Idaho is mostly know for its delicious potatoes. Outside of agricultural circles, though, we tend to think of the state as Micron's cradle, and home to a world-wide behemoth in memory products. DRAM was where the company really got its start, and after a full acquisition of Elpida, it's basically the second-largest memory manufacturer, sandwiched between Samsung and Hynix. Of course, Micron happens to make non-volatile memory too. And these days, Micron and its consumer brand Crucial are all about the flash. When it formed IM Flash Technologies with Intel eight years ago, the sort of solid-state storage we have today was just a dream. Now, it's possible to get your hands on copious capacity thanks to advances in fabrication and SSD research.
Although Crucial got into the SSD game rather early, it only took off after launching the C300. That drive was announced almost four years ago, and was notable for being the first SATA 6Gb/s-capable SSD. It employed Marvell's then-flagship 9174 controller matched up to 32 nm Micron NAND. In 2011, the m4 became a worthy successor based on 64 Gb, 25 nm flash. Successive firmware updates turned the drive into an SSD that we were happy to recommend, especially as prices continued dropping last summer.
How do you follow something like that up? Tough question, right? Without a faster interface to exploit, any successor would still be constrained by a SATA 6Gb/s connection. Performance boosts could only be achieved in a few targeted areas, and pushing beyond the m4's 512 GB limit (economically, at least) required denser flash.

For the M500, the company needed to enhance, update, and improve on whatever it poured into that m4. It stuck to its guns and picked up Marvell's newest controller, the 88SS9187. Then, as 20 nm NAND started improving, transitioning to higher-density 128 Gb die helped achieve larger capacities using the same number of NAND devices.
But here's the awful truth: the main ingredient of solid-state storage is flash memory. If you possess the wherewithal to manufacture it yourself, you're in the best possible position to stay in the black. Once, the rising tide of SSD adoption lifted all boats. Now, the outlook is murkier, since there are so few NAND manufacturers. You have Samsung, Hynix, Flash Forward (Toshiba and SanDisk), and Micron (with Intel). Fortunately, the Crucial brand doesn't have to worry about its flash. With plenty of NAND experience, Micron sorted out its 20 nm flash and created a 960 GB drive with solid technical specifications and a price tag desktop enthusiasts would be able to afford.
That's great and all, but while the big M500 is generally pretty quick, it's not as speedy as the premium offerings from OCZ, Samsung, and SanDisk. The real reason you'd want a large M500 is its lower price and a pseudo-enterprise feature set.

Anyway, the M500 picks up where the m4 left off, and that was never really the fastest drive out there. It was more like the AK-47 of SSDs: practically indestructible and used the world over. Was the m4 perfect? No. But it was a yardstick of sorts, like Intel's X25-M back in the day. It just didn't make sense to keep the m4 around any longer since 25 nm Micron flash is going out of style quickly. As the entire industry shifts to smaller lithography, it's not practical or feasible to keep selling drives based on old stuff.
When Intel and Micron's 25 nm process came online, their flash was more expensive than the 3x nm NAND that preceded it. As yields improved, memory manufactured at 25 nm became much cheaper, fueling the dramatic price drops we've seen on solid-state storage. Shoot, a 256 GB m4 sold for $600 in early 2011. Two years later, they were going for 30% of their original price. Now, for the same $600, you can get an M500 with no less than 1024 GB of flash on-board. Not bad at all. Of course, joining the 1 TB model in Crucial's line-up are three other 2.5" SSDs:
| Crucial M500 2.5" SATA 3.1 | 120 GB | 240 GB | 480 GB | 960 GB | |
|---|---|---|---|---|---|
| MSRP | ~$120 | ~$200 | ~$400 | ~$650 | |
| Controller | Marvell 88SS9187 BLD2 | ||||
| NAND | Micron 20 nm Synchronous, 128 Gb Die | ||||
| Form Factor/Interface | 2.5", 7 mm Z-height, SATA 6Gb/s | ||||
| Warranty | Three Years | ||||
| Seq. Read/Write (MB/s) | 500/130 | 500/250 | 500/400 | ||
| Rand. 4 KB Read, QD 32 (IOPS) | 62,000 | 72,000 | 80,000 | 80,000 | |
| Rand. 4 KB Write, QD 32 (IOPS) | 35,000 | 60,000 | 80,000 | 80,000 | |
| Die Count | 8 | 16 | 32 | 64 | |
Not only is the M500 selling at new capacity points compared to the m4, but it also gets a bevy of new features. For instance, there's thermal protection that steps in to restrict write speeds when the drive gets too hot. This helps with reliability. TCG Opal 2.0 encryption is a boon for data security. In addition, a new power loss protection system removes some issues of data corruption during a sudden power loss event. Lastly, the M500 gets data redundancy in the form of RAIN (Redundant Array of Independent NAND) should one part of a die fail.
In addition to a 2.5" SATA-equipped form factor, the M500 is also available in mSATA trim as well. And unlike the m4, next-gen M.2-based drives are on their way as well, paving the way for next-gen notebooks and possibly desktop systems.
Compared to the pentalobe and Torx screws that many drives employ, the four Phillips-head screws holding the M500's chassis together are easily removed.

The bottom of the enclosure is substantial, which gives the drive some heft. The thick casing also helps transfer some heat away from the SSD's internals. You don't want the flash inside getting too warm; that can throw off expected voltage levels, wreaking havoc on endurance (this is why thermal throttling is an important safeguard nowadays). Should the M500 get too hot, the controller will slow write performance above 70 degrees Celsius.
Bottom
Top
With the PCB freed from its shell, the hardware is fair game for poking and prodding. Pictured above is the 480 GB drive, with eight packages per side. Each hosts two 128 Gb (16 GB) dies. The math there is pretty easy: every package adds 32 GB to this M500's capacity. Micron puts abbreviated codes on its NAND, but running it through the company's decoder gives us a traditional part number: MT29F256G08CECABH6.

Up top, we have Marvell's 88SS9187 BLK2 controller. As the successor to the '9174, which drove Crucial's m4, Plextor's M3s, Intel's SSD 510, and OCZ's Octane, the '9187 adds NAND redundancy features, queued TRIM commands, and even a little extra horsepower. It also supports up to 4 GB of DDR3 cache. Not that these drives need that much, but hey, maybe someday...
Crucial uses that controller with its custom firmware to implement RAIN (Micron's trade name for cross-die redundancy). Like some SandForce-based drives, this technology sets some NAND aside for parity. If one part of a die fails, the controller recovers and rebuilds the data stored in that location. It does sap some capacity (basically, one byte out of every 16), but the feature is a good trade-off in most situations. Some SandForce partners steered clear of the similar RAISE capability, opting instead to make that space accessible for a better price per gigabyte of capacity or use it to over-provision.
The M500's implementation of RAIN cannot recover from an entire die failing prematurely. Whole die failures do happen, unfortunately. But if just part craps out on you (say, one of the constituent planes that make up a die) the M500's RAIN feature should have your back.
This is the bank of itty-bitty capacitors that enables power loss protection (or PLP, if you please). This reliability-oriented value-add can be implemented a couple of different ways. The end goal is to protect information in the volatile write cache should your power go out suddenly. Once those bits hit the flash, they're safe. But they're vulnerable in the DDR3 DRAM and controller. PLP is purposed with safeguarding the entire data path.
Supercapacitors are basically big honkin' batteries, like a mini-UPS on the PCB. They've fallen out of favor over the past few years though, and many drives rely on a series of tantalum caps to keep data flowing when power to the SSD is interrupted. Crucial goes with a slightly different approach on the M500, though. Armed with a series of small caps, the controller flushes most of what's in the cache, while using "NAND tricks" to make up any deficit. At any given time, there's only 1 to 4 MB of user data in the DRAM, so it's not like the hardware is trying to write the Great American Novel in a handful of microseconds.

Our consumer storage test bench is based on Intel's Z77 Platform Controller Hub paired with an Intel Core i5-2400 CPU. Intel's 6- and 7-series chipsets are virtually identical from a storage perspective. We're standardizing on older RST 10.6.1002 drivers for the foreseeable future.

Changes in RST's driver packages occasionally result in subtle performance changes. They can also lead to some truly profound variance in scores and results as well, depending on the driver revision. Some versions flush writes more or less frequently. Others work better in RAID situations. In fact, builds 11.2 and newer support TRIM in RAID as well. Regardless, results obtained with one revision may or may not be comparable to results obtained with another, so sticking with one version across all testing is mandatory.
| Test Hardware | |
|---|---|
| Processor | Intel Core i5-2400 (Sandy Bridge), 32 nm, 3.1 GHz, LGA 1155, 6 MB Shared L3, Turbo Boost Enabled |
| Motherboard | Gigabyte G1.Sniper M3 |
| Memory | G.Skill Ripjaws 8 GB (2 x 4 GB) DDR3-1866 @ DDR3-1333, 1.5 V |
| System Drive | Kingston HyperX 3K 240 GB, Firmware 5.02 |
| Drives Under Test | Crucial M500 120 GB SATA 6Gb/s, Firmware: MU02 |
| Crucial M500 240 GB SATA 6Gb/s, Firmware: MU02 | |
| Crucial M500 480 GB SATA 6Gb/s, Firmware: MU02 | |
| Crucial M500 960 GB SATA 6Gb/s, Firmware: MU02 | |
| Comparison Drives | Samsung 840 EVO 120 GB SATA 6Gb/s, Firmware: EXT0AB0Q |
| Samsung 840 EVO 240 GB SATA 6Gb/s, Firmware: EXT0AB0Q | |
| Samsung 840 EVO 480 GB SATA 6Gb/s, Firmware: EXT0AB0Q | |
| Samsung 840 EVO 1 TB SATA 6Gb/s, Firmware: EXT0AB0Q | |
| SanDisk Ultra Plus 64 GB SATA 6Gb/s, Firmware: X211200 | |
| SanDisk Ultra Plus 128 GB SATA 6Gb/s, Firmware X211200 | |
| SanDisk Ultra Plus 256 GB SATA 6Gb/s, Firmware X211200 | |
| Samsung 840 Pro 256 GB SATA 6Gb/s, Firmware DXM04B0Q | |
| Samsung 840 Pro 128 GB SATA 6Gb/s, Firmware DXM04B0Q | |
| SanDisk Extreme II 120 GB, Firmware: R1311 | |
| SanDisk Extreme II 240 GB, Firmware: R1311 | |
| SanDisk Extreme II 480 GB, Firmware: R1311 | |
| Seagate 600 SSD 240 GB SATA 6Gb/s, Firmware: B660 | |
| Intel SSD 525 30 GB mSATA 6Gb/s, Firmware LLKi | |
| Intel SSD 525 60 GB mSATA 6Gb/s, Firmware LLKi | |
| Intel SSD 525 120 GB mSATA 6Gb/s, Firmware LLKi | |
| Intel SSD 525 180 GB mSATA 6Gb/s, Firmware LLKi | |
| Intel SSD 525 240 GB mSATA 6Gb/s, Firmware LLKi | |
| Intel SSD 335 240 GB SATA 6Gb/s, Firmware: 335s | |
| Intel SSD 510 250 GB SATA 6Gb/s, Firmware: PWG2 | |
| OCZ Vertex 3.20 240 GB SATA 6Gb/s, Firmware: 2.25 | |
| OCZ Vector 256 GB SATA 6Gb/s, Firmware: 2.0 | |
| Samsung 830 512 GB SATA 6Gb/s, Firmware: CXMO3B1Q | |
| Crucial m4 256 GB SATA 6Gb/s Firmware: 000F | |
| Plextor M5 Pro 256 GB SATA 6Gb/s Firmware: 1.02 | |
| Corsair Neutron GTX 240 GB SATA 6Gb/s, Firmware: M206 | |
| Graphics | MSI Cyclone GTX 460 1 GB |
| Power Supply | Seasonic X-650, 650 W 80 PLUS Gold |
| Chassis | Lian Li Pitstop |
| RAID | LSI 9266-8i PCIe x8, FastPath and CacheCade AFK |
| System Software and Drivers | |
| Operating System | Windows 7 x64 Ultimate |
| DirectX | DirectX 11 |
| Drivers | Graphics: Nvidia 314.07 RST: 10.6.1002 IMEI: 7.1.21.1124 |
| Benchmarks | |
|---|---|
| Tom's Hardware Storage Bench v1.0 | Trace-Based |
| Iometer 1.1.0 | # Workers = 1, 4 KB Random: LBA=16 GB, varying QDs, 128 KB Sequential, 8 GB LBA Precondition, Exponential QD Scaling |
| PCMark 7 | Secondary Storage Suite |
| PCM Vantage | Storage Suite |
Fantastic sequential read and write performance is a trademark of modern SSDs. To measure it, we use incompressible data over a 16 GB LBA space, then test at queue depths from one to 16. We're reporting these numbers in binary (where 1 KB equals 1024) instead of decimal numbers (where 1 KB is 1000 bytes). When necessary, we're also limiting the scale of the chart to enhance readability.
128 KB Sequential Read

Even with the Y-axis starting at 300 MB/s, it's pretty obvious that these four M500s are pretty evenly matched in our sequential read performance test. The ramifications of employing 20 nm, 128 Gb dies aren't felt quite yet.
On the other hand, all four knock around at the bottom of our comparison chart when we test with a queue depth of one.

The first thing you might notice if you take a long, hard look at the graph above is the lack of the 840 EVO. Don't worry, you'll be seeing that match-up a bit later on.
So, at lower queue depths, the M500s trail the field in sequential reads. Once you get four commands queued up, though, the '9187-equipped Crucial drives pick up the pace and join the rest of the pack. Intel's SSD 335 240 GB, also based on 20 nm flash, drops under 400 MB/s at a queue depth of one as well. The rest of our comparison hardware groups up tightly between 450 and 500 MB/s.
Here's a break-down of the maximum observed 128 KB sequential read performance during Iometer workload testing.
Although it's true that Crucial's new drives bring up the rear, there's not a ton of difference between the first- and last-place finishers. Frankly, maximum write speed is a more telling metric if we're going to lay out the results like this, bottlenecked by SATA 6Gb/s.
128 KB Sequential Write

Unlike sequential reads, sequential writes tend not to scale according to queue depth. Writes are actually more dependent on the number of dies a given task can be spread across. For example, higher density means that the 120 GB M500 sports just eight dies, whereas the 128 GB m4 had 16. That explains why the two lower-capacity M500s trail the previous generation at their respective capacities.
We have a hard time making a big issue out of this, particularly when the larger models are so much quicker than their predecessors. The quickest m4 topped out near the 265 MB/s mark, while the larger M500s exceed 400 MB/s. The difference isn't earth-shattering, but it's difficult to imagine a desktop application where higher sequential writes might change the experience drastically.

We add in data from elsewhere in the consumer SSD space, and Crucial's larger M500s are up there in the action. It'd be an exaggeration to say that the Extreme II, 840 Pro, and OCZ Vector pistol-whip the 480 and 960 GB Crucial drives. However, a 100 MB/s lead in favor of OCZ's Vector does seem a little brutal. Now's probably a good time to mention that there are SSDs framed as mainstream and others marketed as performance-oriented. The M500 falls into the former category, while the Vector, 840 Pro, and Extreme II are categorically high-end drives where speed is concerned.
If we chart out the maximum observed 128 KB sequential write performance from Iometer, it becomes apparent that the 120 GB M500 is not stellar when it comes to writes. Neither is Samsung's 120 GB 840. Achieving half (or less) of the performance posted by two-bit-per-cell-based SSDs sporting similar capacity, the Crucial and Samsung models are particularly hobbled. One is hamstrung by TLC NAND, while the M500 is hurt by a move to higher-density 128 Gb flash.
The M500 and 840 butt heads again at 240 GB, where Crucial's margin of victory is just 8 MB/s. Samsung's 840 EVO is another matter entirely. Once its Turbo Write buffer runs out, though, it's back down to the regular 840's performance level.
We turn to Iometer as our synthetic metric of choice for testing 4 KB random performance. Technically, "random" translates to a consecutive access that occurs more than one sector away. On a mechanical hard disk, this can lead to significant latencies that hammer performance. Spinning media simply handles sequential accesses much better than random ones, since the heads don't have to be physically repositioned. With SSDs, the random/sequential access distinction is much less relevant. Data are put wherever the controller wants it, so the idea that the operating system sees one piece of information next to another is mostly just an illusion.
4 KB Random Reads

All four M500s turn in similar performance up to a queue depth of four, but begin to diverge with eight commands outstanding. At a queue depth of 16, the differences are even more pronounced, resulting in a nearly 20,000 IOPS delta between the largest and smallest models. Double the queue depth from there, and the 240 GB version pulls up alongside the 960 and 480 GB configurations.

Even the 120 GB M500 comes off smelling like a rose compared to Intel's SSD 335. Overall, though, the m4's successor just can't catch the more elite SSDs at higher queue depths. The situation is far more favorable at lower queue depths, which is good news for Crucial, since that's where a majority of desktop workloads are reflected.
We can isolate maximum 4 KB read performance for a little extra context. There is some separation, but it's not very significant. As with the sequential reads, we're not particularly surprised to see such a small delta between most of the models on this chart. It doesn't take much flash to achieve spectacular read performance. The big differentiation happens in our write tests.

Random write performance is extremely important, no question about it. Early SSDs didn't do well in this metric at all, seizing up in even the most lightweight workloads. Newer SSDs wield more than 100x the performance of drives from 2007, but there's a point of diminishing returns in client environments. When you swap a hard drive out for solid-state storage, your experience improves. Load times, boot times, and system responsiveness all get better. If called upon, your SSD-equipped system could handle a lot more I/O than the spinning media you had in there before. With consumer workloads, it's more about getting to those operations faster, not necessarily handling more of them.
4 KB Random Write

Shining a spotlight on each M500 capacity point gives us a clear look at the impact of parallelism in solid-state storage. Write speed is very much a function of how many dies there are to exploit. We already saw this playing out in the sequential write test, and it's even more pronounced when we write small blocks of data randomly.
Crucial's 120 GB M500 passes 34,000 IOPS at a queue depth of two. The 240 GB drive doubles that number. Both larger M500s continue scaling, but match each others' performance as they plateau. When you're down at a queue depth of one, though, all four SSDs post the same ~30,000 IOPS finish.

Clearly, the 120 GB M500 can't hang with the big dogs, though it does rub up against the 256 GB Ultra Plus. That one model sports the only four-channel controller on our chart, and the impact is painfully obvious.
Crucial's two highest-capacity M500s give performances worthy of front-runner status. The 240 GB version is more modest. Athough it does a lot better than the 120 GB M500, transitioning to 128 Gb flash hurts the benchmark results in a measurable way.

Write Saturation
A saturation test consists of writing to a drive for a specific period of time with a defined workload. Technically, this is a fairly enterprise-class write saturation test, where the entire LBA space of the disk is utilized.

In Iometer and in this write saturation test, all four SSDs start out secure-erased. In the write saturation testing, however, where each drive is peppered across its entire LBA range with 4 KB random writes at a queue depth of 32, the larger M500s start by falling short of the big numbers we saw in Iometer (the were both up above 80,000 IOPS previously).
It's possible that both drives have to adjust to a larger LBA range. But if that's the case, then why don't the smaller models exhibit the same behavior? This could be an artifact of the larger drives' higher performance, whereas the 120 and 240 GB M500s are limited to 35,000 and 60,000 IOPS, respectively.
Given that the 480 and 960 GB models are so similar in terms of speed, it's easy to see that the larger SSD takes roughly twice as long to fill up. With the move from 120 to 240 to 480 GB, capacity jumps in proportion to write performance. That's why the 240 GB M500, despite its substantial advantage in IOPS, takes the same amount of time as the 120 GB drive to drop to a lower performance level on its way to steady-state.
Storage Bench v1.0 (Background Info)
Our Storage Bench incorporates all of the I/O from a trace recorded over two weeks. The process of replaying this sequence to capture performance gives us a bunch of numbers that aren't really intuitive at first glance. Most idle time gets expunged, leaving only the time that each benchmarked drive was actually busy working on host commands. So, by taking the ratio of that busy time and the the amount of data exchanged during the trace, we arrive at an average data rate (in MB/s) metric we can use to compare drives.
It's not quite a perfect system. The original trace captures the TRIM command in transit, but since the trace is played on a drive without a file system, TRIM wouldn't work even if it were sent during the trace replay (which, sadly, it isn't). Still, trace testing is a great way to capture periods of actual storage activity, a great companion to synthetic testing like Iometer.
Incompressible Data and Storage Bench v1.0
Also worth noting is the fact that our trace testing pushes incompressible data through the system's buffers to the drive getting benchmarked. So, when the trace replay plays back write activity, it's writing largely incompressible data. If we run our storage bench on a SandForce-based SSD, we can monitor the SMART attributes for a bit more insight.
| Mushkin Chronos Deluxe 120 GB SMART Attributes | RAW Value Increase |
|---|---|
| #242 Host Reads (in GB) | 84 GB |
| #241 Host Writes (in GB) | 142 GB |
| #233 Compressed NAND Writes (in GB) | 149 GB |
Host reads are greatly outstripped by host writes to be sure. That's all baked into the trace. But with SandForce's inline deduplication/compression, you'd expect that the amount of information written to flash would be less than the host writes (unless the data is mostly incompressible, of course). For every 1 GB the host asked to be written, Mushkin's drive is forced to write 1.05 GB.
If our trace replay was just writing easy-to-compress zeros out of the buffer, we'd see writes to NAND as a fraction of host writes. This puts the tested drives on a more equal footing, regardless of the controller's ability to compress data on the fly.
Average Data Rate
The Storage Bench trace generates more than 140 GB worth of writes during testing. Obviously, this tends to penalize drives smaller than 180 GB and reward those with more than 256 GB of capacity.

The M500s handle their business in this metric, but hardly turn in spectacular performances. The 120 GB M500 gets dinged right out of the gate because of its capacity, though not as severely as, say, the 30 GB Intel SSD 525. It's even slower than the three-bit-per-cell-based Samsung 840 120 GB.
All three larger M500s mix it up in the middle of the pack. For instance, the 240 GB drive reports back that it's 2.55 MB/s slower than the previous-gen m4. The 960 GB model is next-highest, just behind Plextor's 256 GB M5 Pro. We really want to know why the M500s fall in line the way they do, though.
Busy time accumulates when the SSD performs a task initiated by the host. So, when the operating system asks a drive to read or write, measured busy time increases. Take the total elapsed time and the amount of data read/written by the trace, and you get busy time in a far easy to understand MB/s figure. Unfortunately, busy time and the MB/s number generated with it aren't really good at measuring higher queue depth performance.
The corner case testing tells us that the M500s really stand apart from each other as queue depth increases. However, all of these SSDs are basically the same in the background I/O generated by Windows, your Web browser, or most other mainstream applications. It's only when we're presented with lots of reads and writes in a short time that the mid-range and high-end drives stand apart from each other. To test that, we need another metric.
Service Times
Beyond the average data rate reported on the previous page, there's even more information we can collect from Tom's Storage Bench. For instance, mean (average) service times show what responsiveness is like on an average I/O during the trace.
It would be difficult to graph the 10+ million I/Os that make up our test, so looking at the average time to service an I/O makes more sense. For a more nuanced idea of what's transpiring during the trace, we plot mean service times for reads against writes. That way, drives with better latency show up closer to the origin; lower numbers are better.
Write latency is simply the total time it takes an input or output operation to be issued by the host operating system, travel to the storage subsystem, commit to the storage device, and have the drive acknowledge the operation. Read latency is similar. The operating system asks the storage device for data stored in a certain location, the SSD reads that information, and then it's sent to the host. Modern computers are fast and SSDs are zippy, but there's still a significant amount of latency involved in a storage transaction.

In the chart above, drives that finish closer to the bottom offer better write latency. Moreover, the further you go to the left, the better the read latency is.
With that in mind, all four M500s are comparably quick when it comes to read access times in our trace. Write latency is another story, though. We've already established that write speed is heavily dependent on a drive's total number of dies and the controller's ability to utilize them in parallel. Up to a point, more flash devices give you lower write latency. This is evident in the massive spread between 120, 240, 480, and 960 GB M500s. Clearly, they're a lot further apart vertically than horizontally.
The 480 and 960 GB drives are fairly similar in all of our benchmarks. In fact, the 480 GB model gets a slight performance edge, suggesting that's the sweet spot right now with 128 Gb NAND and Marvell's 9187 controller.
All of the SSDs we're testing occupy a very narrow range of service times. Really, every product easily knocks our real-world read-based workload out of the part. If we're to get specific, Samsung's drives are the quickest, though we can't say if that's from their three-core controllers or faster flash memory. It's entirely possible that a slower interface is to blame for the M500's slightly lower results.

This trace has over twice as many read I/Os as writes, though writes account for more throughput. And here's where the testing gets dicey for Crucial's smaller M500s.
The 256 GB-class m4 and M500 behave almost identically. Samsung's 840 EVO drives enjoy a huge advantage due to Turbo Write, which emulates SLC. SanDisk's submissions do as well, since they also sport a technology that replicates the behavior of SLC, called nCache. Likewise, the OCZ Vector leverages a similar capability, though each specific implementation is different. OCZ's approach is the quickest of the three, taking first place in our trace for writes and second place for reads.
Crucial's M500 doesn't benefit from those fancy caching features, but rather holds its own through a more conventional design. The two larger models, specifically, perform admirably.
Futuremark's PCMark 7: Secondary Storage Suite
PCMark 7 uses the same trace-based technology as our Storage Bench v1.0 for its storage suite testing. It employs a geometric mean scoring system to generate a composite, so we end up with PCMarks instead of a megabytes per second. One-thousand points separate the top and bottom, but that encompasses a far larger difference than the score alone indicates.
This test is a big improvement over the older PCMark Vantage, at least for SSD benchmarking. The storage suite is comprised of several small traces. At the end, the geometric mean of those scores is scaled with a number representing the test system's speed. The scores generated are much different from PCMark Vantage, and many manufacturers are predisposed to dislike it for that reason. It's hard to figure out how PCMark 7 "works" because it uses a sliding scale to generate scores. Still, it represents one of the best canned benchmarks for storage, and if nothing else, it helps reinforce the idea that the differences in modern SSD performance don't necessarily amount to a better user experience in average consumer workloads.

The M500s place strangely. They don't light up the scoreboard, and I wasn't expecting that (particularly the two models placing way at the back). You shouldn't get hung up on one benchmark when it comes to characterizing a product's performance, but it's hard not to let the two smallest M500 SSD's scores taint our opinion. Shoot, the 120 GB is sandwiched between two 64 GB class SSDs, while the 240 GB is wedged between two 120 GB disks.
The 480 and 960 GB models register more respectable scores, though.
Futuremark's PCMark Vantage: Hard Drive Suite
PCMark's Vantage isn't the paragon of SSD testing, mainly because it's old and wasn't designed for the massive performance solid-state technology enables. Intended to exploit the new features in Windows Vista, Vantage was certainly at the forefront of consumer storage benching at the time. Vantage works by taking the geometric mean of composite storage scores and then scaling them a lot like PCMark 7 does. But in Vantage's case, this scaling is achieved by arbitrarily multiplying the geometric sub-score mean by 214.65. That scaling factor is supposed to represent an average test system of the day (a system that's now close to a decade behind the times). PCMark 7 improves on this by creating a unique system-dependent scaling factor and newer trace technology.
Why bother including this metric, then? A lot of folks prefer Vantage in spite of or because of the cartoonish scores and widespread adoption. That, and the fact that most every manufacturer uses the aged benchmark in box specs and reviewer-specific guidelines. In fairness, Vantage's Hard Drive suite wasn't designed with SSDs in mind, and is actually quite good as pointing out which 5400 RPM mechanical disk might be preferable.

In the more heavily 4 KB-weighted Vantage testing, the lower-capacity M500s redeem themselves. The 120 GB version even scores above the Intel SSD 335 and OCZ Vertex 3, both with 20 nm flash (though still using 64 Gb dies).
File Copy Performance with Microsoft Robocopy
Microsoft's Robocopy, a CLI directory replication command, gradually replaced the older xcopy. It's multi-threaded, has a ton of options, and generally outperforms vanilla Windows copy operations. Best of all, it's built right in to Redmond's operating system. Especially useful for network copy operations and backups, Robocopy doesn't stop to ask you one hundred questions while it copies over your music collection, either.
The reality of benchmarking file copy performance is that you need something fast to move data from and fast hardware to move it to. This is most important with SSDs. It doesn't matter if your drive can write sequentially at 500 MB/s if the source files are hosted on a USB 2.0-attached external hard drive. We're copying our test files from an Intel SSD DC S3700 to the drives in the chart below, taking source speed out of the equation.
There are 9065 files comprising the 16.2 GB payload. Some of the files are huge (up to 2 GB), while others are best described as tiny. On average, that's around 1.8 MB per file. The files are a mix of music, program, pictures, and random file types.
It's fair to say that this chart would look much different if we were copying from a hard drive to a SSD. Even if the disk drive's sequential throughput wasn't a bottleneck, it'd still choke on the smaller files.

When it comes to desktop SSDs, there are really two tiers of products, mostly invented by the vendors selling them. For instance, Samsung and SanDisk have bifurcated line-ups, split into enthusiast-class products like the 840 Pro and Extreme II, and value-oriented offerings like the 840 EVO and Ultra Plus. Bucking that trend, Crucial only sells one series aimed at the client space: its M500.
We're fine with that. There aren't many distinguishing features to set one group apart from the other. The differences that do exist usually show up as middling benchmark gaps. Crucial does plenty of heavy lifting, and its M500 falls between those often-contrived classifications. The more spacious models are just three and four seconds behind the first-place Corsair Neutron GTX in our 16.2 GB robocopy test, placing them ahead of midfield. Meanwhile, the 240 GB shows up just behind the 1 TB Samsung 840 EVO. Can you honestly claim to tell which of these drives are enthusiast-oriented, and which are designed to get your foot in the door of solid-state storage? It's not that easy. And the M500 gets extra points for features that aren't readily available on competing drives. The attraction is more than just benchmark-deep.
Unfortunately, the 120 GB M500 surfaces behind SanDisk's 64 GB Ultra Plus, which definitely is a value SSD. On the brighter side, it's one place ahead of the now-moribund vanilla 840.
Idle Power Consumption
Idle consumption is the most important power metric for consumer and client SSDs. After all, solid-state drives complete host commands quickly, and then drop back down to idle. Aside from the occasional background garbage collection, a modern SSD spends most of its life doing very little. Enterprise-oriented drives are more frequently used at full tilt, making their idle power numbers far less relevant. But this just isn't the case on the desktop, where the demands of client and consumer computing leave most SSDs sitting on their hands for long stretches of time.

Interestingly, the M500s demonstrate high idle power consumption. That's an important result for mobile users who may be inclined to swap drives and don't want to see their battery life suffer. Of course, with the implementation of device- and host-initiated power management, plus deeper sleep states in Intel's Haswell architecture, the significance of these numbers may be mitigated somewhat.
PCMark 7 Average Power Consumption
If we log power consumption through a workload, even a relatively heavy one, we see that average use is still pretty close to the idle numbers. Max power may spike fiercely, but the usage seen during a PCMark 7 run is pretty light. You can see the drives fall back down to the idle "floor", interspersed with peaks of varying intensity.

All four drives idle around the same wattage, so you mostly see them hanging out at a similar level. Then, when PCMark 7 ramps up its tests, we get spikes corresponding to heavier use, especially those triggered by writes to each drive. Maximum power consumption varies by capacity, so you don't even need the graphics legend to tell the four M500s apart.
The squadron of M500s follows the same pattern as most other drives, registering slightly more power consumption than the idle result during a complete run of the PCMark storage suite. Crucial's entire line-up does end up in the bottom half of this chart, though the results aren't bad considering how power scales up from idle. On average, the increase is only a few more milliwatts.
Maximum Observed Power Consumption
It's even better news that maximum power consumption isn't a critical specification for most desktop workloads. In the enterprise space, yes. This information goes into the calculation for total cost of ownership. But in a client environment, you shouldn't be seeing these numbers for more than short bursts.

It's safe to say that 1 TB of flash is a lot for one drive, and we don't find it unusual that the large 960 GB model uses the most power in Crucial's M500 line-up. Frankly, 5.3 W is a reasonable number for such a high-capacity SSD.
But Samsung's 1 TB 840 EVO places quite a bit more favorably, despite its use of three-bit-per-cell flash.
The M500s And 840 EVOs Mix It Up
Samsung recently sent us four different capacities of its 840 EVO. As it turns out, we also have four Crucial M500s, one of which came to us from our German team, one that was sampled by Crucial, and two we bought online. Naturally, we wanted to know how they match up against each other.
These drives aren't always going to fit the same applications, but they clearly have more than a few attributes in common, despite their dissimilarities. The M500s are built using 128 Gb, two-bit-per-cell MLC, feature RAIN redundancy, and power loss protection. The 840 EVO does its work with three-bit-per-cell flash, emulated SLC caching, and, well, that's it, mostly.
Samsung's new offerings don't offer the same TCG Opal 2.0 and eDrive security...at least not yet. With eDrive, cryptographic functions are hardware-accelerated, and Microsoft's built-into-Windows BitLocker can exploit it in Windows 8. That's going to get added to the 840 EVO in an upcoming firmware release. In contrast, the M500 is loaded with security features. I made the mistake of referring to the M500's security and encryption as just "fairly high-end," and was quickly interrupted by Micron representatives, who pointed out that, at the time, it had the highest-end suite of security features available. That's a fair point, and true as of this moment, though probably not for long.
So how about a few direct performance comparisons?
M500 Vs. 840 EVO : Sequential Read And Write

This is a chart housing four other charts, showing sequential read scaling versus sequential write scaling. The first graph pits the 120 GB M500 against Samsung's 120 GB 840 EVO, and so on. The M500s are shown in blue, and the 840 EVOs are purple.
Samsung enjoys an advantage over the M500s. At every capacity, read and write performance belongs to the 840 EVO. As far as reads go, specifically, the M500s start off a bit slower and don't ramp up to the 840's speed until we use a queue depth of four. This could be an artifact of the way Iometer's tests are performed, but there's a 200 MB/s gap between the drives at a queue depth of one.
The writes are more complicated. Samsung employs its Turbo Write emulated SLC caching scheme to improve performance. At first, we push too much data into the buffer to achieve optimal performance on the 840 EVOs, so the write numbers aren't what they should be. Without that cache, the EVOs are no faster than Samsung's original 840 (in other words, equal to or slower than the M500). But you won't see that here.
The situation gets even stickier when we test with random 4 KB blocks.
M500 Vs. 840 EVO : Random Read And Write

At a queue depth of one, both vendors nail high 4 KB results. The 840 EVO enjoys an edge as the three larger capacities break 10,000 IOPS. The M500s hold their own, though they lose out to the current random read champion at a queue depth of one. From there, though the gap narrows, the 840 EVOs maintain their lead, even at higher queue depths.
Random writes play out a bit differently. The 120 GB 840 EVO chokes early, dropping down into scandalously-low numbers at a queue depth of one. Otherwise, we essentially get a wash from the three larger-capacity SSDs in each family.
As the solid-state drive market matures, we're watching companies take different routes to end up in largely the same place. Some are on the interstate, others are winding around the scenic route, and an unfortunate few became hopelessly lost along the way. That's alright by us; we're interested in both the journey and the destination.
We remain huge proponents of SSDs. Few other technologies can affect the snappiness of your PC quite like solid-state storage. As they become more economical and technically refined, we continue refining the way we test them, compare them, and ultimately recommend them. We've seen this road littered with potholes in the past, but launch after launch, the top contenders impress us over and over again with interface-limited performance, fresh features, and attractive pricing...

...it's just that we sometimes get the impression one or two companies are flying private jets, laughing at the poor suckers stuck in rush-hour traffic.
Sometimes that's Crucial's position, and sometimes its not. Parent company Micron moves tons of drives, and being huge certainly helps. Did we mention that Micron is the number-two DRAM manufacturer next to (surprise, surprise) Samsung.
The M500 is the organization's sole consumer offering at the moment. While it was beset with early availability issues, getting your hands on any drive in the line-up shouldn't be a problem today. That's good news, since affordable 1 TB SSDs are a significant milestone. Previously, one-thousand gigabytes required a pair of 512 GB drives in RAID (not always an elegant solution, to be sure). Now, Crucial achieves this using higher-density dies, but continues leaning on two-bit-per-cell NAND. Samsung proved a while back that three-bit-per-cell flash is viable too, and it's not outside the realm of possibility that Micron might jump on that bandwagon in the future.

This takes us to the unavoidable comparison with Samsung's 840 EVO. We just reviewed that entire line-up in Samsung 840 EVO SSD: Tested At 120, 250, 500, And 1000 GB. Although the 840 EVOs employ triple-level cell memory, it gets a big boost from some quantity of emulated SLC, too. Consequently, the EVO is on par with, and sometimes ahead of the M500 in write performance benchmarks. Read performance is squarely in Samsung's camp, though.
The 840 EVO is missing power loss protection, cross-die redundancy, and it currently lacks Opal 2.0 encryption, which is one of the M500's aces. Samsung says its EVO will get that last feature in time, though the discussion is largely academic until the new 840 shows up for sale anyway. Both drive families sport three years of warranty coverage, but suddenly-lower pricing on the M500s might give them an edge.
What Crucial's latest offerings lack is the EVO's sweet set of consumer-friendly management and cloning tools. The company also isn't selling the M500s with installation kits, though a standalone upgrade package is available separately for $20 from Crucial's Web store. Samsung, on contrast, adds a USB 3.0-to-SATA adapter to its installation kit (though that luxury will add to the 840 EVO's price tag).
At the end of the day, we have to break our recommendations down by capacity. If 128- and 256 GB-class offerings are big enough for you, then it's hard for us to get behind the smaller M500s due to their performance. Conversely, the 480 and 960 GB models are genuinely quick. Speed alone doesn't make an SSD worth buying of course. Enhanced security and reliability features rate way up there, too. Ingenuity might be a better selling point, except that ingenuity rarely shows up on benchmarks. So, we continue plugging away at our M500s and 840 EVOs to get a better sense of how long they'll last.
Not Very Photogenic, Eh?
Now that the 960 GB versions of the M500 are plentiful (they weren't at first), the worst thing we can say about the behemoth is that its metallic label gets marred if you so much as breathe toward it. This makes for bad beauty shots. But we're still pretty enamored with the hardware underneath.





