1. 21
  1.  

  2. 8

    This is a really good talk and I appreciate his argument and others like it. Experimenting with new hardware architectures and systems together holistically is the way to go. It’s been quite demoralizing, for example, to see so much of the RISC-V discussion center on running Linux, when it also opens the gate up for making wholly new system designs that are completely open. There is generally a lack of imagination here and it’s sad, especially considering all the weird OS/hardware/systems ideas we have from the history of computing that perhaps should be revisited

    1. 3

      I agree, but from the perspective of this talk, RISC-V is merely one little detail of a CPU, the ISA used by the operating-system blob sitting in the corner. It has nothing to do with all the other components like power management, graphics or wireless that make up most of an SoC.

    2. 2

      Is there a transcript available? I looked for it and would appreciate it.

      1. 1

        Have you tried the builtin auto transcript from YT?

        1. 2

          Well I tried DeepSpeech and the result is semi-terrible

      2. 1

        Really good talk. I hadn’t realized how much other stuff is going on in modern computing hardware and how many of those components are their own CPUs with their own software.

        I was surprised he didn’t mention Apple at some point, because they’re about the only ones integrating OS and chip design. (“If you’re serious about software, you should build your own hardware” — Alan Kay.) Is it any wonder their ARM systems have been so successful, able to make so much of their power budgets? It’s just a shame we’ll never find out most of what they’re doing.