Announcement

Collapse
No announcement yet.

AMD Threadripper 1950X Linux Benchmarks

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

  • AMD Threadripper 1950X Linux Benchmarks

    Phoronix: AMD Threadripper 1950X Linux Benchmarks

    Last week I was able to finally get my hands on a Threadripper 1950X system thanks to AMD for being able to deliver some Linux tests from this high-end desktop platform. The Threadripper 1950X as a reminder is a 16-core processor with 32 threads via SMT, 3.4GHz base frequency, 4.0GHz boost frequency, quad-channel DDR4 support, and support for 64 PCI-E lanes. Threadripper sits between the Ryzen 7 desktop processors and the AMD EPYC server/workstation processors, which are still soon to be tested at Phoronix. The Ryzen Threadripper 1950X will set you back $999 USD, but compared to the Core i9 7900X at the same price, has six more cores / 12 threads and a slightly higher base clock frequency of 3.4GHz vs. 3.3GHz but a lower boost frequency of 4.0GHz vs. 4.3GHz.

    http://www.phoronix.com/vr.php?view=25106

  • #2
    Thanks for the numbers, Michael.
    But I'm sometimes astonished that some benchmarks... well, don't they scale well on multicore? I'd say we can compare RyZen 7 and Threadripper like apple to apple in a benchmark. They share the same architecture and have similar clocking. But the big Threadripper has twice the cores, still in some benchmarks Ryzen 7 1800X is on top.
    Or is there somwhere else an obstacle in Threadripper (die internal communication between blocks? memory latencies?)?
    Looks strange to me.
    Stop TCPA, stupid software patents and corrupt politicians!

    Comment


    • #3
      While Intel certainly leads in the somewhat ambiguous IPC category, AMDs IPC is not that far off. You have to keep in mind that the hardly threaded benchmarks also cause Intel CPUs to clock somewhat higher and most importantly reach the target clock rate more quickly. So yes, the i9 is better for mixed or single threaded workloads but it isn't a huge difference in IPC but rather better fabrication processes allowing higher frequencies and better state management.

      Comment


      • #4
        Near the end the Sci-Kit benchmark has shown radically different performance levels based on the distro you are using as Phoronix showed in its own testing:

        http://www.phoronix.com/scan.php?pag...-distros&num=5

        Just a point that Ubuntu 17.04 is probably tuned more for Haswell and as we have seen, Ryzen/Threadripper tend to like Haswell optimizations. Skylake X being a new architecture tends to benefit from additional tuning.

        Comment


        • #5
          Already Sold ... ordering one...

          So here are the parts I have decided to order:

          1. PSU --> EVGA SuperNOVA T2 80 Plus Titanium Netzteil 1000 Watt
          2. Motherboard --> ASUS ROG ZENITH EXTREME
          3. CPU --> AMD Ryzen Threadripper 1950X Socket TR4
          4. Cpu cooler --> Noctua NH-U14S TR4-SP3
          5. RAM --> G.Skill Trident Z 32GB DDR4 3600 CL16 (4x8GB) 32GTZSW
          6. Graphic Card --> ZOTAC GeForce GTX 1080Ti AMP! Extreme 11GB
          7. Hard Disk --> Samsung SSD 960 Pro M.2 512GB x2 ( RAID)

          Any suggestions?

          Comment


          • #6
            Originally posted by Adarion View Post
            Thanks for the numbers, Michael.
            But I'm sometimes astonished that some benchmarks... well, don't they scale well on multicore? I'd say we can compare RyZen 7 and Threadripper like apple to apple in a benchmark. They share the same architecture and have similar clocking. But the big Threadripper has twice the cores, still in some benchmarks Ryzen 7 1800X is on top.
            Or is there somwhere else an obstacle in Threadripper (die internal communication between blocks? memory latencies?)?
            Looks strange to me.
            Just the benchmarks not scaling properly. Ryzen has the same intercommunication stuff.

            Comment


            • #7
              Instead of compiling the linux kernel, could you try compiling chromium? There was some odd results on Anandtech where it was slower than the 4c/8t Intel chips, but that makes very little sense, so it would be interesting to see how it performs whening compiling that big a project on Linux.

              Comment


              • #8
                Originally posted by carewolf View Post
                Instead of compiling the linux kernel, could you try compiling chromium? There was some odd results on Anandtech where it was slower than the 4c/8t Intel chips, but that makes very little sense, so it would be interesting to see how it performs whening compiling that big a project on Linux.
                Unfortunately I have no test profile for compiling Chromium unless someone can supply one or some comprehensive script where it would be easy to adapt for PTS.
                Michael Larabel
                http://www.michaellarabel.com/

                Comment


                • #9
                  And where are the promised SWR/LLVMpipe tests?
                  Originally posted by phoronix
                  I'll be back with more SWR/LLVMpipe tests soon, especially if getting the Skylake-X back-end working for SWR as well as seeing how these CPU-based renderers compare on AMD Eypc / Threadripper platforms for curiosity sake.

                  Comment


                  • #10
                    Originally posted by chithanh View Post
                    And where are the promised SWR/LLVMpipe tests?
                    Saved for their own article coming up.
                    Michael Larabel
                    http://www.michaellarabel.com/

                    Comment

                    Working...
                    X