Announcement

Collapse
No announcement yet.

AMD @ Computex 2022 Talks Up Ryzen 7000 Series, Announces Mendocino Budget Laptop APUs

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

  • AMD @ Computex 2022 Talks Up Ryzen 7000 Series, Announces Mendocino Budget Laptop APUs

    Phoronix: AMD @ Computex 2022 Talks Up Ryzen 7000 Series, Announces Mendocino Budget Laptop APUs

    AMD CEO Lisa Su keynoted this morning for Computex 2022 where she talked up some of the company's processor plans for the rest of the year. The focal points were on the much anticipated Ryzen 7000 series desktop processors as well as announcing the "Mendocino" APUs that will be coming to affordable laptops later in the year...

    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
    Interesting, I wonder if the desktop 7000 series will feature a reasonable amount of CUs for the IGP as there was a rather wide disappointing disparity in the 6000 mobile series of either 6 or 12 CUs. DDR5 memory is still painfully expensive though.
    Last edited by Slartifartblast; 23 May 2022, 05:33 AM.

    Comment


    • #3
      Glad to see that we won't have to give up PCIe speed or L3 cache to have an iGPU anymore. The new AM5 platform is surely looking great. We'll see what's the commitment about longevity this time.

      Comment


      • #4
        Originally posted by Slartifartblast View Post
        Interesting, I wonder if the desktop 7000 series will feature a reasonable amount of CUs for the IGP as there was a rather wide disappointing disparity in the 6000 mobile series of either 6 or 12 CUs. DDR5 memory is still painfully expensive though.
        I heard that it would just be a basic GPU with 4 cu's. But that's not official.

        Comment


        • #5
          They say (at 23:46 in the presentation) that AM5 supports WiFi6E.
          Is that in the Chipset? Or a dedicated WiFi chip on the mainboard? Anything in the kernel yet?

          Comment


          • #6
            Generally looking very good, but I'm torn about the integrated GPU. How does switching between iGPU and external GPU is done on AMD side? I guess as a good part, that would allow easy external GPU VM pass-through.

            Comment


            • #7
              Originally posted by agurenko View Post
              Generally looking very good, but I'm torn about the integrated GPU. How does switching between iGPU and external GPU is done on AMD side? I guess as a good part, that would allow easy external GPU VM pass-through.
              No switching is required. Just plug your display into the GPU or the mobo, and boot it up.

              Comment


              • #8
                Originally posted by microcode View Post

                No switching is required. Just plug your display into the GPU or the mobo, and boot it up.
                No, I got that But if you're also using external GPU what's the tool to control which GPU is used?

                Comment


                • #9
                  A lot of people seem to be making a big deal out of the leaked 15% CB ST improvement, in terms of it being too low. Considering that CB is mostly "software rendering", it is a very good indicator of raw general purpose computation, they added some geometry acceleration recently, but that's about it.

                  So that's a 15% gain on the raw computational throughput, which was pretty strong in zen 3 to begin with. Intel managed to pull ahead using other techniques, as sustained compute workloads tend to favor zen 3 to intel 12 gen.

                  Zen 4 will add the first batch of CVML onboard acceleration, avx 512 support, double l2 cache, and in heavy scenarios - significant benefit from all the extra bandwidth from ddr5.

                  So while some see the 15% CB ST gain as amd leaking its "best case", I think what amd is leaking is raw computational improvement + some extra clocks. This benchmark doesn't leverage much of the new stuff that zen 4 gains over zen 3. I do expect to see higher gains in stuff like say geekbench - bursty and common (therefore commonly accelerated) purpose benchmarks, which will benefit from the new instructions, better latencies and them 5.5 ghz boost clocks.

                  Amd itself is saying "more than 15%", which to me sounds like 15% and up, but of course, it could be 15.1% max, it could be significantly higher, but IMO zen 4 addresses most of the things that intel managed to get a healthy perf boost from without having that strong of a compute pipeline as zen.

                  Also, really good news that amd is stepping up their mainstream IO. Decent solution with the two identical chipsets - saves on R&D&P and even manages to eliminate the need for some signal integrity components by keeping tracers shorter and more localized - signal integrity for PCIE5 gets tricky and costly. With ecc memory support, ryzen makes for great budget workstations and servers at a far better value than what similarly priced epyc or zeon systems can provide. AM5 brings consumer performance IO on par of enterprise systems from not that long ago, and I am really digging it.

                  Now if we could only shame intel into enabling ecc support as well.

                  Comment


                  • #10
                    Originally posted by agurenko View Post
                    Generally looking very good, but I'm torn about the integrated GPU. How does switching between iGPU and external GPU is done on AMD side? I guess as a good part, that would allow easy external GPU VM pass-through.
                    IGPU is on the IO die, where also the PCIE controller resides. So it has a very fast pathway to use the iGPU for say video decoding and dma the output to a discrete gpu. There will be your usual bios toggles, but I expect it to be automatic - it either detects a gpu on the slot market as "first" or primary, or falls back to the igpu. I don't expect that anyone would want to disable the igpu even with a discrete gpu, because it will likely contain accelerator circuitry that even if nowhere nearly as "fast" as a fat gpu, will have unparalleled latency as it basically resides on the hardware system agent level. That's amd's whole argumentation in favor of making the tiny igpu "a guarantee" across its entire product range, well that and counting each those as a gpu unit to improve its overall graphics market share, in which intel's been eclipsing both amd and nvidia due to having igpus in every consumer cpu.

                    Comment

                    Working...
                    X