Announcement

Collapse
No announcement yet.

AMD FX-8320E Performance On Linux

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

  • AMD FX-8320E Performance On Linux

    Phoronix: AMD FX-8320E Performance On Linux

    Back in September AMD announced new FX CPUs that included the FX-8370, FX-8370E, and FX-8320E. Back then we reviewed the FX-8370/FX-8370E CPUs under Linux but at the time didn't have our hands on the more affordable FX-8320E processor. In December AMD sent over the FX-8320E and so for the past few weeks I've been happily using this new Vishera CPU.

    http://www.phoronix.com/vr.php?view=21420

  • #2
    As i know for usual around -18% gaming performance (+ stuttering) when cpufreq ondemand is used instead of performance, i will virtualy add it to this test .
    Last edited by dungeon; 01-26-2015, 11:04 AM.

    Comment


    • #3
      One thing to notice between AMD vs Intel chips is 99th percentile frame time. TechReport is one of the few websites that actually perform this benchmark. Despite Intel chips having higher average frame rate, the FX 8 cores maintain a much smoother experience. The exception is the i7 which beats the FX's. The AMD APU's are just garbage all together. Why anyone would buy them for gaming is beyond me.


      http://techreport.com/review/26977/i...sor-reviewed/6

      Comment


      • #4
        ^^ More or less spot on. Intel i3 CPUs are more powerful on average, though their frame times aren't consistent at all, due to being, you know, dual cores. In that regard, no one other then Intel benefits from AMD Freesync tech.

        Hence why i5's are the p/p kings.

        Comment


        • #5
          Originally posted by gamerk2 View Post
          ^^ More or less spot on. Intel i3 CPUs are more powerful on average, though their frame times aren't consistent at all, due to being, you know, dual cores. In that regard, no one other then Intel benefits from AMD Freesync tech.

          Hence why i5's are the p/p kings.
          Yeah, it's sad too. AMD definitely has the better architecture. They screwed up when they let their marketing department call BD an 8 core. It's not. If you consider architecture to be the design of the platform, and the platform to be what's needed for the architecture to function, then Bulldozer and Vishera are 4 cores. As 4 cores they're pretty good. As 8 cores they are mediocre at best. AMD should never have let their marketing department call modules dual cores. They aren't.

          EDIT: What they -need- is one more integer unit per pipeline, with one more issue port per pipeline, and enough decoder bandwidth to keep it all busy. It adds complexity, but they could theoretically improve per thread performance by as much as 11% if they did. That's the entire reason why Thunderbird K7 was so awesome.
          Last edited by duby229; 01-26-2015, 12:03 PM.

          Comment


          • #6
            EDIT: It's a shame to see AMD ignoring all the lessons they learned in the past.

            Comment


            • #7
              I hope now when Jim Keller is back, AMD will release something good this year, or early next year. My AM3+ system still runs good. But probably not for long.

              Comment


              • #8
                Originally posted by duby229 View Post
                Yeah, it's sad too. AMD definitely has the better architecture. They screwed up when they let their marketing department call BD an 8 core. It's not. If you consider architecture to be the design of the platform, and the platform to be what's needed for the architecture to function, then Bulldozer and Vishera are 4 cores. As 4 cores they're pretty good. As 8 cores they are mediocre at best. AMD should never have let their marketing department call modules dual cores. They aren't.

                EDIT: What they -need- is one more integer unit per pipeline, with one more issue port per pipeline, and enough decoder bandwidth to keep it all busy. It adds complexity, but they could theoretically improve per thread performance by as much as 11% if they did. That's the entire reason why Thunderbird K7 was so awesome.
                ?
                it's 8 integer/logic/balls cores but 4 floating point... "units"
                running a program that does not use floats on all cores gives you 8x performance (minus normal threading overhead)

                problem with today's cpus (other then memory access) is instruction decoding
                adding more integer paths would further complicate instruction decoding/ordering

                Comment


                • #9
                  Originally posted by xeekei View Post
                  I hope now when Jim Keller is back, AMD will release something good this year, or early next year. My AM3+ system still runs good. But probably not for long.
                  Yeah, He and Fred Weber made some pretty awesome products for AMD. Although the architecture Jim Keller is working on pisses me off a little bit. He's dropping CMT in his architecture in favor of SMT. It's probably going to have better power efficiency, but I doubt highly it'll perform well. AMD architectures have always had really high latency on pipeline flushes, and SMT is only going to make it worse. I think it's a terrible decision.

                  Comment


                  • #10
                    Originally posted by gens View Post
                    ?
                    it's 8 integer/logic/balls cores but 4 floating point... "units"
                    running a program that does not use floats on all cores gives you 8x performance (minus normal threading overhead)

                    problem with today's cpus (other then memory access) is instruction decoding
                    adding more integer paths would further complicate instruction decoding/ordering
                    AMD's CMT architecture is organized in what AMD calls "modules". They chose that name because they decoupled the integer pipeline from the floating point pipeline. In doing so it allowed them to add as many integer pipelines as they wanted. In Bulldozer and Vishera there are two integer pipelines and one floating point pipeline. Each integer pipeline has two full function integer units, with two issue ports assigned to them. This configuration is called CMT "Concurrent multi threading". The decoder is shared across all three pipelines. Concerning just the integer pipelines, 89% of the time x86 code will decode into two or less instructions. The other 11% of the time it will decode into more than that. Right now that 11% has no choice but to wait for the next free cycle.

                    Going back in time, what made Thunderbird K7 so great was that it had 3 full function integer units, each with it's own dedicated issue port.

                    If AMD had done the same thing here, Bulldozer and Vishera would have been much more competitive with Intel's architecture. Which has three full function integer units, plus one part function integer unit and 4 issue ports.

                    AMD's architecture can only issue 2 instructions per pipeline, Intel's can issue up to 4. Although 4 isn't very likely, 3 does happen often enough to make a huge difference.
                    Last edited by duby229; 01-26-2015, 12:38 PM.

                    Comment

                    Working...
                    X