1. 55
  1.  

  2. 33

    First they…

    Ah, you all know how it goes. Congratulations to the RISC-V guys on making another step on that particular ladder.

    1. 30

      yesterday afternoon, the community reply https://www.arm-basics.com/

      1. 3

        That was a nice reply :D

        1. 4

          Eh, it’s a pretty cheap shot, morally on the same level as the original page and much weaker in content. I hope it’s not representative of the larger RISC-V project.

          1. 13

            I’m not too impressed with the counter-FUD but I think it’s hilarious that the riscv-basics.com people didn’t think to register arm-basics.com while they were at it.

        2. 1

          Absolutely brilliant

        3. 23

          I think that ARM do not realizes what they just did.

          Besides stupid idea aka ‘Get the Facts’ from Microsoft now people start to acknowledge what RISC-V is and that its an alternative to ARM.

          Before ARM made that site people did not even knew RISC-V existed :)

          1. 9

            Before ARM made that site people did not even knew RISC-V existed :)

            I’m just one data-point and I’m more of a software rather than hardware person, so I don’t really matter, but yes. I had no idea about RISC-V before Matthew Garret tweeted about this page. Nice to see an open design. This would definitely be something to consider if I ever have to deal with hardware at this level.

            1. 4

              I’m a little new to RISC-V but I see a whole lot of very familiar names up on this wall: https://pbs.twimg.com/media/DgyJOMwX0AAeSgx.jpg:large

              So while it might not be as mainstream as ARM, my impression is that the industry knows about RISC-V and is watching it very carefully.

            2. 17

              This is great news for RISC-V, and I can’t wait to see the next steps.

              1. 15

                Open hardware is critical for user freedom. It is great to see that ARM thinks RISC-V is a viable alternative.

                1. 10

                  There’s a lot of discussion about how authentic or childish this is. Although I still can’t find it, I did read an article on a site for embedded/chip people from an ARM spokesperson quite a while ago saying a lot of the same. Essentially, ARM’s position is that its ecosystem of existing, field-proven IP and supporting tooling presents a huge advantage in terms of creating your next chip with lowest risk. You can move faster than with a new ISA like RISC-V. You will run into fewer troubles. More important, ARM will be around supporting its IP for a long time. Whereas, we get new bankruptcies or acqui-kills every year on competitors to ARM, GPU companies, FPGA companies, and so on. They look so good until they’re gone, mainly due to opponents’ ecosystems.

                  So, those arguments are legit when you’re a company deciding what to invest hundreds of thousands to tens of millions in that has to work when you need it to and be around later. ARM is a great choice for that with all their existing customers corroborating it. They also have best SoC’s for things like smartphones where processor is about the least, difficult feature to create. MIPS is a good option, too, if you want to go cheaper. It’s often used in home routers as an example. They have high-end stuff, too, but I don’t know how many use it.

                  And then there’s a crowd that might be able to do all their own IP implementations, work with unproven toolchains, take less competitive products now to invest in a freerer/better solution down the road, or take other risks. That crowd might benefit from RISC-V. Most companies are better off going with ARM or MIPS if aiming for optimal risk/reward ratio. Good news is that RISC-V ecosystem is improving all the time with big companies backing it. :)

                  1. 5

                    I think this is a pretty good summary of ARM’s position.

                    Still, it’s worth pointing out that they are a “fabless” design firm, whose capital is all in their IP. They only make money on licensing their designs, so an open standard like RISC-V presents a similar kind of existential threat to them as Linux did to proprietary server OSs in the late 90s and early 2000s. The smartphone trend has been very profitable for ARM, but could very well dry up in just a few years: it’s dominated by giants like Apple / Samsung / LG / HTC, all of whom can afford to invest in processor design. As the phone market expansion slows, it will become more competitive on margins, and those players who are able would love to shed those licensing fees.

                    Despite the significant technical merits of the clean-slate approach, there’s a long way to go before RISC-V is competitive with ARM cores on mobile performance. I imagine that RISC-V will start to encroach on their business mainly in the embedded space. We’re already seeing signs of that, for example from WD last year.

                    1. 2

                      They only make money on licensing their designs, so an open standard like RISC-V presents a similar kind of existential threat to them as Linux did to proprietary server OSs in the late 90s and early 2000s.

                      I think this got more true when a bunch of big companies that make hardware or silicon IP started backing RISC-V. Right now, it’s a non-threat that could become a major one as it picks up momentum. It could be existential later. I think that’s why ARM started doing DesignStart. Now press releases.

                      “ I imagine that RISC-V will start to encroach on their business mainly in the embedded space. We’re already seeing signs of that, for example from WD last year.”

                      That was my prediction. Except I thought they’d be taking away market share from 8-16-bit MCU’s and 32-64-bit MIPS more than ARM. The SoC’s in that space have a lot of peripheral and I/O support. Like I thought for Leon3, a straight-forward option might be to start designing that I.P. on inexpensive nodes to mix with a size/power-optimized, mini-RISC-V core on 28nm like SiFive’s. Idea being the better, free core becomes as cheap as an 8-16-bitter on older node with bundled stuff getting more competitive over time. Start with low end that needs few peripherals, get some volume deals on that, step up to one with a bit more, rinse, repeat. Simultaneously, go straight for 32-64-bit market for stuff like routers or automotive integrating with low-cost, 3rd-party IP. Compete more on better tooling than just ISA pricing with open replacements for that custom, proprietary stuff embedded people normally buy for five digits.

                      Anyway, that was just my speculative strategy for this stuff. I’m follow that industry from a distance. I can only have so much confidence in the proposal. I know peripheral’s, I/O, unit pricing, and toolchains matter a lot to them, though.

                  2. 7

                    Note that there is no proof that this actually a page built by ARM as far as I can tell, and the domain is registered with domain privacy on a different registrar than other ARM or SoftBank owned domains.

                    I’m not saying this is not an official ARM page, but it might just as well be some prankster trying to make ARM look bad by creating this foot-in-mouth campaign page which can’t not be compared to the old and eternally criticized Microsoft “Get the facts” campaign of yesteryear.

                    Edit: and it’s been taken down. Either ARM realized the negative PR this was causing, or it’s been taken down because of unauthorized use of the ARM trademarks, or the prankster’s Azure credit ran out?

                    1. 16

                      Note that there is no proof that this actually a page built by ARM […]

                      The infographic is hosted on ARM’s servers.

                      1. 6

                        Well that’s pretty conclusive, never mind me!

                    2. 6

                      down and it looks like archive.org doesn’t have the page :(

                      https://web.archive.org/web/*/riscv-basics.com

                      anyone have a copy?

                        1. 3

                          Lobsters’ built-in “cached” button worked for me just fine: https://archive.is/https%3A%2F%2Friscv-basics.com%2F

                          I wonder if they got embarrassed and took it down permanently.

                          1. 2

                            It works if you prepend “https://” to it: https://web.archive.org/web/*/https://riscv-basics.com

                            There’s actually quite a few snapshot over less than 48 hours, I guess there were lots of people who though this might get taken down pretty quick.

                          2. 3
                              1. 1

                                Advertising and press releases like this aren’t really something we want here on Lobsters, right? Even if it is kinda satisfying to root for the underdog.

                                1. 10

                                  Respectfully disagree. I think this is a very interesting story for those of us interested in free / open hardware.

                                2. 1

                                  NXDOMAIN for me.

                                    1. 1

                                      Sorry. This URL has been excluded from the Wayback Machine.

                                      Wow. I know Wayback Machine will honour robots.txt but I’d never seen a retroactive takedown before.