Study Opens Route To Ultra-Low-Power Microchips (mit.edu)

Freshly Exhumed writes: A new approach to controlling magnetism in a microchip could open the doors to memory, computing, and sensing devices that consume drastically less power than existing versions. The approach could also overcome some of the inherent physical limitations that have been slowing progress in this area until now.

Researchers at MIT and at Brookhaven National Laboratory have demonstrated that they can control the magnetic properties of a thin-film material simply by applying a small voltage. Changes in magnetic orientation made in this way remain in their new state without the need for any ongoing power, unlike today's standard memory chips, the team has found. The new finding is being reported today in the journal Nature Materials, in a paper by Geoffrey Beach, a professor of materials science and engineering and co-director of the MIT Materials Research Laboratory; graduate student Aik Jun Tan; and eight others at MIT and Brookhaven.

As silicon microchips draw closer to fundamental physical limits that could cap their ability to continue increasing their capabilities while decreasing their power consumption, researchers have been exploring a variety of new technologies that might get around these limits. One of the promising alternatives is an approach called spintronics, which makes use of a property of electrons called spin, instead of their electrical charge. Because spintronic devices can retain their magnetic properties without the need for constant power, which silicon memory chips require, they need far less power to operate. They also generate far less heat -- another major limiting factor for today's devices.

Comments

Toriod core memory anyone?

[bobbied]

Sounds like an improvement on core memory..

Re:Toriod core memory anyone?

[ShanghaiBill]

Sounds like an improvement on core memory..

... in the same way that fractional-step UV photolithography sounds like an improvement on a hammer and chisel.

Will storage and memory eventually become one?

[SuperKendall]

Reading through the abstract something that struck me was the statement "with no degradation in magnetic properties after >2,000 cycles".

With the increase in speed of SSD's all the time, and advances like this that don't suffer degradation, it made me wonder if at some point there would be no need for separation of RAM and SSD, if storage were fast enough you could just use as much of it as you liked for system memory.

Looking around at some specs it seems like at this point RAM may be just 10x faster than the best SSD's around, probably less now. I'm sure there will always be even faster L1/L2 cache memory chips to speed things up, but just thinking of the system RAM we all have today - there has to be a point where the primary storage is fast enough to take on that role and gain greatly improved system memory as a result.

Re:Will storage and memory eventually become one?

[fahrbot-bot]

Reading through the abstract something that struck me was the statement "with no degradation in magnetic properties after >2,000 cycles".

Just don't use it *before* 2,001 cycles, 'cause, man, it's all over the place.

Re:Will storage and memory eventually become one?

[shess]

Reading through the abstract something that struck me was the statement "with no degradation in magnetic properties after >2,000 cycles".

With the increase in speed of SSD's all the time, and advances like this that don't suffer degradation, it made me wonder if at some point there would be no need for separation of RAM and SSD, if storage were fast enough you could just use as much of it as you liked for system memory.

Looking around at some specs it seems like at this point RAM may be just 10x faster than the best SSD's around, probably less now. I'm sure there will always be even faster L1/L2 cache memory chips to speed things up, but just thinking of the system RAM we all have today - there has to be a point where the primary storage is fast enough to take on that role and gain greatly improved system memory as a result.

For a long time, the time it takes the transfer data from RAM to the CPU has been the bottleneck in performance. That is a large part of why so much of a CPU's die is devoted to cache - putting a ton of computational elements on the die is pointless if you can't keep them busy, and cache is also more straightforward to design. Switching out RAM for flash would cause a HUGE performance hit.

Beyond that, RAM can cycle continuously with no upper limit. Due to caches, most DRAM isn't going to cycle as frequently as SRAM in the cache will, but 2000 cycles is a trivial number when you're talking about RAM, you can pass that in fractions of a second. Many people are uncomfortable using SSDs for swap, which is a primitive way of using it as RAM.

I would be very surprised if we ever reach a point where there's some sort of unified storage and memory hardware, unless we freeze the target. Basically, anytime someone invents new technology, we generally use it to expand the parameters of the space to address new problems, rather than only using it to simplify the existing space.

Re:Will storage and memory eventually become one?

[mentil]

Indeed. If it were used for storage, it'd certainly cause a Space Oddity.

Everything old is new again

[Locke2005]

Anyone remember magnetic core memory? You used to be able to turn off the machine, then turn it on a week later have have it still be in the same state. We've just greatly improved the density of magnetic cores. (Sharp also had patents on magnetic memory chips.)

Re:Can retain their magnetic properties

[Ungrounded Lightning]

For how long?

Geologic time (absent extreme temperatures or magnetic fields) wouldn't surprise me.

Seabed minerals have retained the magnetization that they fossilized when they cooled below their curie point, creating a geologic record of Earth's magnetic field reversals in the spreading seabed. (That's how we know that/when the Earth's field reversed from time to time.)

While this isn't the same structure, electron spins flipping in an environment conducive to them being stable is not something that tends to happen spontaneously.