IBM Makes First Racetrack Memory Chip

holy_calamity writes "For several years, we've followed the progress of IBM's revolutionary 'racetrack' memory, which stores data inside nanowires for several years. Now Big Blue has made the first prototype integrated onto a single chip, using the CMOS processing technique used in commercial chip fabs. It's still a research prototype, but goes some way to validate IBM's claim that the technology could be commercialized."

Comments

I was disappointed

[Ukab the Great]

to find that the article has nothing to do with Deep Blue simulating a bunch of dogs chasing a fake rabbit.

Re:I was disappointed

[Forty Two Tenfold]

For several years, we've followed the progress of IBM's revolutionary 'racetrack' memory, which stores data inside nanowires for several years.

For several years, we've been complaining about slashdot's editors' inadequate skills for several years.

Re:I was disappointed

[cyclomedia]

I thought it meant the data could be stored for several years?

Finally

[GameboyRMH]

This will give computers the ability to gain performance through the use of red paint and stickers!

Re:Finally

[Anonymous Coward]

Stay away from the NASCAR version. It only works for programs stuck in loops.

Re:Finally

[jellomizer]

Will this will allow Programming to become a full fledged sport. Surround by folks around the country taking it as a opportunity to drink of oversize "beer" cans. Cheering as code fails to compile or causes a major system failure in the last 10 minutes.

Re:Finally

[demonbug]

Stay away from the NASCAR version. It only works for programs stuck in loops.

That's why I prefer the P2P versions, like Rally. Even with the F1 version Europeans seem partial to, no matter how fancy the execution gets you are still just running through a loop and end up back in the same place you started.

Dec 2010 Slashdot Comments

[clyde_cadiddlehopper]

Go back and read the comments on the Dec 2010 Slashdot item. A great deal of ire was vented over racetrack memory being in the 10-to-15-year-where's-my-flying-car distant future. And here we sit 12 months later with a functioning CMOS chip. I can't wait to fly around the block.

Re:Dec 2010 Slashdot Comments

[Rockoon]

And here we sit 12 months later with a functioning CMOS chip that stores only one bit per wire instead of actually being racetrack memory.

There, fixed that for you.

IBM rules

[lucm]

People underestimate the value of all the R&D IBM is doing. They spend a lot of money on this kind of research, and they do it seriously. And they don't look for the latest fad to blow the mind of consumers - they build for the long run.

Ok, their GUIs usually make my eyes bleed and the setup for some of their products is painful (Tivoli anyone?). But IBM is moving forward; their cloud offering, which was a complete joke a few years ago, is getting pretty good. Their stuff does not shine like Apple, it does not integrate like Microsoft, but it works pretty well.

Re:IBM rules

[gentryx]

IBM is the company which gets the most patents awarded. Every single year. Since decades. The don't do research out of goodwill, but for profit. Yes, not just shortsighted, but for the long haul. That's why they still exist. Since 100 years.

It's hard to compare IBM to Apple, since they target completely different customers: Apple is cosumers, IBM is business.

Re:IBM rules

[swillden]

When I was at IBM they admitted that they mostly used the patents to defend against lawsuits from other companies which were claiming infringement with their own patents.

Mmm, I think you misunderstood, or the presentation only addressed one sort of patents. IBM makes lots of money from licensing its patents. There is a difference, though, which is that the patents that IBM licenses (by and large, there may be exceptions) are "legitimate". They're real, serious advances in the art which are of real value to their licensees, and not something that any random engineer would come up with off the top of their head when faced with a similar problem. Most of the patents IBM licenses for big bucks are for things like lithography processes, techniques for increasing disk storage density... and this racetrack memory.

I no longer work for IBM, but I did spend 14 years there and while I have many (many!) other criticisms of the company, I think their approach to patents is a good one.

Re:IBM rules

[MobyDisk]

IBM may well cite that prior art in their patent application. Contrary to what Slashdot may make you believe, prior art does not invalidate patents. Most patents explicitly list the prior art that led them their. The fact that someone did this with a mercury delay line would in no way invalidate a patent on doing it electronically. This is certainly novel and not obvious. You wouldn't invalidate a patent on a transistor because someone created a switch out of water and gears 100 years ago.

This is exactly the kind of good research that patents are intended to protect. Companies spending time and money to try and solve a problem no one has solved before in order to advance technology. If IBM truly delivers a memory chip that is an order of magnitude smaller and/or faster than DRAM they deserve the royalties from that patent. We should happily pay it in each chip we buy knowing that the patent system gave them an incentive to push technology.

Re:IBM rules

[TheRaven64]

That was my first thought to. Of course, it's a reimplementation of mercury delay lines in the same way that an ion drive is a reimplementation of a V2, so it probably does deserve some credit...

Re:IBM rules

[lucm]

38% Unix market share and growing. Just dropped some dollars on Power7 HW running AIX a few months ago.

Power7 is making me question the theory of evolution because it is "Intelligent Design" ;-)

Re:IBM rules

[dkf]

IBM is the company which gets the most patents awarded. Every single year.

Sadly, in and of itself that means nothing. Patents vary widely in quality.

Yeah, but most of IBM's patents are the good ones. Won't say all — someone's bound to find a counterexample — but the vast majority are in areas where patents work fine and have always worked. (For example, if you're doing advanced chip design then you're working in an area where there masses of IBM patents, mostly expired.)

Patents are a problem only when they are unclear, over-broad, and insufficiently innovative. Alas, too many in the computing area are like that, but IBM's seem to be much less of a problem than most. Try picking on Apple or a patent troll instead.