Slashdot Mirror
'Breakthrough' LI-RAM Material Can Store Data With Light (ctvnews.ca)
A Vancouver researcher has patented a new material that uses light instead of electricity to store data. An anonymous reader writes: LI-RAM -- that's light induced magnetoresistive random-access memory -- promises supercomputer speeds for your cellphones and laptops, according to Natia Frank, the materials scientist at the University of Victoria who developed the new material as part of an international effort to reduce the heat and power consumption of modern processors. She envisions a world of LI-RAM mobile devices which are faster, thinner, and able to hold much more data -- all while consuming less power and producing less heat.
And best of all, they'd last twice as long on a single charge (while producing almost no heat), according to a report on CTV News, which describes this as "a breakthrough material" that will not only make smartphones faster and more durable, but also more energy-efficient. The University of Victoria calculates that's 10% of the world's electricity is consumed by "information communications technology," so LI-RAM phones could conceivably cut that figure in half.
They also report that the researcher is "working with international electronics manufacturers to optimize and commercialize the technology, and says it could be available on the market in the next 10 years."
And best of all, they'd last twice as long on a single charge (while producing almost no heat), according to a report on CTV News, which describes this as "a breakthrough material" that will not only make smartphones faster and more durable, but also more energy-efficient. The University of Victoria calculates that's 10% of the world's electricity is consumed by "information communications technology," so LI-RAM phones could conceivably cut that figure in half.
They also report that the researcher is "working with international electronics manufacturers to optimize and commercialize the technology, and says it could be available on the market in the next 10 years."
Um, 10% of the world's electricity is not consumed by phones. And even if they actually meant all computing and networking equipment combined, how is a RAM advancement supposed to cut all power consumed by computers, switches, etc in half?
Facepalm.
Very well; let this abomination unto the Lord begin!
The stupid summary leaps to the absurd conclusion that mobiles represent 100% of the power consumption made by ALL IT equipment. I'm pretty sure it's actually more like 1%; perhaps less.
It's the technology news/pr machine. I think we've had "breakthroughs" like this for ages, but what we didn't have for most of the time was a relentless, hype-oriented technology "press" that made us aware of them, and also spun them up into the next big thing. They were what they were, quiet little advancements that might or might not ever see the light of the day.
Yeah.... 5 years out for a tech claim means "we have a bit of experimental data that shows something might work. (Please fund me)".
10 years out in the tech world means "this is wild speculation and might never even become a technical demonstration. (Please fund my startup)"
So I did a very quick search on the internet looking for Light induced RAM and Light induced magnetoresistance and only found one article that predates the slashdot article and the one it links to. (Ok, I'm procrastinating from doing other stuff).
http://www.uvic.ca/home/about/...
This university published article is just as short on details and has no links to any published research. It's also a bit laughable: "new material allows computer chips to exist at a molecular level" which means what exactly? Computer chips currently don't exist at the molecular level? Anyway, don't mean to give their communications department a hard time, I just want more solid info.
It's clear that some of the claims from the hyped article that slashdot cites are ridiculous (at least the university release doesn't make those claims). The journalists, lacking any background in science probably called up some "experts" and said (out of context) "if you had a material that could do such and such" what would be the advantages. So, these experts, whether or not they actually know anything, just started making things up like it'll cut down on energy consumption (true but not a huge amount) and that it would prevent fires like the Samsung smart phone (probably not because the modest power savings from this RAM would not allow the battery to be designed differently which was the cause of the fires).
Unfortunately, the heat (and power) problems are not in the RAM but in the processor (amongst other things) which this technology does not address. In the university article they say that it is part of an effort to reduce the power and heat of processors but does not say this technology does this. Apparently, from the article, it is only suitable for RAM; hence the name LI-RAM. So while it may be faster (good) and not give off much heat (also good) it doesn't live up to the hype in the distorted media interpretations of the university article (which the slashdot submitter then chopped up and republished). This all assumes that they can get this to work at the fantastic performance and density levels of modern RAM all while not introducing new sources of heat and power to make it work (it requires "green light' presumably from a laser).
Anyway, if you want to waste some time, take a look at the Slashdot link and then look at the university article and you'll see how information can be mangled and hyped up by people who don't have a background in the subject. Of course, since we all like "free" (or ad supported) news, we aren't exactly encouraging accurate journalism :(
Yeah, you tell em! AI is never going to amount to anything, and batteries haven't changed in decades!
Now if you'll excuse me, I'm going to pull out my phone with a 3200 mAh battery the size of a couple-millimeter-thick business card and tell Google Photos to search through my thousands of unlabeled photos to show me just those that contain pictures of an arbitrary object.
Very well; let this abomination unto the Lord begin!
I'm sorry, but the person gave two examples, both of which *have* undergone major advances and made their way into our everyday lives. And if neural nets (what drives the image recognition) aren't from AI research then what are they from?
Very well; let this abomination unto the Lord begin!
Yeah.... 5 years out for a tech claim means "we have a bit of experimental data that shows something might work. (Please fund me)".
10 years out in the tech world means "this is wild speculation and might never even become a technical demonstration. (Please fund my startup)"
So, what's your solution? For all of the whining and moaning, and hand wringing, it seems that the answer for so many slashdotters is "Jeezuz NO! not another change! Not a breakthrough! Stop reporting on stuff!"
I can always see who works in science/tech - and some times who shouldn't be - by these posts.
Technology does not spring forth fully formed and beautiful, like Venus from the sea. Stuff takes time. In my field, it typically takes 20 years to develop a concept into a finished product. A few I've worked on were 50+ years from someone's concept to end development. Just depends on how far ahead of the curve the ideas go.
It isn't to say that there isn't bullshit. I recall a super radio antenna design from Rutgers that was claimed to be so efficient that the 100 watts pumped into it caused the antenna to melt. Parse that for a bit, and see if you can't come to a completely different conclusion.
Which is all to say, if we have a good reason to believe that a story is bogus, like in the above instance, an antenna melting is the very antithesis of efficient - by all means point it out.
But back to the idea that if there is a time period before deployment of 5 or 10 years, that it is bogus, well, you are applying a metric to technology that is just plain wrong.
A good hint is battery tech. They aren't inventing new elements, and we know what combinations will produce what. For a log time we've known. https://en.wikipedia.org/wiki/... Breakthroughs aren't often breakthroughs. The concepts and lab results very often need years of technological advancements in manufacturing processes to catch up. In the meantime, getting pissed or completely pessimistic about it is kinda a get off my lawn reaction.
The shepherds did so well protecting the flock that the sheep no longer believed that wolves existed.
So, what's your solution? For all of the whining and moaning, and hand wringing, it seems that the answer for so many slashdotters is "Jeezuz NO! not another change! Not a breakthrough! Stop reporting on stuff!"
My proposal would be to stop reporting on stuff that is 100% fluff, and 0% technical details.
The fact that they mention "smartphones and laptops" as if they're the only computers in existence is a hint that they don't realize that these two groups pale in comparison to desktop computers and servers in terms of power consumption.
#DeleteFacebook
Right, dead end, which is why they keep becoming better and better at various tasks, to the point that they're now entering our everyday lives? Because surely that's the very definition of a dead-end.
Does anyone remember how terrible voice recognition used to be 1-2 decades ago? Because I sure do. The concept that you could have things like Siri, Google Now, devices like Alexa, etc get it right the vast majority of the time would have been laughable. Neural nets used to be too bad to use in these tasks at all. When Google made their first neural-net based voice recognition system (rather than the algorithmic matching ones from before) it got a 25% error rate. Now it's down to 8%. They're cropping up everywhere - most recently Skype's real-time translation service. Which is starting out a bit imperfect, and I guarantee you, once the neural nets get better with time, people will promptly forget how they didn't used to be as good as they'll have become, just like happens with everything else neural nets do.
On the image side, it's not just about image recognition (say, Google Photos). Facial recognition has gone from fringe to dangerously accurate. In my last job (medical imaging) we used neural nets to segment the brain. Which I find to be a rather amusing concept, artificial neural networks studying biological neural networks ;) They started off rather poor at the job, but by the time I left they were doing a better job at it than humans. Neural nets are also better lipreaders than humans. Really, the number of fields they've been expanding into in the past decade, and the progress over the past decade, is really staggering. One "hard AI" task after the next, they're getting better than humans. Remember this XKCD comic from just a few years ago? You can now download software to do just that sort of thing. It's not just about computing power advances, either; the learning algorithms themselves have been advancing by leaps and bounds recently.
Now, if you don't mean neural networks when you say AI, then what the heck do you mean when you say AI?
Very well; let this abomination unto the Lord begin!