IBM New Memory: Faster, Better, than SSDs
IBM is developing a new memory technology called "Racetrack" that could replace all current forms of memory and storage solutions.
Earlier this month, IBM announced that it is currently working on a new form of storage-memory technology that could very well become universal, meaning that no longer will consumers need to choose between various types of memory/storage mediums. According to the company, this new memory--based on spintronics--is capable of storing more data in the same amount of space used in today's hard drives (10 to 100 times more data than flash on a single nanowire apparently), and also offers "lightning-fast" boot times, lower cost, stability, and durability.
What will make this new technology beneficial is that it combines the performance and reliability of solid state random access flash memory with the price and storage size of the magnetic hard disk drive, but without the cost and hardware limitations of both. As an example, hard drives may cost less and provided large volumes of storage, but are much slower than flash memory and feature moving parts that will eventually fail over time. Flash memory may be compact and fast to read data and write but has flash wear issues as well as being quite expensive. The new "racetrack" technology doesn't rely on an electronic charge to store data, but rather uses the "spin" of electrons and its associated magnetic moment. The new technology also doesn't use moving parts, so there's nothing to wear out, and unlike flash SSDs, can be rewritten to endlessly.
So how does this new technology work? A video provided by IBM offered a simplified definition, explaining that information is stored as a magnetic pattern on a nanowire. Pulses of special "spin-polarized" electrical current actually cause the patterns to "race" along the wire track; data can be read and written in less than a nanosecond, and can even be moved and read in either direction. Data is written by placing a second nanowire (with a special pattern) near the first nanowire. The data on the first wire changes by moving the pattern along the second wire.
IBM's Almaden Research Center, the team behind the project, believes that its "racetrack" memory could lead to solid state electronic devices within the next ten years. The group gives an example of an MP3 player using the technology, saying that it would be able to store around 500,000 songs or around 3,500 movies using the same amount of space offered today, built with far lower costs and requiring less power consumption, thereby generating less heat. Ultimately, the device would be indestructible, would not feature any moving parts, run on a single battery for weeks, and even last for decades.
"It has been an exciting adventure to have been involved with research into metal spintronics since its inception almost 20 years ago with our work on spin-valve structures," said Dr. Stuart Parkin, an experimental physicist leading the team. "The combination of extraordinarily interesting physics and spintronic materials engineering, one atomic layer at a time, continues to be highly challenging and very rewarding. The promise of racetrack memory - for example, the ability to carry massive amounts of information in your pocket - could unleash creativity leading to devices and applications that nobody has imagined yet."
Although the project is still in the early stages of development, Dr. Parkin and his team plans to move the racetrack technology into the third dimension with the construction of a "novel" 3D racetrack memory device.
- Intel Cuts Prices of Core 2 Duo, Quad Chips
- Dept. of Defense to Announce Cyber Command
- Time Warner Holding Back Super-fast Broadband?
- First Android Netbook Costs Just $100
- Duke Nukem Forever: Close to Completion?
- AMD Launches Low-Power 40-Watt Opteron EE
- Bluetooth 3.0 is Fast Like Ricky Bobby
- QOTD: Have You Bought An SSD Yet?
- Intel Says 6-Cores Not Delayed
- Report: Intel SSD Price Cut on April 26
- Seagate's Energy-friendly 5,900 RPM HDDs
- AMD: 6-Core Istanbul Available in June
- White House May Step In for Cybersecurity
- Find Illegal Torrents Using Google
- Report: MSI to Unveil Android Netbook in June
- Pirate Bay Judge Accused of Bias
- Apple: Netbooks Don't Deserve Mac Brand
- Nvidia CEO Denies VIA Investment Rumors
so when can I put this in raid for the home pc?
So, what's the catch?
But really, this sounds great. I would like to know when it's expected to come out rather than IBM saying "its "racetrack" memory could lead to solid state electronic devices within the next ten years."
What other options are we looking at in 10 years? Really by then this technology will probably outpaced by something else.
Im surprised IBM didnt call it Token Ring memory. The explination of how it works sounds alot like old Token Ring networks. Can you guess who invented Token Ring? Yep IBM.
Lee
Tom's hardware reported about this racetrack memory a couple of years ago. There isn't any information here that is new compared to that article.
Wow! This really sounds promising. Almost sounds too good to be true.
Remember spintronics helped bring us the huge hds we have today. Hopefully they move thru with development as quickly as possible so we can see this in a consumer product in under 10 years, that would be great
Wow! This really sounds promising. Almost sounds too good to be true.
Remember spintronics helped bring us the huge hds we have today. Hopefully they move thru with development as quickly as possible so we can see this in a consumer product in under 10 years, that would be great
well, it sure beats laser holographic storage mediums. Course 10 years away, we could be using optic fiber threaded processors or nanotube chips along with graphic cards that will multicore ghz processing power.
I'm going to cream my pants if I keep reading.... oh crap. 10 years? Nice, and hopefully they have a fully functional model soon, but for now HDDs are my choice.
I would love a 10TB thumb music/video player with 1G/sec+ read/writes that cost only a few dollars.
Not going to happen if IBM sells hard drives still. This sounds so good that they will either replace their whole hardware ecosystem or they will stuff it under the rug as an "experiment".
I will begin modeling racecars in 3DS Max immediately so that I will have appropriate data to store on it 10 years from now.
Same quotes even, IBM move forward on this project, get some demos going and show us the awesome. The only real difference i see is that the spin technology is more within reach now.
reading that made me happy.
so when can I put this in raid for the home pc?
5-10 years or more.
almost sounds like a DNA structure!
I'm sure the drive could be very solidly made, but calling it 'indestructible';especially the data?
Anything that relies on magnetic fields can be magnetically manipulated!
Meaning place the drive over a strong magnet,and what would happen?
Especially one atom layer material should be extremely sensitive!
almost sounds like a DNA structure!I'm sure the drive could be very solidly made, but calling it 'indestructible';especially the data?Anything that relies on magnetic fields can be magnetically manipulated!Meaning place the drive over a strong magnet,and what would happen?Especially one atom layer material should be extremely sensitive!
They secretly put some Chuck Norris sweat into the prototypes. Thus they were indestructible.
What other options are we looking at in 10 years? Really by then this technology will probably outpaced by something else.
Actually I think nanowire technology is the holy grail. 'Within 10 years' is different than '10 years'. It could be 2 years. It could be 5. It could be 10. It's hard to conceive of something that is more bleeding edge than this. After all, it uses spintronics.
I wonder if this has anything to do with the rumors of that UFO crashing near IBM's headquarters... j/k, j/k
nice sounding.
i wonder if the HD companies will squash this one too. seems like a lot of really cool sounding storage technologies surface and then nobody hears from them again.
The catch? Reliably engineering the materials.
If we could do what needs to be done to make this right now 32nm processors would (probably) be old hat.
They are talking about atomic level material manipulation, not nano level. Pretty big difference with a separate set of hurdles. Its a material scientist's wet dream, to reliably control structure at an atomic level.
10 years is 53 dog years. This is way to long for technology.
In 10 years I'll already have a quad Raid 0 SSD setup AND by then SATA III or faster will be released. In 10 years we will all be using different computers that are way faster than they are today. Buy then there will be a higher demand for a faster higher capacity SSD as stated.
In 10 years I'll already have a quad Raid 0 SSD setup AND by then SATA III or faster will be released. In 10 years we will all be using different computers that are way faster than they are today. Buy then there will be a higher demand for a faster higher capacity SSD as stated.
SATAIII was developed at the same time as SATA1 and 2. I believe I read about it on Tom's Hardware (a few years ago!), and I expected all 3 to be released in close succession for various applications. Nope... it's more profitable to release in slow increments to make people want to keep replacing stuff.