HA!
-
Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.
Computing @ speed of light. New chip, millions of times faster than current machines.
- Thread starter mocoworm
- Start date
HA!
Chittagong
Gold Member
Finally with the speed of light the physics of Gran Turismo are possible in a game!
KojiKnight
Member
No. That's quantum computing. This is fundamentally the same as what we have now. Bits are carried by light instead of electrons. Quantum computers are the ones that can factorise huge numbers for breakfast.
efyu_lemonardo
May I have a cookie?
Could someone explain to me what electronic component this device could replace, in terms of functionality?
Because if it's meant to operate like a transistor, it's fucking huge..
That kind of seems to defeat the purpose, if it can't be shrunk down.
Because if it's meant to operate like a transistor, it's fucking huge..
That kind of seems to defeat the purpose, if it can't be shrunk down.
efyu_lemonardo
May I have a cookie?
Nothing happened, it's still decades away of becoming something that could be used in consumer products.
It might have happened, but it might not.
Ba-doom-tish.
Not so facetious a reply tho - a company called D-Wave reckon they've cracked it, others think their process is still classical computing.
no. NO!
Miami is 7500 miles away from Tokyo as the crow flies. Light in fiber optics takes 115 ms to make a round trip. That's a 115ms ping. Too much to play Street Fighter without lag, even with one fiber cable going straight from Miami to Tokyo.
115 ms >> 1 ms
If we manage to send messages through the center of the earth using neutrinos (This is pure scifi btw), at the maximum speed of light in a vacuum, the very best ping we can get is 47ms between the 2 most distant points on earth. 47ms is still 3 frames of input delay in street fighter @ 60fps, still not quite good enough.
YianGaruga
Banned
Sadly, we only got it working in the other timeline. Sucks for those of us that are stuck here
Yeah, I realised that when I re-read his comment and saw 'ms' and not second. Brain fart moment.
The ethical question now is so I edit the post...
If this happens, then automation will guarantee we will lose 70% of labor via deep learning and AI. We're going to lose between 30-50% of the labor force in the next 1-3 decades from said trends, but stuff like this would be even bigger boosts.
It can be so amazing yet so frightening, because we're still a very ignorant species.
It can be so amazing yet so frightening, because we're still a very ignorant species.
maniac-kun
Member
So quantum computing vs. Light computing. Both will take 10- 20 years to be realized for home devices.
ALL the GigaHertz
Lanrutcon
Member
How many N64s taped together?
slavesnyder
Member
Will it be better with Kinect?
MaKTaiL
Member
Is this going to replace transistors or just make them even smaller? Either way it is going to be pretty revolutionary if that's the case.
No because Kinect uses the Cloud, not the Light.
Ping is still limited by speed of light
Edit: Best part of this tech is using less power while being much faster. Battery tech didn't even have time to catch up :lol
1.21Gigawatts
Banned
Games will still run at 30FPS. Developers will just improve the graphics even further.
It's actually limited to the Speed of light in a medium, in this case light, which is significantly lower then c.
I'm not understanding what the Speed of light has to do with computational power, that part went right over my head in the article.....
Devil's advocate
Member
It could be used in special purpose processors, which ususally need much less transistors than general purpose ones.
That's why they talk about using it in specialized applications at first.
Linkstrikesback
Member
Quantum computing uses the polarised states of lights as the bits, which I might naively think that's what this chip is supposed to be doing using the beam splitter (Photon goes one way for one polarisation, and the other way for the other polarisation).
The other people in my physics office who work on this seem pretty convinced that it's very difficult to separate the two, at least just through using a beam splitter.
I remain sceptical of this until they have something that they've shown actually works, this is probably no better than D-waves "quantum" computer.
Typical bits in computers are either a "0" or "1" state. Occasionally, called up/down, or a number of other physics names.
At the moment, the states are typically changed by electrons. If you could instead change them with photons of light, which move orders of magnitude faster than electrons, you could in principle speed processing up hundreds of times.That is much easier said than done though.
Lanrutcon
Member
Haven't you heard? Polygon says consoles are dying, so by the time this tech is out we'll all be part of the master race
efyu_lemonardo
May I have a cookie?
Thanks for the explanation. Could you give some examples of special purpose processors that would benefit from this?
OminoMichelin
Member
It exists, but it doesn't have error correction yet (because how exactly are you going to correct something you cannot see?). I don't see it being remotely viable, with consumer computing.
This could, maybe, potentially, possibly be huge, instead.
Terra Firma
Banned
Will wait for v2.0.
C is the speed of light in a Vacuum. This is not significantly lower, it is fractionally lower.
Bjoern the Smexy
Member
Is this really true?
Because holy shit. I can't even wrap my head around this thing being another millions of times faster than current machines.
Unbelievable.
Welcome to the future.
Because holy shit. I can't even wrap my head around this thing being another millions of times faster than current machines.
Unbelievable.
Welcome to the future.
Juxo
Member
Underrated post.
So we will still have silicon but just drive it with light instead of electors. Can we really make a piece of silicon that will change its state reliably millions of times faster then the current ones we have? If not we will just transmit data from one transistor to next and to other CPUs in the server (farm) a lot faster then currently. And we are still going to be left with one of the biggest bottlenecks in computing we currently (slow RAM speeds)
TylerDurden4321
Member
Ansibles
Nemesis121
Member
Leyasu
Banned
Sigh. This is being misreported, as always. Photonic computing is not going to give us "million times" faster CPUs.
We compute with electrons because they interact with each other strongly. Light only interacts with itself weakly. You need much more power and larger components to make it happen and this is a limit imposed by physics, not engineering. A photonic computer might be possible but it would have to be either very dumb or very big and hot.
Where photonic computing *will* make a difference is in interconnects.
It would be wonderful if all RAM was on the same die as the CPU, so that lookup times were close to zero, but this just isn't feasible from manufacturing yield and heat management POVs. So instead we have a few MB of precious cache on-chip, with the rest of RAM on chips. But electrical signals only move so fast; waiting for the data to arrive from RAM puts a bottleneck on data-intensive computing.
Photonic computing has the potential to do those connections with light, making all RAM effectively as fast as cache.
If you want to see a radical increase in CPU power, look to graphene. That has the potential to allow the same sort of computing we already do with electrons, but 100x faster. Still plenty of work to be done there.
We compute with electrons because they interact with each other strongly. Light only interacts with itself weakly. You need much more power and larger components to make it happen and this is a limit imposed by physics, not engineering. A photonic computer might be possible but it would have to be either very dumb or very big and hot.
Where photonic computing *will* make a difference is in interconnects.
It would be wonderful if all RAM was on the same die as the CPU, so that lookup times were close to zero, but this just isn't feasible from manufacturing yield and heat management POVs. So instead we have a few MB of precious cache on-chip, with the rest of RAM on chips. But electrical signals only move so fast; waiting for the data to arrive from RAM puts a bottleneck on data-intensive computing.
Photonic computing has the potential to do those connections with light, making all RAM effectively as fast as cache.
If you want to see a radical increase in CPU power, look to graphene. That has the potential to allow the same sort of computing we already do with electrons, but 100x faster. Still plenty of work to be done there.
What about entangled particles? Let's have entangled polygons!
Do you have a link to the correct information then?