New Material for Spintronics Discovered
Cpt_Corelli writes "Researchers at Uppsala University and the Swedish Royal Institute of Technology have discovered a new material with properties suitable for creating spintronic devices at room temperature. Previously this was only believed to be available at very low temperatures. The material is a combination of zinc oxide and manganite. The breakthrough is the cover item of the October issue of Nature Materials. If this new material proves viable for production there is an enormous potential for smaller and faster processors. Could this be the beginning of a new era in processor development?"
I read that the previous record -- from just a year or so back -- was -101c.
This is apparently huge, if the PR-blitz is to be believed.
Belief is the currency of delusion.
Here we come, won't that be great. 10Mfps in Quake4D, milliseconds from start to crash in windows.
But still connected to a low bandwidth connection (2Mbps) to an unreliable network with high contention rates and collisions.
Fast processors ceased to become something to get excited about since about 1999, 90% of people don't need them, 8% need more memory instead, and the final 2% do nuclear and climate simulations, work in industrial modelling, or SFX and animation.
An Eye for an Eye will make the whole world blind - Gandhi
Core Memory was around a long time ago. It provided non-volatile memory for a computer.
Isn't this just a molecular version of this idea?
Perhaps this is going to be the one that is going to change the bottleneck in the system from the slow memory to the newly slow processor. And the very slow HDD. And the very slow I/O.
Having made which cynical observation, I wonder what impact this could have on database client server? Keeping the database in memory? Multiway processors? It looks like the only people really able to make use of the technology are going to be at IBM, and possibly Sun.
Panurge has posted for the last time. Thanks for the positive moderations.
I remember reading an article on this technology about 15 years ago. The article said it would hold a few terabytes non-volitile in the size of a sugar-cube (2cm^2).
My immediate reaction was how would this affect programming and OS when the line between memory and storage is disolved. Not sure if the interface to CPU would be as fast as current memory, which means it would just be a storage mechanism.
If it could be used for primary memory, what happens to files and how they are viewed (logistically not physically). Would we need 'virtual' files on a RAM-disk or something more abstract?
Time will tell.....
-Steve
$7.95/mo, 200 GB disk, 2TBxfer, MySQL, PHP, RoR.
New material. Got that. But what makes it so special?
UNIX? They're not even circumcised! Savages!
It uses less power, too. MRAM is going to revolutionize every aspect of computing... big-horsepower things like PCs, yes, but ESPECIALLY PDAs.
I can't wait.
+++ATH0
It means that in 5 years, Moore's Law will stall out. Industry giants like Intel will refuse to make the huge investment to bring spintronics and other technology to market. Moore's Law will only continue at a crawl, and it will become only a function of heat sink size and weight. Processors 10x faster will only be so because their HSF will be 10x bigger. Prepare for extremely heavy desktop towers that become hot to the touch because the case itself becomes the heat sink.