Announcement

Collapse
No announcement yet.

After Years Of Waiting, Hands On With The AMD ARM Board

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

  • #11
    Originally posted by M@yeulC View Post
    AMD could roll their own instruction set
    Uhm, no they could not. It takes a lot more than just designing the ISA and developing silicon to create a new ISA that can be actually used in desktops and servers. There is a lot of software to be written (Linux support, compiler support, etc.). A custom ARM design is far more likely anyway, or even RISC-V, though I think the latter is very unlikely to happen.

    Comment


    • #12
      “LeMaker has decided to restart production of these boards, even though it still has a communication issue between the processor and PCIe x16 slot. Those who don’t mind, can pre-order it for $299 dollars and because PCIe x16 slot isn’t working, users will get a free heatsink and fan. The PCIe x16 problem, hopefully will be fixed with a firmware sometime in the future.”

      This is why it was delayed in the first place. The x16 slot wouldn't work on the first run. Who would buy a board with a non-functioning slot? Hopefully, the firmware on the copy Phoronix has is current.

      Comment


      • #13
        There was a rumor that instead of shipping with the a1120, these are shipping with the a1150 or a1170, which is higher end but also more power hungry.

        You can see this easily if it has 8 cores rather than the 4 cores. Please confirm if this is true when testing.

        Thank you.

        Comment


        • #14
          Originally posted by chuckula View Post

          Actually it's quite clear: AMD has abandoned ARM. It could be fun to see what the board does in a review article but these things are basically white elephants that will never go anywhere.
          Lisa as the head of the company had to made many hard decisions to turn the company products and brand back on track. It was more cost effective to improve their current products than designing a new product & platform for an already very competitive market which regulates or limit some features for the sake of cross chip compatibility.

          But still to see the long time effect. 1. How AMD will be able to answer Intel when they catch up (and they will like they did on the Athlon days), and 2. The x86 platform is loosing ground to ARM devices. One of these days ARM will come to the desktop to stay. ARM is the future, do you like it or not.

          Comment


          • #15
            Originally posted by darkcoder View Post
            One of these days ARM will come to the desktop to stay. ARM is the future, do you like it or not.
            If we were still in an Intel only world, where Intel could demand the ludicrous margins they have been making per chip for the last few years, I'd absolutely agree.

            But now Zen is actually viable, and assuming it works in APUs as well as in large chips its going to actually cut Intel's margins a bit.

            And every margin cut eases pressure off ARM entering the desktop, because the primary motivator for ARM encroaching on Intel's turf is not power efficiency (it matters a lot less here) but the price / performance.

            I expect ARM boards to remain boutique, mostly because of how non-standard ARM environments are. You need custom firmware and drivers for every board, which is not nicely scalable to modular systems. If AMD can keep Intel in check computers will get cheaper again for once.

            Comment


            • #16
              Originally posted by starshipeleven View Post
              ^
              |
              This intel fanboi is scared.

              Really, for now AMD is just focusing on getting out of the sandpit which means good x86 parts, and will likely stay at it for at least another year.
              It's still too soon to state so vehemently that they have abandoned ARM for good.
              x2, the x86 business is AMD's bread and butter. They *have* to get that side of the house in order and profitable - before they move forward with other projects like ARM. Thankfully, the new Ryzen and Epyc look to be extremely competitive, so expect AMD's fortunes to improve in the coming years, and along with it, side projects like 64 bit ARM servers.

              Comment


              • #17
                Now that Phoronix has an AMD ARM board in its hands, all we have left is a Phytium T series 64 core, an Applied Micro X-Gene 3, a POWER 9, a Cavium Thunder X2 and the Qualcomm Centriq 2400.

                Then the world is complete.
                Last edited by edwaleni; 12 July 2017, 04:33 PM. Reason: I forget to mention the POWER 9

                Comment


                • #18
                  Lots of interesting comments here.

                  First off ive seen nothing from AMD indicating that ARM has been completely abandoned. What we have seen is a management team in save the company mode. This does not imply that there is no future for ARM at AMD, we wont know for sure until a public statement is made.

                  Second; the idea that power usage isnt important on the desktop or in the server room is non sense. If anything it becomes more and more important as a greater percentage of companies see a real impact in their electrical bill.

                  Third; ARMs greatest potential on the desktop lies with Apple. For one they have the high performance cores. Surprisingly i don't think ARMs future will be driven by their cores, but rather what is integrated around those cores. ARMs future will be shaped by the ability of companies to engineer their own SoC to bring new technologies like hardware acceleration of AI / ML technologies. Once these capabilities start to hit the market the old processor cores wont mean much to most users. We could start to see the transition in 2018 with the first enhanced SoC. In a way if you are focused on ARMs cores you are looking towards the past, the future is even bigger SoC with new tech integrated.

                  Fourth; related to the above is the fact that ARM cores especially if stripped of all 32 bit instruction set support, are tiny. This is a huge advantage in the world of SoC design as it frees up space for other hardware, saves power and allows for far more onboard memory (caches).

                  Comment


                  • #19
                    another ISA can't take over the desktop until it's as fast as x86 on compiled code and last gen games use enough cores to throw extra cores at them while emulating

                    Comment


                    • #20
                      Originally posted by wizard69 View Post
                      Surprisingly i don't think ARMs future will be driven by their cores, but rather what is integrated around those cores. ARMs future will be shaped by the ability of companies to engineer their own SoC to bring new technologies like hardware acceleration of AI / ML technologies.
                      Something that could be very interesting is an ARM SoC that integrates very fast networking and a lot of PCIe lanes with a reasonable number of cores on a single die. Those would be great for storage. Intel currently doesn't seem to have a successor to the Xeon-D which was wildly successful in taking market share from ARM, so if I were building ARM cores I'd be looking at this. Cavium ThunderX2 looks promising in that regard.

                      Comment

                      Working...
                      X