Announcement

Collapse
No announcement yet.

Intel Publishes "X86-S" Specification For 64-bit Only Architecture

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

  • #91
    Originally posted by cl333r View Post
    How about Intel skips this intermediate step and goes straight to ARM or RISC-V.
    Why not simply open up the actual ISA that they use under their current translation layer?

    Comment


    • #92
      Intel lost the right to have any say in the future of microprocessor architecture when they tried to force Itanium on the globe, and then spent decades charging 4 to 6 times reasonable cost for pitiful two or four core microprocessors out of spite.

      In fact if not for AMD we'd be paying $4,000+ for a crappy four core Intel microprocessor at this very moment.

      Add their horrific corporate history of destroying any company or engineer who dare challenge their thievery and Intel has earned only one thing -

      The right to pound sand.

      Comment


      • #93
        Originally posted by Laughing1 View Post

        Libre hardware?
        Some options exist. Take a look at https://riscv.org/exchange/ and filter by hardware.

        There's also some OSH high performance core in the works:

        Comment


        • #94
          Originally posted by F.Ultra View Post

          Why not simply open up the actual ISA that they use under their current translation layer?
          Because that would lock them into the translator layer they're using, making it very difficult to re-architect their microarchitecture.

          A good ISA does not force microarchitectural details upon you.

          Comment


          • #95
            Originally posted by Min1123 View Post

            But, software is not on our side for this.
            If you want a vibrant ecosystem, RISC-V is where the momentum is.

            Comment


            • #96
              Originally posted by Min1123 View Post
              Still with what's here, it's garbage collection. The proposal specifies removal of basic modes for 16 and 32-bit, removal of some opcodes, and forcing things to use state registers instead of older flags, but I didn't see a single mention of consolidating or making a baseline of SIMD instructions. That means x87, MMX, MMXExt, SSE 1-4xx , AVX 1/2/512, and others all are still there, and all still have/need competing levels/hierarchies of overlapping functionality.
              Yeah, because they need this to be basically 100% compatible with things compiled for current x86_64. This isn't a 'new' architecture, it's re-baselining on the current one in a 100% compatible way. As long as someone can write a BIOS, bootloader, and kernel that starts at x86_64, this will work, and that's the point. Under the hardware hood, a bunch of stuff to support older modes will go away, and just like today, if you want to target AVX-whatever and newer, you can, but stuff compiled today will still run.

              Comment


              • #97
                Originally posted by ayumu View Post

                If you want a vibrant ecosystem, RISC-V is where the momentum is.
                I'm there too. I have Nezha, D1 of several platforms, HiFive Unmatched, and a VisionFive2. Software support is coming along, but so far the OpenSBI from these devices is far behind OpenFirmware from several decades ago, and far closer to the paradigm of ARM makers, where everyone has a different incompatible way of booting, Also, the PowerVR GPUs haven't been doing us any favors just yet either.

                Comment


                • #98
                  Originally posted by muncrief View Post
                  Intel lost the right to have any say in the future of microprocessor architecture when they tried to force Itanium on the globe, and then spent decades charging 4 to 6 times reasonable cost for pitiful two or four core microprocessors out of spite.

                  In fact if not for AMD we'd be paying $4,000+ for a crappy four core Intel microprocessor at this very moment.

                  Add their horrific corporate history of destroying any company or engineer who dare challenge their thievery and Intel has earned only one thing -

                  The right to pound sand.
                  Intel didn't increase prices of CPU through all the years they had dominance since core 2 duo. In fact AMD are experts in increasing prices, Literally first mainsteam CPU model since like ~~2000s consumer market 1031$ cpu is ... AMD-FX57, 18 years ago and if you adjust for inflation price that is 1600$. In fact Intel is the one that dropped prices over the years, Intel had few 999$ cpus like i7-975, etc. but just a little less clocked same version of CPU (960) was 562$. Now 13900k MSRP is joke comparing to those old prices and we are after massive inflation.

                  I literally don't f... get why people in this forum hate Intel. They literally are only company among big trio that doesn't increase prices for their products, in fact if you look at previous vs current gen they are the ones that lowered their prices, they provided you 1st class linux support, open source and tons of contributions ages before AMD and still are commited to it, and most of time they provided drivers for upcomming stuff months ahead so typical distro like ubuntu can support them out of box in day of premiere.

                  You literally have situation right now, that AMD has to drop MSRP of 7950X by 125$, because Intel instead of upping price to match AMD greedyness, they kept prices of 13th gen CPUs the way they did. Intel literally could set up MSRP of their 13th gen CPUs higher, and if i was shareholder at Intel i would probably vote for it, but they didn't do it to please shareholders, they did it to please you.

                  Like literally how people can support AMD. The moment AMD has minimal advantage (5000 ryzen CPUs vs Intel 11th gen) they instantly raised prices by 50-100$ across entire lineup.

                  PS: This post doesn't talk about server CPUs. But typical server CPU buyer is company and most companies are rich and greedy so even if Intel had too high pricing here, i don't give a F.

                  Like AMD - 5000/7000 series price increases, all the price increases with FX serie CPU, killing basic threadripper (i would consider it half consumer/half serious platform) the moment competition died from intel and massive markup with threadripper pro etc.

                  Nvidia also increases prices, 2000, 3000, 4000 serie GPUs are priced way higher then inflation would suggest some moments in long past also had smaller incidents.

                  Intel... I genuinly cannot list one such incidents. They had some other types incidents like bribing OEMs to not showcase PC with AMD cpus, but AMD also had some big incidents like RDNA 5600XT BIOS, recent motherboards frying AMD cpus from multiple manufacturers, etc...
                  Last edited by piotrj3; 20 May 2023, 07:22 PM.

                  Comment


                  • #99
                    Originally posted by Min1123 View Post

                    I'm there too. I have Nezha, D1 of several platforms, HiFive Unmatched, and a VisionFive2. Software support is coming along, but so far the OpenSBI from these devices is far behind OpenFirmware from several decades ago, and far closer to the paradigm of ARM makers, where everyone has a different incompatible way of booting, Also, the PowerVR GPUs haven't been doing us any favors just yet either.
                    I have a D1 board and the VF2, besides a bunch of RISC-V MCUs.

                    Reference the SBI spec. OpenSBI isn't meant to do much. I am familiar with sun ultrasparc fw; it's cool but it's outside the role of SBI.

                    u-boot SPL (before opensbi, its primary job is to set RAM up) and u-boot (after SBI) are imho great, and much more useful than than openfirmware ever was.

                    VisionFive2 now has UEFI (edk2). That provides drivers and filesystems, and a working UEFI shell. It also provides a runtime that the OS can use and/or shut down entirely (like SBI does, but much richer).

                    RISC-V is doing WAY better than ARM, because the effort in boot specifications was put in there quite early: Every relevant RISC-V implementation seems to be compliant.

                    The platform specification (OS-A specifically) does go much further, and unlike the IBM PC clones, it isn't an accidental platform carried forward from the 80s, but one that's designed and today.
                    Last edited by ayumu; 20 May 2023, 08:40 PM.

                    Comment


                    • Originally posted by skeevy420 View Post
                      As far as the Macintosh-Windows 7 holdout guy is concerned, if he hates Windows 10 and 11, which I totally get, then his only real choices are Free Unix or Non-Free Unix; Linux/BSD or Apple. If he has customers that he recommends software and hardware to he's probably be better off choosing Apple over anything else. That's what I'd pick. It isn't that I want to pick them, but they're Unix-based and professional enough that macOS has proper color reproduction for working with imaging software and printing. We gotta work with what's available and that's pretty much it. KDE, HDR, Wayland, and proper color reproduction has me very excited.
                      It's more a matter of weighing up various personal preferences and deciding which one is best for his personal use, given that he already uses Linux in various support roles and used to daily-drive macs in the 68k and PPC eras.

                      For his customers, I assume he's just going to continue supporting whatever they want to run. After all, he's getting paid to put up with Windows in that context.​

                      Comment

                      Working...
                      X