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

121–130 of 138 posts

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

#121

Post by unixhero »

52-6F-62 wrote: post #80
mitchty wrote: post #64

Earlier quoted context omitted.

Doubtful, the first purpose built memristor existed in 2008. Though to be honest we've seen them since the 1800's as oddities in certain electrical circuits. They've been theorized since 1971 mathmatically but until fairly recently haven't been a thing. It is the fourth fundamental circuit, give it a while before discounting it. And be happy you're alive when the fourth fundamental circuit was discovered and made rea…

Feynman also ranted about how much processing within the memory would speed computing in the 80's. He challenged those "computer scientists" to figure it out :P .... this lecture: http://cse.unl.edu/~seth/434/Web%20Links%20and%20Docs/Feynma...

Very insightful post. Thanks for that.

I skimmed the paper and enjoyed the illustrations, and the photos at the end :).

So Memristors are about reversible gates instead of non-reversible NAND gates?

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

#122

Post by eradicatethots »

coldtea wrote: post #45
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…

> I've never understood why people don't get more excited about memristors For the same reason they aren't excited about flying cars? Because they wait until something is actually delivered that in a form they can use?

Excitement happens before things are delivered by definition. This comes off as a pretty rude comment to me, with the excessive question marks implying that obviously people aren’t so stupid as to get excited about anything that isn’t 100% guarenteed to happen. Would like to see less comments like this on HN.

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

#123

Post by e12e »

mitchty wrote: post #65
1024core wrote: post #58

Earlier quoted context omitted.

10 years? I've been seeing them hyped as the "next big thing" since the 90s. HP keeps pumping out press releases every so often.

How? The first memristor was created in 2008.

Well, if they were 10 years away in 2008, that should mean the first 90% is done this year, then it'll be another 9 years for the remaining 10%? So 2026? ;-)

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

#124

Post by tim333 »

52-6F-62 wrote: post #80
mitchty wrote: post #64

Earlier quoted context omitted.

Doubtful, the first purpose built memristor existed in 2008. Though to be honest we've seen them since the 1800's as oddities in certain electrical circuits. They've been theorized since 1971 mathmatically but until fairly recently haven't been a thing. It is the fourth fundamental circuit, give it a while before discounting it. And be happy you're alive when the fourth fundamental circuit was discovered and made rea…

Feynman also ranted about how much processing within the memory would speed computing in the 80's. He challenged those "computer scientists" to figure it out :P .... this lecture: http://cse.unl.edu/~seth/434/Web%20Links%20and%20Docs/Feynma...

I think maybe the relevant bit - kind of a job to find in 16 pages:

>. Another plan is to distribute very large numbers of very simple central processors all over the memory. Each one deals with just a small part of the memory and there is an elaborate system of interconnections between them. An example of such a machine is the Connection Machine made at M.I.T. It has 64,000 processors and a system of routing in which every 16 can talk to any other 16 and thus 4000 routing connection possibilities.

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

#125

Post by colejohnson66 »

CamperBob2 wrote: post #115
smilekzs wrote: post #93

Earlier quoted context omitted.

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

The answer is no, but the sad part is, there's no real reason why not. An FPGA isn't really a "gate array," but a huge ensemble of small lookup tables that are made with what amounts to very fast static RAM. If major FPGA manufacturers didn't guard their bitfile formats with their lives, it's possible that entirely new fields of research would have emerged to find new ways to take advantage of the existing hardware i…

> If major FPGA manufacturers didn't guard their bitfile formats with their lives[...]

Why do they?

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

#126

Post by tim333 »

I got a little confused by the

>scientists have developed the first “in-memory computing”

as your normal GPUs have register and cache memory mixed with the processors. I think the novel feature is they are mixing the processors with non volatile flash like memory rather than with RAM. Which I guess is interesting but the "will speed up computers by 200 times" presumable refers to an old school architecture rather than something like the 15/125 TFLOP Volta GPU which I'd imagine is faster than their thing. (https://news.ycombinator.com/item?id=14309756).

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

#127

Post by ngold »

blennon wrote: post #3

"The result of the computation is also stored in the memory devices, and in this sense the concept is loosely inspired by how the brain computes." For anyone who is interested in a simple model of how the brain does this, check out "associative memories". The basic idea is that networks of neurons both store memory (in their synapses) and perform the computations to retrieve or recall those memories. A simple example…

Really fascinating, the low energy aspect is intriguing.

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

#128

Post by tigershark »

nyolfen wrote: post #24
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…

> I've never understood why people don't get more excited about memristors. personally, mostly because i've been seeing articles about how they're just about to totally upend computing for the last ten years

I still remember something like 25 years ago the revolution of clockless circuits published in SciAm..

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

#129

Post by cmiles74 »

MikeDoesCode wrote: post #44

Didn't HP design and I think prototype something like this with memristors, calling it 'The Machine'? edit So HP built one with 160Tb of memory, I remember it being proposed with memristors but haven't been able to check if the prototype used them... Does anyone know what is different about IBM's that let's them claim this as a first though?

As shmerl noted, The Machine turned out to be vaporware and they released something significantly different under the same name. HP couldn't mass produce memsistors. :-(

https://www.extremetech.com/extreme/207897-hp-kills-the-mach...

Post reply on HN