Announcement

Collapse
No announcement yet.

ASUS Unveils The Tinker V As Their First RISC-V Board

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

  • #11
    Originally posted by kiffmet View Post
    caligula and yet RPi still doesn't have VAAPI and relies on a custom kernel to get the most out of it. I've had a better experience with Rockchip SBCs tbh.
    stormcrow the part about side channel attacks also applies to A53 and A55 cores, since those are in-order aswell.

    What's disappointing about RISCV for me is that the CPU performance just isn't there yet. The AndesCore AX45MP in the Asus tinkerboard is clock for clock slower than a Cortex A15 from 2012. On the other hand, it's likely that ARM SBCs will start shipping A77 cores this year and IMO, with that CPU and following generations, things are starting to get really interesting…
    That many extensions of the RISC-V ISA are still drafts isn't particularily nice either, since compatibility may break in the future.
    *shrugs* Anyone that deploys these at scale gets what's coming to them when they try to replace dead or malfunctioning units down the line if the software deployed on them depends on quirks in the drafts. No one should be forgetting that prototype, draft spec boards are just for early preview, technology evaluation and adaptation.

    As for performance, I don't see that as a problem. The reason being is that many industrial IoT applications aren't CPU bound to begin with. Sensor sampling, switch control, register flipping, etc. don't need fast CPUs. If they're properly isolated from external threats, the only thing they have to worry about is potentially dealing with passing authentication tokens and there's a new NIST standard for just this sort of application environment. As a technology evaluation board, it can still be expensive and companies will buy it so long as the roadmap includes much cheaper boards for production deployment in the future.

    Regardless of the advisability of widespread adoption, proprietary extensions and GPUs will be the Achilles heel on RISC-V as traditional OEMs try to scramble for product differentiation in an increasingly crowded field - just like in the ARM sector. The ISA specifications will only take people so far.

    Comment


    • #12
      At this low performance level I'd be hard pressed to find an application for the board. An ESP32 (which has multi-core and RISC-V variants) would probably be more suitable, or an older generation RPi.

      Comment


      • #13
        what might be cool would be hacking on riscv cores that already exist in other commercially sold embedded products, for example an hdd condroller, or something like that. finding the programming port and hooking it up to platformio / openocd / vscode debugger. perhaps not sufficiet to run linux, but maybe as an mcu, flick some gpios on and off, and query some simple stuff over jtag or whatever? could that be worth bothering for just to have a play around? (and to avoid having to pay hundreds of dollars). because boy those riscv evaluation boards are still far too expensive

        Comment


        • #14
          another way might be if rpi made a switch from broadcom / arm to riscv platform. for example for the rpi5. then that would be able to get the cost down... only problem is: it would then end up being slower than the rpi4. lol

          Comment


          • #15
            Originally posted by dreamcat4 View Post
            what might be cool would be hacking on riscv cores that already exist in other commercially sold embedded products [...] because boy those riscv evaluation boards are still far too expensive
            The ESP32-C3 development boards which are RISC-V based can be had for under $4 quantity 1. Programmable with Arduino SDK or VSCode and PlatformIO.

            Comment


            • #16
              The problem I've encountered with all the RISC development boards I've considered is that they're under powered while simultaneously being astronomically expensive. So far they all seem to be geared towards existing developers who must absolutely have them, so they have no choice but to pay the price.

              If RISC wants to spread more rapidly it needs to provide low priced boards to entice those who might be considering it to experiment with it, otherwise it's going to continue to advance at close to stagnation levels.

              Comment


              • #17
                Originally posted by macemoneta View Post

                The ESP32-C3 development boards which are RISC-V based can be had for under $4 quantity 1. Programmable with Arduino SDK or VSCode and PlatformIO.
                ah cool man thanks! didnt realize. yeah that would be far easier

                Comment


                • #18
                  Originally posted by macemoneta View Post

                  The ESP32-C3 development boards which are RISC-V based can be had for under $4 quantity 1. Programmable with Arduino SDK or VSCode and PlatformIO.
                  And for people that aren't aware, that's the import price after shipping. They're 2 for $13 or 3 for $18 on Amazon if you're in a hurry. I'm waiting on an ESP32S and a USB-C breakout board for some stuff I'm wanting to work on.

                  Granted, I'm also waiting on a soldering gun from China to do that work, a Pinecil V2, so I'm going to be waiting for a bit.

                  Comment


                  • #19
                    Originally posted by ezst036 View Post
                    I hope this eventually leads toward standard RISC-V mATX, ITX, and eventually ATX boards with a Coreboot/etc OSS bios from the factory.
                    SiFive have been working toward this for some time. They've announced a microATX board called the HiFive Pro P550. (Edit: But I guess they don't support coreboot.)

                    Comment


                    • #20
                      Originally posted by Old Grouch View Post
                      It could be an interesting target for OpenWrt - with two GigE ports, one faces the WAN, and the other a local switch. Given the cpu specifications, I doubt it could route a full GigE of traffic, but it would be nice to get more RISC-V options for OpenWrt.
                      It would be interesting to know how the GbE ports are connected. Are they USB like on Raspberry Pi?

                      Comment

                      Working...
                      X