Re:IBM scientists say radical new ‘in-memory’ architecture w...
by otakucode on Friday October 27, 2017 @04:16PM
(#15571865)
Memristors are any device whose resistance changes in response to the voltage that's gone across it in a predictable and reversible way. The phase-change memory they describe is one type. Memristors can form any logical gate. They are essentially resistors whose value can be changed dynamically and persistently. Rather than having them store data directly (which you can do, sure, and we'll probably see devices doing that soon) they can form logic gates in much the same way transistors do - except they can be changed in configuration while running. They can form a set of gates that function as a traditional memory latch one nanosecond, then the next be part of an adder. It's like if re-programming was so profoundly free on an FPGA as to be exactly as burdensome as performing an AND.
Re:IBM scientists say radical new ‘in-memory’ architecture w...
by com2kid on Friday October 27, 2017 @04:46PM
(#15572098)
That sounds like the optimization problems with a VLIW architecture taken to a whole new level.
Re:IBM scientists say radical new ‘in-memory’ architecture w...
by ericd on Friday October 27, 2017 @01:55PM
(#15570697)
I'm not an expert on this, but I think you might be thinking of nvram/non volatile RAM, "memristor" is a portmanteau of memory and transistor, so I'd expect that they have properties of both.
Re:IBM scientists say radical new ‘in-memory’ architecture w...
by digi_owl on Friday October 27, 2017 @02:07PM
(#15570804)
Not quite. memristors are resistors with a memory.
Specifically as current is run across it, it builds up more and more resistance. Then as the current is reversed (the plus and minus ends of the resistor is switched) this resistance will go back down.
Thus it can be used as non-volatile memory in the sense that certain tiers of resistance can be designated as a bit.
Note that as it is fundamentally an analog component one can potentially also used this property to emulate a neuron.
Re:IBM scientists say radical new ‘in-memory’ architecture w...
by ericd on Friday October 27, 2017 @06:12PM
(#15572615)
Thanks very much for the correction! Removing my foot from my mouth, now.
That's incredibly interesting, especially the idea that it could be used to emulate neurons. I expect that it'd be hard to fabricate them with a very high branching factor, but as I've amply demonstrated, I know very little about this topic.
Re:IBM scientists say radical new ‘in-memory’ architecture w...
by markdog12 on Friday October 27, 2017 @02:19PM
(#15570913)
Currency? Is it a crypto-currency that runs across it? :)
Re:IBM scientists say radical new ‘in-memory’ architecture w...
by digi_owl on Friday October 27, 2017 @02:24PM
(#15570942)
bleeping butterfingers...
Fixed.
Re:IBM scientists say radical new ‘in-memory’ architecture w...
by deepnotderp on Friday October 27, 2017 @02:10PM
(#15570825)
Actually it's a portmanteau of "memory" and "resistor". That's important since it's basically a variable resistance resistor that "remembers" its resistance.
Is this article about memristors? That wasn't clear. Can memristors perform computations?