Slashdot Mirror

← Back to Stories (view on slashdot.org)

Clockless Computing

Posted by michael on from the beat-of-a-different-drummer dept.

ender81b writes "Scientific American is carrying a nice article on asynchronous chips. In general, the article advocates that eventually all computer systems will have to move to an asynchronous design. The article focuses on Sun's efforts but gives a nice overview of the general concept of asynchronous chip design." We had another story about this last year.

3 of 342 comments (clear)

  1. Re:clockless computing? by tripletwentie · · Score: 0, Flamebait

    they can't take away the clock. how will i know what time it is?
    or when it's time to patch up XP

  2. Re:Explanation, sorta by Beliskner · · Score: 1, Flamebait
    Slashdot is guilty of spreading FUD in this case. It has to be clocked somewhere, the processor's going to have to at least have D-latches on its inputs and outputs as the FSB is going to be synchronous (unless I missed a REALLY important memo). Processors are already asynchronous between the D-Latches. This completely asynchronous chip is a rebellion against Intel hyperpipelining. Any EE knows that you get the highest clock speed when the slowest component on your ASIC is the D-Latch, but this doesn't necessarily mean the highest performance, simple operations must go through too many D-Latches.

    Decreasing the clock speed and the number of D-Latches is a sensible idea, as D-Latches take up space on the die and make heat. This is a simple protest vote against Intel, rightfully so. In increasing the CLK of their processors at every other expense so much so they even hack their own processor by translatin it into microcode subinstructions. Stupid.

    --
    A caveman dreams of being us, the incalculable power and riches. We dream of being Q, then what?
  3. Re:Cooling actually does speed up asynch CPUs by Critter92 · · Score: 1, Flamebait

    He's totally making this shit up. Everyone knows that early 90's Caltech research into async computing was doomed since Professor Apgar was funneling NSF funds into his "do-undergrads-in-my-hew-hot-tub" fund and the grad students were spending all their time in beanbag chairs huffing nitrous and finding new uses for superglue. I doubt anyone from that lab even remembers 1993, let alone details about the test harness or latch design.