5 ms·

I agree open source hardware would be great but SiFive is a closed source vendor like many others. It's RISC-V itself that has a permissive license, allowing a

I agree open source hardware would be great but SiFive is a closed source vendor like many others. It's RISC-V itself that has a permissive license, allowing anyone to make an implementation open or closed source.

cestith 4y ago
An open ISA is a good start. It means short of any closed-source extensions the code can move to another chip vendor. With smart modularization of the code, even parts using closed-source extensions could be potentially retargeted faster than moving to a whole new ISA. Finding a radiation hardened complete solution that's totally open source could be a tough request to fill.
- goodpoint 4y ago
  > short of any closed-source extensions the code can move to another chip vendor That's a huge limitation. It's unfortunate that RISC-V allows for closed implementations (IP) and also closed ISA extensions.
  - cestith 4y ago
    I think it's a rather pragmatic decision given corporate capitalism is the economic model under which most processor design and trade happens, at least toward the higher end. The good news is since the main ISA and several open extensions exist, any open source hardware implementation of those aren't hindered by the ISA. That's also true for SPARC, POWER, and a few others of course. But RISC-V seems to be the belle of the ball and apparently simpler to build a hardware design around. Someone, somewhere may eventually form a coalition to pay TSMC or Samsung to generate an in-house spec for a high-performance RISC-V processor with an open-source license on the hardware design, too. Some places use RISC-V in products where the processor is another line item in the BOM and not the centerpiece of a SOC or dev board, like Western Digital using it in storage products. A group of companies and institutions could very conceivably get things where we'd wish now that the open ISA has set the stage.
    - goodpoint 4y ago
      > rather pragmatic decision given corporate capitalism Supporting a system that centralized designs of critical semiconductors in the hands of few companies and manufacturing in the hands of TSMC is not very pragmatic. If anything it's very ideological.
      - cestith 4y ago
        Acknowledging that's the system under which something has happened does not imply, let alone assure, personal support for that system overall. The real world doesn't bow silently in deference to our hopes and dreams.
        goodpoint 4y ago
        > The real world doesn't bow silently in deference to our hopes and dreams. You are implying that socioeconomical systems are "the real world" and somehow they just happen to exist independently from our "hopes and dreams". It's the complete opposite. Unlike solar flares and ocean tides, socioeconomical systems are entirely the product of human decisions.
  - snvzz 4y ago
    >It's unfortunate that RISC-V allows for closed implementations (IP) and also closed ISA extensions. It is rather fortunate. Else, it couldn't possibly aspire to be the standard ISA for absolutely everything.

cestith 4y ago

An open ISA is a good start. It means short of any closed-source extensions the code can move to another chip vendor. With smart modularization of the code, even parts using closed-source extensions could be potentially retargeted faster than moving to a whole new ISA. Finding a radiation hardened complete solution that's totally open source could be a tough request to fill.

goodpoint 4y ago
> short of any closed-source extensions the code can move to another chip vendor That's a huge limitation. It's unfortunate that RISC-V allows for closed implementations (IP) and also closed ISA extensions.
- cestith 4y ago
  I think it's a rather pragmatic decision given corporate capitalism is the economic model under which most processor design and trade happens, at least toward the higher end. The good news is since the main ISA and several open extensions exist, any open source hardware implementation of those aren't hindered by the ISA. That's also true for SPARC, POWER, and a few others of course. But RISC-V seems to be the belle of the ball and apparently simpler to build a hardware design around. Someone, somewhere may eventually form a coalition to pay TSMC or Samsung to generate an in-house spec for a high-performance RISC-V processor with an open-source license on the hardware design, too. Some places use RISC-V in products where the processor is another line item in the BOM and not the centerpiece of a SOC or dev board, like Western Digital using it in storage products. A group of companies and institutions could very conceivably get things where we'd wish now that the open ISA has set the stage.
  - goodpoint 4y ago
    > rather pragmatic decision given corporate capitalism Supporting a system that centralized designs of critical semiconductors in the hands of few companies and manufacturing in the hands of TSMC is not very pragmatic. If anything it's very ideological.
    - cestith 4y ago
      Acknowledging that's the system under which something has happened does not imply, let alone assure, personal support for that system overall. The real world doesn't bow silently in deference to our hopes and dreams.
      - goodpoint 4y ago
        > The real world doesn't bow silently in deference to our hopes and dreams. You are implying that socioeconomical systems are "the real world" and somehow they just happen to exist independently from our "hopes and dreams". It's the complete opposite. Unlike solar flares and ocean tides, socioeconomical systems are entirely the product of human decisions.
- snvzz 4y ago
  >It's unfortunate that RISC-V allows for closed implementations (IP) and also closed ISA extensions. It is rather fortunate. Else, it couldn't possibly aspire to be the standard ISA for absolutely everything.