Announcement

Collapse
No announcement yet.

Coreboot Improvements For FU540 Land Following SiFive's Open-Source Boot Code

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Coreboot Improvements For FU540 Land Following SiFive's Open-Source Boot Code

    Phoronix: Coreboot Improvements For FU540 Land Following SiFive's Open-Source Boot Code

    Last week SiFive published their HiFive Unleashed open-source boot-loader code for this first RISC-V SoC on their Linux-friendly development board. This code being open-sourced has already helped improve the support for the FU540 SoC within Coreboot...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    Hope we can soon see some cheap RISC-V boards, something like a Raspberry Pi 3.

    Comment


    • #3
      Originally posted by uid313 View Post
      Hope we can soon see some cheap RISC-V boards, something like a Raspberry Pi 3.
      We will, when RISC-V is mass adopted.

      Comment


      • #4
        Originally posted by tildearrow View Post

        We will, when RISC-V is mass adopted.
        Well its not going to get mass adopted before it is cheap.

        Comment


        • #5
          Originally posted by uid313 View Post
          Hope we can soon see some cheap RISC-V boards, something like a Raspberry Pi 3.
          Unlikely. Current estimates seem to point at more than 5 years in the future.

          Comment


          • #6
            Originally posted by uid313 View Post
            Well its not going to get mass adopted before it is cheap.
            And it's not going to become cheap until it is mass-adopted. Catch-22. Consumer market is not a target and is going to be overall irrelevant for its success or failure. ARM didn't get into cheap crap like Raspi until it was very well-established in other (far more profitable) markets.

            It's going to be all decided by the large IC corporations that make their own custom ASICs, if they think it's worth pursuing then it will proceed to next stage.

            Comment


            • #7
              Originally posted by starshipeleven View Post
              Unlikely. Current estimates seem to point at more than 5 years in the future.
              It saddens me to hear estimation of so slow progress. Would really love to see RISC-V take over the world, and a cheap RISC-V board similar to the Raspberry Pi.

              Originally posted by starshipeleven View Post
              And it's not going to become cheap until it is mass-adopted. Catch-22. Consumer market is not a target and is going to be overall irrelevant for its success or failure. ARM didn't get into cheap crap like Raspi until it was very well-established in other (far more profitable) markets.

              It's going to be all decided by the large IC corporations that make their own custom ASICs, if they think it's worth pursuing then it will proceed to next stage.
              Yeah, I guess you're right.

              So far Nvidia seems to be on board, and Western Digital. Samsung and Seagate along with Google is on the membership roster too.

              Hopefully Siemens standardize on it for industrial systems and SCADA.
              Hopefully some military adopts RISC-V.
              The Chinese and the Russians should definitly get behind RISC-V so they can have an open architecture and open processor and be free the U.S.
              The U.S should produce RISC-V in factories in the United States so they don't get their chips backdoored at the factories in China.

              Comment


              • #8
                Originally posted by uid313 View Post
                It saddens me to hear estimation of so slow progress. Would really love to see RISC-V take over the world, and a cheap RISC-V board similar to the Raspberry Pi.
                Your unrealistic expectations are getting annoying.

                I want a PC I can actually trust, and Risc-V isn't really a silver bullet for that. Nothing stops any manufacturer from just making a risc-V Management Engine core and integrating with another (bigger) Risc-V "app processor" (which is what the CPU of a PC is), and given the companies these things are backed by, I'm not really holding my breath for that.

                Risc-v is supposed to make designing electronic devices easier for OEMs, and to be more scalable than other designs as it learned the lessons of the last 2 decades, that's all, it offers no guarantee to the users that the opensource core designs will not be used to screw over the customers.

                So no, I don't see why Risc-v taking over the world would even matter. And I'm not seeing why you would want it in cheap fire-and-forget crap embedded devices. It's not adding any benefit over ARM, for all its users it's going to be transparent change.
                It won't help risc-v spread in the slightest, as it will reach that kind of market only after it has a stronghold in the OEM embedded device market already.

                I'm much more interested in what Raptor Computing Systems is going to do in a few months, and really hope they have a decent mini-atx mobo design with a single CPU socket for Power CPUs, with a (relatively, for what it is) affordable price, like 1k dollars or so.

                I mean, I'd buy the Talos II Lite too if it wasn't a half-populated dual-socket E-ATX board with just two pcie slots.

                Hopefully Siemens standardize on it for industrial systems and SCADA.
                Hopefully some military adopts RISC-V.
                The Chinese and the Russians should definitly get behind RISC-V so they can have an open architecture and open processor and be free the U.S.
                The U.S should produce RISC-V in factories in the United States so they don't get their chips backdoored at the factories in China.
                This is just naive. Because of course US will never allow any company to install a backdoor in their own hardware right? Intel ME and AMD PSP are collective hallucinations, they never truly existed.

                Really, for Risc-V to matter for what I said above (i.e. making hardware that is not screwing the user) a massive reduction in manufacturing complexity for integrated circuits has to happen, so that many smaller companies can appear around the world, and rich people or small companies can actually get their SoC designs manufactured for a small niche of users. This is the only thing that would allow people to still have access to hardware that isn't designed to screw them over with ME-lookalikes.

                Like what happened with electronic board manufacturing, now you can get low prices even for short runs (as long as your board is hobbyist-type, so a couple layers, and not very large). But I'm not seeing that happening with silicon. It's well into the land of diminishing returns, the push is to consolidate all manufacturers into a single large one to decrease the costs. Even Intel is starting to offload to external fabs https://www.extremetech.com/computin...uction-to-tsmc

                Comment


                • #9
                  unapproved post for uid313 above this

                  Comment

                  Working...
                  X