing, we've just reviewed a new precision soldering station for through hole and SMD soldering.

Best Soldering Irons and Stations

The Best Soldering Irons 2026

Best Smart Soldering Iron

$26 isn’t a lot of money in the world of best soldering irons. It can buy you a cheap kit, loaded with accessories to sweeten the deal. Or it can buy you what is probably the most useful soldering iron a maker could have. We loved the Pinceil v2 so much we gave it an Innovation Award.

Pinecil V2, is a refinement of V1, and brings a fully temperature controlled soldering station into the palm of your hand. No really! Don’t let its size fool you. This is a capable iron that can be used for delicate soldering tasks, or for larger jobs such as speaker cables and other connections with a large thermal mass. Just change the tip (it is compatible with TS-100, TS-101 tips) and you can solder connections both small and large.

Pinecil V2 is quick to heat up, hitting 350 degrees Celsius in just 20 seconds when connected to a 20V USB-C power adapter. That's really fast for any soldering iron and great for those of us that need to do a job quickly.

The “smart” aspect of Pinecil v2 is the OS. Yes we have an OS on a soldering iron, and with it we can tweak the iron to our needs. Setting a custom temperature profile for specific tasks is just a few clicks away. Customizing the UI for left or right handed users, setting sleep times to keep the iron hot while waiting for the user, then reacting to movement and ramping the iron to your preferred working temperature.

The downside of Pinecil v2 is that it doesn’t come with a stand. The overall shape of the iron means that it does not roll around your bench. If you need a stand, they can be purchased for just a few more dollars.

Given the size of Pinecil v2 it is the ideal soldering iron for younger or inexperienced makers. It feels more like a pencil than a soldering iron and the quick heat up and cooldown times mean there is less chance of injury. That said, always supervise learners when using any new tool.

Best Soldering Station

Hakko are the Rolls-Royce of the best soldering irons. Its color-scheme may look a little Fisher-Price, but this soldering iron is a professional piece of kit. The Hakko FX-888D is a soldering station that offers excellent thermal performance, with a soldering iron that can reach 480 degrees Celsius. Tips can be easily sourced and changed, enabling precision or heavy duty soldering.

The soldering station control unit has only two buttons but from there we can change the temperature and create presets for quickly moving from one type of job to another. The soldering iron has a great feel, with a non-slip coating and a flexible silicone cord to reduce accidental melting.

We soldered up a Velleman kit (PIC experiment board K8048) and the Hakko worked rather well. The default 350 degrees Celsius felt a little too low for the rather thick PCB so we upped the temperature to 400 and the lead solder flowed perfectly.

This is a pro level soldering station and with that it commands a pro level price. If you are going to solder a lot, or for professional use then this should be on your wish list.

Best All Rounder

The TS101 continues the greatness of its progenitor, the TS100 but provides a greater choice of power supply options. The iron is easy-to-use, quick to heat up and precise in its output. It's also compatible with TS100 soldering tips, opening up a world of choice.

In our review we loved how the TS101 felt in the hand, the button placement is perfect and the tip compatibility with the plethora of Pinecil and TS100 tips means that it caters to every user. Younger or inexperienced makers will find the TS101 slightly larger than say Pinecil v2, but the button placement and balance make this an ideal alternative for makers learning to solder.

The $50 price tag is double that of Pinecil V2 and there isn’t much difference between them. Both soldering irons support the same power options and soldering iron tips. What does separate them is comfort. If we were soldering all-day long, then the TS101 would be our choice.

Best Soldering Iron for Beginners

A cheap soldering iron kit is how many of us start our soldering journey. It is how I started way back in the 1980s. Sometimes a cheap kit will put us off, other times it offers a low cost point of entry for a new skill. The Tabinger solder iron kit is low cost, under $20 and comes with plenty of extras. This is normally a warning on the quality, in this case the iron is rather good despite our first impression.

In the kit we get lead free solder, spare tips and a folding stand (which works but isn’t the nicest). Sure all of these aren’t the best quality, but if we are starting out they will do a good job until we move onwards.

The iron is light in the hand and features a dial to set your temperature. Tips can be easily changed, and the kit comes with a selection of precision and chunky tips. Changing a tip involves unscrewing the collar and sliding the cold tip off, reversing the process to secure your choice in place.

Tabinger’s soldering iron melts solder well; just increase the temperature on the dial to be a little over your ideal choice. We normally solder at 350 degrees Celsius, but the solder was a little tacky with this iron, so we went to 400 and all was well. We did notice that the tip became “scorched” rather quickly. In our experience cheaper soldering irons can burn out quite quickly, so bear that in mind. Replacements can be easily sourced from Amazon or Aliexpress.

The Tabinger 60W Soldering Iron is a decent iron to get you started. The low cost makes it ideal for dipping your toe into soldering.

Best Hot Air Rework Station

The best soldering station with a hot air rework is a dream purchase for most makers. You can buy brand names for hundreds of dollars, but the Yihua 995D+ offers brand name performance for a fraction of the price.

We’ll start with the soldering iron: a rather generic iron which uses the collar system to retain a tip over a heating element. In this case being generic is a good thing. Replacement tips can be easily sourced (even Haako tips will work). The iron feels good in the hand and the silicone cable never gets in the way thanks to its own weight keeping it fixed to the bench.

The hot air gun also feels good in the hand and has a similar cable. Controlling the iron and hot air is a little tricky at first, requiring the correct controls to be selected (on the left are the hot air controls, on the right the soldering iron) before using the central knob to set the temperature and air flow.

The tall, thin design keeps your bench space clear - the included soldering iron stand takes a little more space but that's ok. The included stand is solidly built, doesn't slip around the bench and has the choice of brass or sponge to clean the tip. The Yihua 995D+ is an excellent soldering station for beginners and pros alike. It gets the job done, looks good and keeps more cash in your pocket.

Best Soldering Gun

Weller is a known brand in the world of soldering irons. They produce good quality irons and this 140W soldering gun represents another quality product.

Soldering guns come in many forms, some feed solder to the tip, others, like this require us to manually feed the solder with a spare hand. The Weller 140W soldering gun is a beast, and lays down a surprising amount of heat in six seconds. From ambient to 480 degrees Celsius in six seconds, you can solder heavy, big joints. If you need to solder some speaker cables, large power connections on a robot or vehicle, then this gun will make the job so much easier.

Slightly pressing the trigger unleashes 140W of power to the tip of the iron. Talking of tips, it comes with a selection for precise and blunt tips for different soldering scenarios. This isn’t a general purpose soldering iron, but when you need raw power, for a large surface area, then this will do the job just fine.

Best Travelling Soldering Iron

Fnirsi’s HS-01 smart soldering iron is a challenger for Pinecil’s crown. It shares the same price-point, around $30 and it uses USB C power. But the HS-01 is a much better travel companion.

First of all is the cap. Instead of a stand we get a cap which firmly covers the hot tip of the soldering iron. Sure the cap will get hot if you put it on right after soldering, but the iron cools down quick, and heats up even quicker! The screw collet, designed to hold the tip in place, means that we do not need a screwdriver to replace tips, just wait for the iron to cool and you can swap the tip for a compatible replacement. TS100 / TS101 / Pinecil tip compatibility would be the ideal, but a set of Hs-01 compatible tips is around $15.

Soldering with the HS-01 is a joy. The included tip is a good all-rounder and for our review it performed great on solder joints big and small. For $30, if the Pinecil V2 is out of stock, or you need a soldering iron for your traveling maker kit, buy this instead!

Read: Fnirsi HS-01 Review

Best for the Professional Maker

The Ifixit Fixhub Portable Soldering Station is a great all-rounder. In your home workspace, the makerspace and the garage, the Ifixit Fixhub Portable Soldering Station has the power and precision for all of your soldering needs. It feels like a Hakko soldering station, but portable. The mix of portability, power and precision it is going to make a dent on your wallet. If you are a professional maker who needs a powerful soldering system around your shop, then the Ifixit Portable Soldering Station is for you.

The replaceable tips are great quality, and they can put the heat exactly where we need it. We used the conical, knife and bevel tips and each of them exceeded our expectations. The USB-C cable is great quality, and won’t get in the way when soldering. Take a look at our full review to learn everything about this great soldering station.

Read: Ifixit Fixhub Portable Soldering Station Review

Picking The Best Soldering Iron For You

Picking The Best Soldering Iron For You

Finding the best soldering iron for you is important. You need to be comfortable with the soldering iron in order to concentrate on the task at hand. But which type of soldering iron is for you?

• Basic Soldering irons: If you are just trying your hand with soldering then perhaps an advanced soldering setup is not for you, yet. A simple soldering iron heats up to a set temperature, giving you the confidence to start soldering without having to tweak the settings.
  
• Smart Soldering Irons: If space is at a premium, but you need a temperature controlled soldering station, then smart soldering irons are for you. Typically powered using USB C or DC power supplies, these irons provide precise temperature control without dominating the bench.
  
• Soldering Station: If you need a soldering iron with precise temperature control, along with the best thermal balance, then a soldering station is for you. Stations have an external control unit which contains the power circuitry necessary to deliver precise thermal output. The soldering iron is typically housed in a separate stand giving the user flexibility on the bench.
  
• Hot Air Soldering Station: These stations offer a temperature controlled soldering iron, along with a hot air gun that is used for surface mount soldering, reflowing circuits and desoldering components from boards.
  
• Soldering Gun: A soldering gun is a brute force tool, used for larger solder joints. If you are soldering heavy duty cables, speakers or terminals, then a soldering gun has the power you need. For finer, detailed work they are clumsy and awkward, so you will still probably want a soldering iron to go with your gun.

Other Soldering Irons We Tested

Other Soldering Irons We Tested

In our tests to find the best soldering iron we tested a number of other irons. Some from well known manufacturers, and others from some lesser known manufacturers. These didn’t make the grade in our tests but we have included the full results of the tests, so that you can make your own decision based on quality and price.

Fanttik T1 Max

This smart soldering iron is so simple to use, and it heats up and cools down fast. It is completely wireless and comes with a selection of soldering tips and kit to get you started. So why didn’t it make the cut? The handle is a little too thick, making it uncomfortable for smaller hands. The weight balance is a little off and despite the easy dial controls, the Fanttik T1 Max hasn’t got the grunt for larger solder joints. But this is a great soldering iron for gifting to a new maker. I’ve got a full review and analysis to help you make up your mind.

WEP 982 III Precision Soldering Station

I really liked this precision soldering station. It heats up fast, the 200C standby temperature keeps the iron ready for use. Tip selection is great, there are precision tips and big(ish) chisel tips for larger joints. You can “hot swap” the soldering tips using the stand. I also found this soldering iron to be very precise. So if it reviewed that well, why isn’t it on the list? The problems that I encountered were that it was easy to over shoot your soldering temperature, not a big deal, but it was irksome. But the biggest issue that I faced was that my review unit arrived damaged but the damage took a few hours to manifest itself. Hence I would only recommend buying this soldering iron from a retailer with a good returns policy.

Miniware TS80P

How did this iron not make the list? It looks great, feels great and it has a spring loaded tip ejection system. What let it down was the price, $95 is a lot for an iron. It can put down heat, but you need to change the included precision tip for something more general purpose in order to get it into a board. We love the iron’s user interface and it works really well, but so does Pinecil V2, which retails for a third of the price.

Weller WLSK3023G Soldering Station

Coming from the Weller pedigree we had high hopes for this soldering station. Alas it wasn’t meant to be. The LED light ring is annoying, the iron felt imprecise, and the hot tip of the iron pointing out from the stand felt counter-intuitive. It isn’t a bad iron, but compared to others, it just wasn’t worth the $70 asking price.

YIHUA 938BD+ Soldering Station

It was a tough call between the 938BD+ and the 995D+. The 938BD+ has a wider footprint, but the user interface felt a little less cumbersome. The call was tough, and if the 938BD+ is offered at a reduced price to the 995D+, then jump on it.

Preciva 8786D Hot Air Gun Rework Station

It may look a little dated, but this is a solid, low-cost soldering station. So why didn’t it make the list? The hot air rework gun felt weak, we had to ramp the temperature to the max in order to melt the solder on a heavy joint. The soldering iron felt fine, much like the Yihua stations (Preciva and Yihua are one and the same) but it lacked a silicone cord, instead using PVC. PVC cords are fine, but they are prone to creeping along your bench, getting in the way.

New Acalox Soldering Gun

This gun was the antithesis of the Weller. It took a while to heat up and required solder to be fed via a ratchet system, through the gun. It felt cheap, and the easily removable plastic cover exposed mains voltage connections. If you need a soldering gun, spend the extra money and get the Weller. This cheaper version is not safe for general use.

How We Test The Best Soldering Irons

How We Test The Best Soldering Irons

Each soldering iron was tested with a Velleman PCB kit, chosen for its mixture of small through-hole components and large soldering joints (mechanical joints to anchor components to the board). We used the same solder, lead based 60 / 40 with tin and rosin flux core across all tests, to ensure that our results were consistent.

From Cold to Hot: Getting to A Working Temperature

For each soldering iron / station / gun we tested how long it would take to get to a working temperature from cold. What is a working temperature? Well that is a personal choice. Some prefer to solder at 350 degrees Celsius, others much higher. Rather than set a temperature target, we chose to pick the moment where solder instantly melts on the tip of the iron. In the case of the hot air rework stations we chose the moment that a heavy solder joint would fully melt. All times are measured in seconds, and a lower time is better.

Basic Soldering Iron Times

These basic irons are there for “plug and play” soldering and the Tabiger 60W iron gets to a solder melting temperature much quicker (42 seconds) than the venerable Antex XS25 (116 seconds). The Tabiger has basic temperature control, and almost three times the power of the Antex iron. That said, Antex is a respected brand and offers superb performance, making it the ideal starter soldering iron. Its what I used for a very long time!

Hot Air Rework Station Times

These stations all feature a soldering iron and a hot air gun. The irons all heated up fast with the Yihua 995D taking first place in 24 seconds. The Yihua 938BD+ claimed second place at 33 seconds. For hot air, the 938BD+ was way slower than the others, 112 seconds versus a joint 92 seconds for the 995D and Preciva 8786D.

Smart Soldering Iron Times

Smart soldering irons are the F1 cars of the soldering world. They get hot, fast! The Fnirsi HS-01 took first place with a time of 9 seconds. Second was the Pinecil V2 which offers plenty of power for its low price! Any of these smart soldering irons would be a great addition to a maker’s toolbox. But for under $40, Pinecil v2 is hard to resist.

Soldering Gun Times

If you absolutely have to solder large solder joints, then a soldering gun is for you. These things get hot, and have the thermal mass to dump solder onto the target. Weller’s soldering gun is impressive. It heats up within six seconds, and can sustain large solder joints for a considerable amount of time. The Acalox soldering gun is slow to heat up (26 seconds) and it feels rather cheaply made. Avoid it and pay the extra for Weller.

Soldering Station Times

Weller may have won the soldering gun round, but here it loses out to Hakko’s FX888D. Heating up in 25 seconds, versus Weller’s 76 seconds, the Hakko FX888D is a sublime soldering experience, but we expect that given its price. If you are going to be soldering professionally it is always best to have the right tool for the job (as Star Trek’s Scotty can attest.)

Soldering irons frequently asked questions 2026