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

Superconducting Microprocessors? Turns Out They're Ultra-Efficient (2021)

spectrum.ieee.org

Post reply on HN
Print view View original

81–90 of 100 posts

Re: Superconducting Microprocessors? Turns Out They're Ultra-Efficient (2021)

#81

Post by Terr_ »

coryrc wrote: post #79
JumpCrisscross wrote: post #6

Earlier quoted context omitted.

> someone else needs to prove a negative, that there exists no possible technology that will "disrupt" cryocoolers Look at the Wikipedia references for crycoolers [1]. Note the dates and volume. Now look at room-tempuerature superconductors [2]. 1990 vs 2023. 5 vs 57. OP is arguing that a greater fraction of high-temperature superconducting research dollars might find purchase in improving the cryocooler than we pres…

The latter aren't being researched for CPUs, which are small things. They're for applications like long-distance power transmission, electric motors, and more. Things which aren't feasible to cryocool.

> long-distance power transmission

IIRC unlikely to change quickly even with higher-temp superconductors, since it would mean splicing in new power-grid segments that transfer direct-current instead of alternating, and then you have losses in conversion too.

Re: Superconducting Microprocessors? Turns Out They're Ultra-Efficient (2021)

#82

Post by Jensson »

Grimblewald wrote: post #29
moralestapia wrote: post #22

Earlier quoted context omitted.

>even accounting for cooling No.

stuff like this makes me wonder what the distribution on human context limits is.

It is very small, but humans tend to internalize knowledge as they read giving essentially infinite but lossy context length. Those posters failed to internalize the message here so they get the wrong knowledge out of the message.

Re: Superconducting Microprocessors? Turns Out They're Ultra-Efficient (2021)

#83

Post by numpad0 »

actinium226 wrote: post #54
timschmidt wrote: post #42

Earlier quoted context omitted.

Some versions of the venerable MOS Technologies 6502 have only 3,218 transistors. The Intel 8080 has somewhere between 4,500 and 6,000. 5k transistors is square in the middle of "plenty for a classic 8 bit micro". Enough to run a basic *nix or embedded RTOS.

There's something I'm missing, they say the prototype hits 2.5 GHz, how is that possible if they only have the equivalent of 5k transistors? Or is clock cycle independent of transistor count?

CPUs are like fancy tally clickers with an auto-clicker attached. A CPU operating at 2.5 GHz means the clicker is depressed 2.5x10^9 times/second.

What to do with that click speed is technically an entirely different matter.

Re: Superconducting Microprocessors? Turns Out They're Ultra-Efficient (2021)

#84

Post by jonjenk »

marcosdumay wrote: post #69
lacoolj wrote: post #62

This will be incredible when the cooling at consumer-level has been figured out. I do wonder what kind of material requirements/availability there will be though. Even things like touchscreens and other rare earth metals are either getting scarce or controlled by one or two countries. Regardless, exciting news here for all of us

Lots of people handle liquid nitrogen in all kinds of different field applications. Cooling things is not that hard.

Right. But this requires cooling to liquid helium temperatures. That's a separate level of hard and cost compared to liquid nitrogen.

Re: Superconducting Microprocessors? Turns Out They're Ultra-Efficient (2021)

#85

Post by AnthonyMouse »

Tuna-Fish wrote: post #73
pennomi wrote: post #71

Earlier quoted context omitted.

Doesn’t have to be compatible with life, just compatible with computing.

You misunderstand. The problem with using time dilation is that you can only make the clock go slower, not faster. If you wanted to exploit time dilation to get a computer result faster (in subjective time), you don't send the computer to an exotic locale, you go there yourself , while you leave the computer to do work in normal space.

TIL that time dilation can only really be used to make something go slower than normal space, not faster.

Sci-fi authors owe some apologies. I'm looking at you, Star Trek.

Re: Superconducting Microprocessors? Turns Out They're Ultra-Efficient (2021)

#87

Post by heavenlyblue »

oceanplexian wrote: post #31
philipswood wrote: post #25

Earlier quoted context omitted.

Energy is expended when you zero or set a bit. If you compute reversibly you need use special logic gates to not throw any bits away during the computation, like the Toffoli gate. All your operations need to have the same number of input and output bits and needs to be able to run forwards and backwards. Effectively you set or zero no bits during the entire computation that can't be losslessly reversed. If you struct…

My analogy probably isn’t perfect but isn’t this how an abacus or a slide rule works? You expend energy to set the values for computation but the act of computing and then reading them expends no energy.

That's a bad analogy because moving stuff in abacus is the computation itself.

Re: Superconducting Microprocessors? Turns Out They're Ultra-Efficient (2021)

#88

Post by Animats »

4 bit CPU, about 10,000 gates, 5 GHz. So just a proof of concept at this point.

Cyrogenic computing has been too far outside the mainstream. The mainstream technology improved faster than the cyrogenic stuff.

NSA put large amounts of money into cyrogenic computing, from the 1960s on. "I want a thousand-megacycle computer. I'll get you the money!" - an NSA director. It never really worked out, although at one point some special purpose device, probably a key tester, was actually built. The first round of that cyrogenic technology used cyrotrons. Cyrotrons were fast, but, being magnetic devices, not small enough. The second round used Josephson junctions. NSA finally gave up on that around the time ordinary CMOS passed 1GHz. Lately there's been some interest again.[1]

[1] https://spectrum.ieee.org/will-the-nsa-finally-build-its-sup...

Re: Superconducting Microprocessors? Turns Out They're Ultra-Efficient (2021)

#89

Post by tcgv »

Animats wrote: post #88

4 bit CPU, about 10,000 gates, 5 GHz. So just a proof of concept at this point. Cyrogenic computing has been too far outside the mainstream. The mainstream technology improved faster than the cyrogenic stuff. NSA put large amounts of money into cyrogenic computing, from the 1960s on. "I want a thousand-megacycle computer. I'll get you the money!" - an NSA director. It never really worked out, although at one point so…

That's the way technology evolves, some bets pay off, while others don't.

Re: Superconducting Microprocessors? Turns Out They're Ultra-Efficient (2021)

#90

Post by lupusreal »

heavenlyblue wrote: post #87
oceanplexian wrote: post #31

Earlier quoted context omitted.

My analogy probably isn’t perfect but isn’t this how an abacus or a slide rule works? You expend energy to set the values for computation but the act of computing and then reading them expends no energy.

That's a bad analogy because moving stuff in abacus is the computation itself.

Suppose you put the abacus sideways, such that setting a bead requires working against gravity, but then returning the bead to the original position gives you that energy back?
Post reply on HN