Announcement

Collapse
No announcement yet.

GeForce GTX 1080 Ti: Core i7 7700K vs. Ryzen 7 1800X Linux Gaming Performance

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

  • #41
    Keep in mind that ryzens inter core connect as well as cache bandwidth is tied to the memory frequency! All the reviewer's got like the fastest possible RAM which still runs fine with Zen's current memory controller!
    in the end, CPU performance might be limited by cache bandwidth, as the execution pipelines are starved of work... Memory clock frequency has a much bigger impact on the Zen platform than on the Intel platform..
    Normaly you won't run into memory bandwidth saturation in games (Arma 3 is a edgecase) but L3 Cache BW is saturated pretty quickly if core scheduling is bad...

    Comment


    • #42
      To be honest, all i see is bad driver optimization and poor software code in games. I think that kernel and BIOS/EFI is lesser problem here (still aprt of the problem). I think biggest problem is non-existant optimization in nevidia blob, and excilent optimization for "best gaming CPU" on market in last ~4 years (before anyone argue that 7700k is new CPU, it is Sandy Bridge architecture with arch. improvements and quite substantional frequency boost).

      Originally posted by Spacefish View Post
      Keep in mind that ryzens inter core connect as well as cache bandwidth is tied to the memory frequency! All the reviewer's got like the fastest possible RAM which still runs fine with Zen's current memory controller!
      in the end, CPU performance might be limited by cache bandwidth, as the execution pipelines are starved of work... Memory clock frequency has a much bigger impact on the Zen platform than on the Intel platform..
      Normaly you won't run into memory bandwidth saturation in games (Arma 3 is a edgecase) but L3 Cache BW is saturated pretty quickly if core scheduling is bad...
      That can be fixed, but read my comment above your quote, most of the legacy software (games) will probably not get too much gain. But that will change, "AAA" games developers will for one optimize games for new architecture, and for two, they will work with hardware manufacturers (as they always did) to push people to buy new stuff, so expect 4-cores CPU's to become limitation in next year or so for few selected games, and in ~2+ years from now for quitea bit of new titles.
      I'm back! With a new mic, but it's still crap... On topic, back during the HotChips conference last year, it should have been clear that Zen needed the Windo...
      Last edited by leipero; 13 March 2017, 05:09 AM.

      Comment


      • #43
        Originally posted by bug77 View Post

        Look at the single thread results here: http://www.anandtech.com/show/11170/...0x-and-1700/22
        Ryzen's IPC is comparable to Intel's. But given the clock speed difference, single core performance isn't. It's in the same ballpark, but lagging. Though, not by the margin Michael has found in many games...
        Unless the games rely on AVX256 somewhere critical.. Still seems odd. The GPU would do vectorizable code better.

        Comment


        • #44
          Originally posted by mlau View Post

          Right. 20% perf difference could be explained by the clock difference (i.e. 7700k max turbo is >20% higher than 1800x's), but the rest seems like result of nvidia driver differences between this and Michaels last Zen tests. Obviously there's a lot of optimizations to be had with the nvidia driver.

          @Michael: but you didn't disable the turbo on both systems?
          Which "nvidia driver differences"? Both processors are running the same driver(s).

          Comment


          • #45
            Intel Compiler / Math library's degrading Ryzen's performance.

            Comment


            • #46
              Originally posted by schmidtbag View Post
              @indepe
              Ryzen's single-core performance is just fine; it is directly comparable to Intel's. I don't understand where people are getting this from. Even if just looking at game benchmarks, most of the games Ryzen does poorly in are multi-threaded.
              I would indeed love to see this test with HT turned off and 6 of the cores turned off using cpuset or cgroups....
              There can be very bad locking mechanisms at play.

              Comment


              • #47
                I wonder whether anyone realizes how much this thread reads like a thread about Bulldozer:
                - it's a new architecture, the OS needs to catch up (check)
                - it's a scheduler issue (check)
                - software is optimized towards intel (check)
                - software is not modern enough to use the latest CPU (check)

                Obviously, Ryzen is not the flop Bulldozer was. But the discussion around it seems to have made no progress over the years.

                Comment


                • #48
                  Originally posted by bug77 View Post
                  I wonder whether anyone realizes how much this thread reads like a thread about Bulldozer:
                  - it's a new architecture, the OS needs to catch up (check)
                  - it's a scheduler issue (check)
                  - software is optimized towards intel (check)
                  - software is not modern enough to use the latest CPU (check)

                  Obviously, Ryzen is not the flop Bulldozer was. But the discussion around it seems to have made no progress over the years.
                  Thats because no serious gamer bought Bulldozer after the reviews. If software would have adapted as they would have for Intel, Bulldozer would stay in a much better light (even today) - but nobody gave a dime about it.
                  With Ryzen, a lot of people give a dime and they demand improvements or it might have an effect on the sales of a game - and this is where software companies start to give a dime too. On top of that, AMD has released Dev-Kits and is in contact with the game studios, as well as M$ (finally) heard the shot and starts working on adaptations too. So yeah, work is being done or at least the demand for it is there. Lets have a tea and look at the results in 3 months.

                  Comment


                  • #49
                    bug77
                    But bulldozer was never a flop. Sandy Bridge i5's beated FX on regular basis in games till recently, now FX takes the cake, by large margin in any properly multithreaded game. Ryzen is nothing to be compared to bulldozer architecture..., first, bulldozer had 50-60% lower IPC compared to Sandy Bridge, Ryzen have on pair IPC with Kaby Lake. Your comparison is unjust and invalid.

                    I could write something I'm saying for years now..., but i will wait till BIOS and software fixes come out. But i can predict what will happen, and remmember my words, Ryzen will still be behind Intel CPU's at same frequencuies in games, even tho it beats them in any other task. And if that happens (and I'm 99% sure it will), i will tell you my opinion why (tho, you can assume).

                    Comment


                    • #50
                      In Bulldozer's defense people were actually right about it being under-utilized with most applications, specially games, which back then were still relying on one or two threads for most of the work. There may not have been all that many of them, but games using Mantle did show a pretty good size boost in performance on Bulldozer compared to their Intel equivalents meaning that games at large adopting low level APIs like Vulkan and DX12 could be a pretty big boon for AMD and highly multithreaded parts in general. We're only now seeing games properly take advantage of four or more threads and if you look at benches of properly multithreaded games like the Hitman and Doom reboots you'll see Bulldozer parts performing a lot better than what they do in other games.

                      So Ryzen's less than stellar performance in some areas is partially the same problem that plagued Bulldozer. However there's still plenty of bugs and quick-and-dirty implementations to be worked out of BIOS:es so performance could improve quite a lot without application developers having to do anything. Let's also not forget that most of the roadblocks for proper multithreading have since been cleared with multithreading features in newer versions of languages getting widespread adoption and compiler support along with low level graphics APIs finally becoming a hot ticket item.
                      Last edited by L_A_G; 13 March 2017, 08:04 AM.

                      Comment

                      Working...
                      X