Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

IBM scientists say radical new ‘in-memory’ architecture will speed up computers

kurzweilai.net

Post reply on HN
Print view View original

91–100 of 138 posts

Re: IBM scientists say radical new ‘in-memory’ architecture will speed up computers

#91

Post by ericd »

digi_owl wrote: post #21
ericd wrote: post #17

Earlier quoted context omitted.

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.

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 compone…

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 will speed up computers

#92

Post by fao_ »

mitchty wrote: post #83
cag_ii wrote: post #81

Earlier quoted context omitted.

The concept goes back to at least '71 http://ieeexplore.ieee.org/document/1083337/

Correct on concept, but up until 2008 nobody produced one reliably. You can find references and hints to them all the way back to maxwell's day. They tended to be regarded as anomalous circuits back then however.

The fact that nobody has produced a fusion generator, does not mean that the amount of articles claiming that fusion is 'just around the corner' is any less.

Re: IBM scientists say radical new ‘in-memory’ architecture will speed up computers

#93

Post by smilekzs »

blattimwind wrote: post #77
otakucode wrote: post #73

Earlier quoted context omitted.

FPGA reprogramming is slow. It's like saying we can make more oil because all we have to do is wait a few hundreds million years for all our plant matter to rot. And correct me if I'm wrong, but can you even reprogram an FPGA while it is actively being used? I don't believe so. Much less reprogram it from the gates which are active.

> And correct me if I'm wrong, but can you even reprogram an FPGA while it is actively being used? You're wrong, this is called partial reconfiguration.

Can you reprogram the part of an FPGA that is actively being used?

Re: IBM scientists say radical new ‘in-memory’ architecture will speed up computers

#95

Post by iainmerrick »

hinkley wrote: post #51
BugsJustFindMe wrote: post #48

Earlier quoted context omitted.

> They were right. Who was right about what? > They are the equivalent of second generation laser printers. What does this even mean?

Roentgen and Retina displays are in the neighborhood of 200 ppi, which is about when laser printers became 'clear' and legible, which was the thesis behind Roentgen. But now you've got smart phones pushing in the region of 400 PPI. It's smoother, and make Apple displays look like laggards. But they're beyond the point where most eyes notice the difference without magnification.

But now you've got smart phones pushing in the region of 400 PPI. It's smoother, and make Apple displays look like laggards.

I think Apple displays look completely fine (much better than average, in fact, since they focus so much on color accuracy).

The ridiculous Android PPI arms race is one reason why iPhones are way ahead in all performance benchmarks. What's the point of all those extra pixels you can't even see if the GPU can barely manage to fill them?

Re: IBM scientists say radical new ‘in-memory’ architecture will speed up computers

#96

Post by posterboy »

vvanders wrote: post #29
jerf wrote: post #25

Earlier quoted context omitted.

"this ability to change from functioning as a bit of memory to being a bunch of functional logic on the fly at the speed of a memory read?" My excitement is tempered by considering the areal demands on the silicon for the putative "smart memory". Suppose, just for the sake of argument, you want your smart memory to be able to take a 4K block of 64-bit integers and add them together. It happens incredible quickly, sur…

Yeah, this seems like the exact same problems that we already have with FPGAs. 1. They're really hard to program for in a way that is easy to understand and scale. 2. They eat a ton of power. Heat = power = max speed. If you can't make it better then today existing ASICs are still going to be used. The intersection of ASIC + reconfigurable serial processes(aka sequential programming) strikes a really sweet point betw…

> 2.

so you don't see the potential combination of this and that?

Re: IBM scientists say radical new ‘in-memory’ architecture will speed up computers

#98

Post by deepnotderp »

otakucode wrote: post #2

This is the promise of memristors. Despite innumerable articles being written about neuromorphic architectures like they'll be something miraculous, this ability to change from functioning as a bit of memory to being a bunch of functional logic on the fly at the speed of a memory read? That's going to be crazy. It will open up possibilities that we probably can't even imagine right now. I've never understood why peop…

Well, first off, their endurance is currently not even good enough to replace DRAM, let alone replacing critical path transistors. They also take significantly more power to switch and are much slower. They also tend to be highly unreliable (they even have variable write time!) and hard to manufacture. Really, the only useful advantage they do have is their theoretically incredible density, but even most of this comes from the possibility of 3D stacking (also possible with transistors) and from the possibility of a crossbar memory, but crossbars have epic sneak path issues.

Re: IBM scientists say radical new ‘in-memory’ architecture will speed up computers

#99

Post by squeaky-clean »

Alex3917 wrote: post #56
cududa wrote: post #36

Earlier quoted context omitted.

Pretty sure I first read about them in a PC World magazine I got at the airport on a family vacation in like 2003

As far as I know I submitted the first ever article about memristors to HN, which was this in 2008: https://news.ycombinator.com/item?id=177865

HN has only been around since 2007

Re: IBM scientists say radical new ‘in-memory’ architecture will speed up computers

#100

Post by yigitdemirag »

So sorry to find this thread on HN while intoxication. I need to write a blogpost about it in a couple of days. In short, these devices are more capable and realizable than most can imagine.

Disclaimer: Neuromorphic computing with PCRAM devices is my MSc and future PhD thesis topic.

Post reply on HN