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

New Optical Switch Up to 1000x Faster Than Transistors

spectrum.ieee.org

Post reply on HN
Print view View original

41–50 of 94 posts

Re: New Optical Switch Up to 1000x Faster Than Transistors

#41

Post by singularity2001 »

dogber1 wrote: post #29
pg_bot wrote: post #13

Earlier quoted context omitted.

Isn't this just a trade off? Is there never a scenario where you would trade transistor density for switching speed and lower power consumption?

That's indeed done all the time in electronics: for example, RF CMOS usually trailing on a node three or four generations behind the bleeding edge. However, all-optical/photonic computing is just intrinsically so much worse than electronics. On top of the issues that I touched on, there are also other fundamental problems, e.g. distribution of power: photons like to get absorbed by nearby electrons. How do you then s…

Could nonlinear wave interactions be applied in near vacuum, isolated from the lasers, amplifiers and counters? Think 100000*100000 imprecise loss-full tensor/matrix multiplications.

Re: New Optical Switch Up to 1000x Faster Than Transistors

#43

Post by ithkuil »

tragictrash wrote: post #19
dogber1 wrote: post #5

In reality, all-optical computing is mostly a terrible idea: fundamentally, it cannot reach the integration density of electronics. It boils down to the elementary differences between Fermions (electrons, neutrons, etc.) and Bosons (photons, etc.). Their intrinsic behavior determines the interaction with matter, i.e. conductive/absorptive properties. As a result, optical wires (waveguides) have to be sized roughly at…

This sounds to me like doctors claiming they know so much about the human body when in reality we are at the infancy of our understanding

We're in an interesting point: there is so much we don't know but in order to learn more everything we do must fit in the already immense amount of knowledge that we accumulated so far. In the vast majority of cases this requires that the people who want to nudge the frontier a bit further must first dedicate a good portion of their life's studying what we know, and as the result sounding a bit arrogant when they explain to a layperson that actually they know what they're talking about. Yes, in some cases, they may be erring on the side of too confidence, but in many many cases is the layperson who doesn't fully grasp the ramifications of the innocent looking alternatives.

Re: New Optical Switch Up to 1000x Faster Than Transistors

#44

Post by hrgiger »

perihelions wrote: post #17

I *think* this is a recent-ish preprint of the same paper they're talking about. (The current Nature version is neither open-access nor accessible on Sci-Hub). https://arxiv.org/abs/2005.05811 (E.g., both papers contain this [0] text string pointing to the same URI. But, the arXiv preprint doesn't mention the Nature submission). [0] " All data supporting this study are openly available from the University of Southamp…

Also IBM announcement here: https://research.ibm.com/blog/single-photon-switching

Re: New Optical Switch Up to 1000x Faster Than Transistors

#45

Post by ajuc »

singularity2001 wrote: post #40
pg_bot wrote: post #13

Earlier quoted context omitted.

Isn't this just a trade off? Is there never a scenario where you would trade transistor density for switching speed and lower power consumption?

Exactly. While this speed vs space trade off makes less sense in mobiles, it might make perfect sense in industrial settings. Imagine 3D computers the size of a room (Craigh 2) but a 1000 times faster than any TPU only cluster.

If you can trade single-core speed for parallelism cluster of traditional electronics make more sense, but for some algorithms you can't.

Imagine a CPU with the complexity of Arduino but running at 100 GHz.

Re: New Optical Switch Up to 1000x Faster Than Transistors

#46

Post by Goz3rr »

rasguanabana wrote: post #39
dogber1 wrote: post #5

In reality, all-optical computing is mostly a terrible idea: fundamentally, it cannot reach the integration density of electronics. It boils down to the elementary differences between Fermions (electrons, neutrons, etc.) and Bosons (photons, etc.). Their intrinsic behavior determines the interaction with matter, i.e. conductive/absorptive properties. As a result, optical wires (waveguides) have to be sized roughly at…

It’s not all about computing. It’s about avoiding conversion from electrical to optical signal (and back) at every network node, which is costly.

Don't you need a certain amount of computing at each network node anyways to see what to do and where to send the optical signal next? In additional to error correction/amplifying the signal?

Re: New Optical Switch Up to 1000x Faster Than Transistors

#47

Post by rocqua »

Goz3rr wrote: post #46
rasguanabana wrote: post #39

Earlier quoted context omitted.

It’s not all about computing. It’s about avoiding conversion from electrical to optical signal (and back) at every network node, which is costly.

Don't you need a certain amount of computing at each network node anyways to see what to do and where to send the optical signal next? In additional to error correction/amplifying the signal?

Generally you only need to read the 'header'. If that is little enough computation maybe that can be done optically, gaining the advantage of not needing to convert twice.

Re: New Optical Switch Up to 1000x Faster Than Transistors

#48

Post by hrgiger »

rewq4321 wrote: post #38
jcims wrote: post #31

Earlier quoted context omitted.

(I swear I’m not in Fridman’s payroll.) As a layperson I found this episode with Jeffrey Shainline an interesting discussion tangential to the topic of optoelectronic computing. The basic gist was that photons are good for communication, electrons are good for compute. https://youtu.be/EwueqdgIvq4

The specific timestamp for that part of the conversation: https://youtu.be/EwueqdgIvq4?t=2793

Not in the domain but as far as I understand from video for the counter argument: "Adding a matter of component and creating this kind of liquid of the light you add interactions in the system so practically light interacts more efficient with the light"

https://youtu.be/Kv25Dw-IuCM?t=675

Re: New Optical Switch Up to 1000x Faster Than Transistors

#49

Post by rocqua »

neltnerb wrote: post #11
sleepy_keita wrote: post #3

Earlier quoted context omitted.

> the new optical switch works at room temperature

These sound a lot like room temperature optical qubits... I'm certain I am missing something but this sounds pretty neat.

They probably aren't in a superposition, let alone entangled with each other.

Re: New Optical Switch Up to 1000x Faster Than Transistors

#50

Post by iptrans »

dogber1 wrote: post #10
anko wrote: post #7

Earlier quoted context omitted.

how about for an optical switch? for switching network packets over fibre?

Switching/routing usually requires significant information processing (e.g. decode packet header, match destination address against routing tables, etc.). This necessitates 10k or more gates. All-optical computing can't deliver this level of integration density, nor the performance at reasonable power levels. Maybe there will be some smart way to pre-encode routing information onto packets to reduce processing requir…

Being able to do any kind of computation in the optical domain would benefit telecommunications immensely.

Basic things like optical muxing/demuxing and serialization/deserialization would be fantastic.

Post reply on HN