• atzanteol@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    16
    ·
    edit-2
    24 days ago

    Nifty. I love the idea of risc v but fear there is no real consumer market for it that isn’t better served by arm though.

      • atzanteol@sh.itjust.works
        link
        fedilink
        English
        arrow-up
        10
        ·
        24 days ago

        I’m not sure that’s big enough of an advantage though. I hope it is but history shows that most people are more than willing to pay for convenience over freedom.

        • Semperverus@lemmy.world
          link
          fedilink
          English
          arrow-up
          8
          ·
          edit-2
          23 days ago

          Check out MilkV, they have a 64 core RISCV workstation that supports PCIe at full speeds and has NVMe and SATA slots like a completely normal x86 motherboard.

          People have gotten modern AAA games running on it.

      • Alex
        link
        fedilink
        arrow-up
        5
        ·
        24 days ago

        If you license a design from someone you’ll still be paying something. Sure there are also free implementations but they are aimed at microcontrollers, you won’t get any server class chips for free.

      • LeFantome@programming.dev
        link
        fedilink
        arrow-up
        3
        ·
        23 days ago

        I do not think it is about the royalties in most cases. I mean, RISC-V royalties may be the reason you choose it over ARM for a custom chips ( say in the bajillion SSDs you are going to ship ). Perhaps you were going to choose a different ISA for a microcontroller and the lack of license fee makes RISC-V attractive.

        For chip maker, it is the freedom that matters as that is what “convenience” means to them. And it means less risk. Look at the Qualcomm / ARM lawsuits right now. That would not happen if Qualcomm had chosen ARM.

        And if you are a chip maker licensing core designs, do you want your ISA to force you into a monopoly? ARM is more mature today but the role that ARM the company plays is being filled by multiple RISC-V suppliers ( HiFive, Milk-V, etc ). More players means more completion means more choice and probably better prices. ARM’s core business is licensing chip designs and they are about to have a lot of competition from RISC-C.

        And in the end, competition from and within the RISC-V space will drive down prices for consumers. That is what consumers are going to care about. The lower prices will not really be because of lower license fees ( though that will help of course ). And it all comes with a large and open software ecosystem. So the “convenience” will be there too.

  • LeFantome@programming.dev
    link
    fedilink
    arrow-up
    10
    ·
    23 days ago

    I am torn on this chip. One way of looking at it is negative as they are adding custom instructions that are not part of the RISC-V standard. Part of me hates the fragmentation.

    On the other hand, the alternative is that they release another MIPS chip ( MIPS ISA, not RISC-V ). That obviously fragments the CPU space even more and does nothing to drive the RISC-V space forward.

    If I take a step back, this is exactly the freedom that RISC-V represents. Not only did it make sense for MIPS to adopt RISC-V over their own ISA but this is the kind of thing that would not be possible if they went with ARM.

    What makes RISC-V better than ARM is the freedom, not the lack of licensing fees. I think this is an example of how RISC-V wins in the end.

    Equivalent instructions will make it into the RISC-V spec ( official extensions ) and future MIPS chips will no doubt use the standard at some point.

    In the end, this just creates more demand for and more support for RISC-V on Linux and Open Source RISC-V toolchains ( such as compilers ).

    Anything that moves RISC-V forward is positive.