Announcement

Collapse
No announcement yet.

Ryzen 9 3900X/3950X vs. Core i9 10900K In 380+ Benchmarks

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

  • #21
    Originally posted by atomsymbol

    Could you please rerun (at least the above benchmark) on the Ryzen 9 3950X with PBO (precision boost overdrive) enabled? If I remember correctly, some other benchmarking sites have reported quite a large difference between default and PBO with Ryzen 9 on highly parallel workloads (something like +25% better performance), though the power consumption might be quite high. If 3950X isn't limited by power in this case, then it is limited by memory/L3-cache bandwidth and/or by SSD bandwidth or by the build system (Makefiles). I wasn't able to find a default-vs-PBO 3950X result at https://openbenchmarking.org/
    GCC compilation time is really only interesting for GCC times itself... Due to GCC's multi-stage build process, it isn't as parallelized as other code compilation tests.
    Michael Larabel
    https://www.michaellarabel.com/

    Comment


    • #22
      Originally posted by Michael View Post

      Note it's due to the test repetition.... due to standard deviation being above 3.0%, it ran extra times. On the benchmark result you can see the actual average, this is the elapsed time.
      Oh sorry, I missed that the numbers on the right were individual runtimes.

      Comment


      • #23
        Polyhedron Fortran Benchmarks
        Benchmark: mp_prop_design
        Core i9 10900K: 8 Minutes, 12 Seconds 492s
        Ryzen 9 3900X: 52 Minutes, 6 Seconds 3126s
        Ryzen 9 3950X: 8 Minutes, 54 Seconds 534s
        Something went wrong here, no?

        Comment


        • #24
          Originally posted by reach3r View Post
          Was the 10900K tested with TDP adhering to specification or with unlimited boosting?
          IIRC Hardware Unboxed or GamersNexus mentioned that most vendors have the default on Z490 at unlimited power.
          Does AMD adhere to its specs?!

          AMD Ryzen 7 3700X is rated 65W, consumes 90W no matter how long a test runs.
          AMD Ryzen 7 3800X is rated 90W, consumes ~120W no matter how long a test runs.

          AMD fanboys have to just shut the f*ck up sometimes. Intel 10900K does adhere to its TDP rating which is exactly 125W:

          Comment


          • #25
            Funny how AMD fanboys bury Intel at every turn, yet the 10900K is often faster than 3950X despite the latter having a whopping 60% MOAR cores.

            And I won't repeat what I've already said on TPU.
            Last edited by birdie; 28 May 2020, 04:18 PM.

            Comment


            • #26
              Originally posted by birdie View Post

              Does AMD adhere to its specs?!

              AMD Ryzen 7 3700X is rated 65W, consumes 90W no matter how long a test runs.
              AMD Ryzen 7 3800X is rated 90W, consumes ~120W no matter how long a test runs.

              AMD fanboys have to just shut the f*ck up sometimes. Intel 10900K does adhere to its TDP rating which is exactly 125W:

              lol.
              dude, chill.

              I'm aware how much more power above nominal TDP current Intel and AMD chips pull during boost states and that this is within spec.

              What i am asking is whether the specs were adhered to or not. This is relevant because
              a) as mentioned some motherboards come with unlimited boosting as the default
              b) there is a very real difference in cooling 125W with occasional 250W peaks versus 250W sustained.

              Michael can you specify this about your tests?

              Comment


              • #27
                Originally posted by birdie View Post
                ...intel fangirl drivel...
                1: Unoptimized, single threaded legacy code that cannot properly utilize powerful CPUs

                2:

                3: Get a life

                Comment


                • #28
                  Originally posted by birdie View Post
                  Funny how AMD fanboys bury Intel at every turn, yet the 10900K is often faster than 3950X despite the latter having a whopping 60% MOAR cores.

                  And I won't repeat what I've already said on TPU.
                  Good on AMD for finally piling some competitive pressure on intel -- instead of lambasting AMD fanboys, has it ever occurred to you to thank AMD for making your favourite CPU + motherboard combos cheaper?

                  In either case, the consumer is the winner. I still think Intel has a shit segmentation strategy however, and will thus refrain from lining their pockets as long as AMD ships ECC and VT-d/AMD-Vi capable frequency unlocked CPUs.

                  That 3900XT July refresh is looking like a mighty tempting upgrade over my current 2700X.

                  Comment


                  • #29
                    Originally posted by birdie View Post

                    Does AMD adhere to its specs?!

                    AMD Ryzen 7 3700X is rated 65W, consumes 90W no matter how long a test runs.
                    AMD Ryzen 7 3800X is rated 90W, consumes ~120W no matter how long a test runs.

                    AMD fanboys have to just shut the f*ck up sometimes. Intel 10900K does adhere to its TDP rating which is exactly 125W:

                    Is that a VW emissions test graph ?

                    Comment


                    • #30
                      Originally posted by ddriver View Post

                      1: Unoptimized, single threaded legacy code that cannot properly utilize powerful CPUs

                      2:

                      3: Get a life
                      2. That's called PL1 and it is within thermal dissipation limits, asshole. Read the Anandtech review closely, because surely you haven't read a single line from it and just saw this useless screenshot. The fact that AMD is lying through their teeth about their CPUs TDP doesn't move you one bit it seems.

                      3. Beat it, as well.

                      Most of the code ever written cannot or is extremely expensive to be properly parallelized/optimized, asshole. Even most modern video codecs for fuck's sake, e.g. x265 which barely scales beyond 16 cores. Libaom (av1) barely uses more than two. For 99% of people out there 15 cores of your super duper Ryzen 9 3950X are worthless. You can have 1000 slow cores but if you have a task which fully saturates just one core, your additional 999 cores are worth literal crap. And again, absolute most applications run this way.

                      I'm not saying I'm against having more cores available for the average Joe. I'm against crapping on Intel at every turn because they don't have access to an advanced node like AMD does. Their uArchs are extremely competitive. This effing comparison pits an architecture from 2015 to a brand new AMD arch from 2019 for fuck's sake and Intel still beats the crap out of AMD in a lot of tests. In fact Intel is only losing in truly parallel benchmarks because again, surprise, 16 >> 10. By retarded AMD logic an ARM CPU with 64 cores is better than an AMD CPU with 16 cores. God, AMD fanboys (actually fanboys of almost all companies) literally have crap instead of gray matter.

                      Originally posted by Slartifartblast View Post

                      Is that a VW emissions test graph ?
                      Ha ha ha ha. So funny.

                      Originally posted by ermo View Post

                      Good on AMD for finally piling some competitive pressure on intel -- instead of lambasting AMD fanboys, has it ever occurred to you to thank AMD for making your favourite CPU + motherboard combos cheaper?

                      In either case, the consumer is the winner. I still think Intel has a shit segmentation strategy however, and will thus refrain from lining their pockets as long as AMD ships ECC and VT-d/AMD-Vi capable frequency unlocked CPUs.

                      That 3900XT July refresh is looking like a mighty tempting upgrade over my current 2700X.
                      I'm lambasting the idiots who are parroting lies and nonsense, and trust me, Intel fanboys are not better. Cheaper? CPUs, maybe. Motherboards? Are you effing kidding me? The X570 motherboards have been expensive as hell.

                      "The consumer is the winner" - I'm not entirely sure about that. I'm all for progress but motherboards have become ridiculously expensive recently. CPUs as well. GPUs as well. Remember what was just 20 years ago and cry inside. And inflation doesn't cover the price increases we've seen so far.
                      Last edited by birdie; 28 May 2020, 05:08 PM.

                      Comment

                      Working...
                      X