Akhetonics Photonic Computing: 1,000x Faster CPUs
medium.com52 points by peter_d_sherman 3 days ago
52 points by peter_d_sherman 3 days ago
So they have some kind of nonlinear material that can make light modulate other light, that they managed to get working at teraherz frequencies?
Sounds incredible, as in: can't believe it.
By the way, isn't Nvidia researching this?
https://en.wikipedia.org/wiki/Nonlinear_optics
Very real stuff (I have a PhD on this)
This is the second sentence:
> Nonlinear phenomena become relevant only when the input light is very intense. Typically, in order to observe nonlinear phenomena, an intensity of the electromagnetic field of light larger than 108 V/m (and thus comparable to the atomic electric field of ~1011 V/m) is required.
I don’t know anything about this field. Can you address this point? How close are we to overcoming this in academia?
10^8 V/m, yeah that is quite a lot, 100V per micron or 1MV/cm. Even if it works it's going to have reliability issues, especially since the structures are not going to be perfect.
At milliwatt power, and terahertz frequency?
I guess it's a theoretical limit as in the material can change properties that fast given a fast light signal.
Yes; this material is called "snake oil".
Let's see if the investors bite.
I wouldn't infer much from that. Investors are in the business of taking risks and making gambles.
Also "FooCo raises $X zillion round" doesn't mean they got $X zillion deposited into their bank account. There are always several milestones along the way to unlock each tranche. But the press releases never mention that.
Finally! This paves the way towards practical BrainF*** execution.
We're all adults here, you can spell out BrainFuck just fine.
maybe it's brainf**, the pointer heavy formally verified brainfuck dialect
It's just as if he had the brilliant idea to do photonics as an outcast even though there is already a 6" wafer semi-production level fablab being built for it in The Netherlands. There is even a version of TinyTapeout for Photonics.
This comes up every couple of years, but it hasn't seem to have made any advances since the first time I've seen this.
The channel's other uploads may be giving us a glimpse into this project's future.
Interesting topic, unreadable due to sloppification.
"The first thing to understand about Akhetonics’ optical computer is that it’s not one thing yet." Great, that is such a good explanation. Thanks ChatGPT I guess.
> Photonics
So also 1000x smaller integration
Why?
Light is hundreds of nanometers, smallest "transistor" cannot be much smaller than wavelength.
I didn’t read but I’d also like to point out photonics are 1000x larger than leading edge electronics with minimal path to scaling smaller. I don’t see a free lunch
Basically the density of early CPUs (1970s), but 1000x faster than those, or current CPUs? Like a 1GHz 6502, or a 1THz 6502? The former is not really better, but the latter might interesting.
There ain't no free lunch, but real progress rarely depend on it. It's usually lunch we weren't eating before, or a different way to eat the same lunch but with much more satisfaction.
In honour of Akhenaten? Cool.
References of Akhenaten is making rounds in modern pagan/occult circles as well. The "Law of One" group is really pushing ancient Egyptian 'pyramid power' stuff.
This has been tried by Cray like 15 years ago and the university of Colorado 11 years ago; there were also other initiatives, all failed.
I'm not excluding that now we have more budget/time/AI to make it work (if it can), I'd just say exercise caution reading this (very AI slop) article (and anything that looks too good to be true)
interesting that it is still 2 bit
Do you mean as opposed to operating in an analog fashion? My guess is because keeping it digital means it is as close a possible to a drop-in replacement for today's electronic digital computers, including all the firmware, OS, and software that has already been written for them.
akhetonics-photonic-computing 1000x-faster-cpus, in mice
If you're looking for the limits of this approach: it's about 1000x faster than electric BUT the individual components have to be way larger (and can't be shrinked down)
... which hopefully means a LOT of fabs will be able to produce these (because a 2nm node makes no sense for this)
It also means this will be incredible for radio chips, interconnects, that sort of thing and will do exactly nothing for the first 20 years for cpu's, never mind gpus.
[dead]
[flagged]