Jump to content
AIDA64 Discussion Forum

Search the Community

Showing results for tags 'benchmark'.



More search options

  • Search By Tags

    Type tags separated by commas.
  • Search By Author

Content Type


Forums

  • AIDA64
    • General Discussion
    • Brainstorming
    • Bug reports
    • Hardware monitoring
    • Benchmarking, system performance
    • Network audit, change tracking, SQL databases
    • Localization, language modules
    • 3rd party solutions
  • App Forum
    • Android Forum
    • iOS Forum
    • Sailfish Forum
    • Tizen Forum
    • Ubuntu Touch Forum
    • Windows Phone Forum

Find results in...

Find results that contain...


Date Created

  • Start

    End


Last Updated

  • Start

    End


Filter by number of...

Joined

  • Start

    End


Group


AIM


MSN


Website URL


ICQ


Yahoo


Jabber


Skype


Location


Interests

Found 9 results

  1. Due to the enermous number of requests we've been receiving since the Ryzen market launch, concerning various confusions about AIDA64 and Ryzen, hereby we post a clarification. AIDA64 is currently not 100% compatible with the recently unveiled AMD Ryzen high-performance x86 processors. It's because AIDA64 still has a few issues that we need to fix. However, in order to fix those issues, we first need to run a series of very long benchmark tests on Ryzen, and that -- among with the bug fixes themselves -- will take several days to complete. As for the bugs and limitations we so far discovered: UPDATE: We have fixed #3, #4, #5 and #7 in the latest AIDA64 v5.90.4200 stable update, which is now 100% compatible with AMD Ryzen processors: https://www.aida64.com/downloads/latesta64xe 1) A number of minor hardware detection issues were already fixed in the latest AIDA64 beta update. 2) The list of Turbo and XFR PStates are invisible on Ryzen, so it's not possible to properly enumerate or track them using a software. When a core of the AMD Ryzen processor goes into idle, the core will report the clockspeed of the P2 power state (e.g. 2200MHz on the Ryzen 7 1800X) and enter into the core-c1 (CC1) or core-c6 (CC6) sleep state. While the VID remains detectable in these states, the states are power gated and the true frequency is not known to the OS or monitoring utilities. As indicated from the “fine-grained Pstate” commentary released at Ryzen Tech Day, the AMD Ryzen’s processor true frequencies in these modes are significantly lower than reported via the “last known” P2 reading. AMD engineering tells us that V/f changes can be executed at 1ms intervals, indicating that the act of monitoring the states with the resolution necessary to accurately capture this behavior would also prevent cores from entering into the ultra low-power CC1 or CC6 states. 3) L1 cache bandwidth and latency scores, as well as memory bandwidth and latency scores are already accurately measured. 4) L2 cache and L3 cache scores indicate a lower performance than the peak performance of Ryzen. The scores AIDA64 measure are actually not incorrect, they just show the average performance of the L2 and L3 caches rather than the peak performance. It will of course be fixed soon. 5) Even though AIDA64 warns about a potential lack of optimization, the CPU and FPU benchmarks should be indicative of the full potential of Ryzen. We may be able to tweak e.g. the FPU Julia benchmark to squeeze even more performance out of Ryzen, but we don't expect the improvement to be substantial. 6) The CPU Hash benchmark provides an exceptionally great score on Ryzen due to the hardware accelerated SHA instructions capability of Ryzen. It's absolutely normal that hardware acceleration boosts CPU computing performance by such a margin. 7) AM4 motherboards are not yet supported by the latest AIDA64 stable build of v5.80.4000. Make sure to use the latest AIDA64 beta build to have accurate sensor measurements on ASRock, Asus, Biostar, Gigabyte, and MSI AM4 motherboards. We will post further updates to this topic as we progress with our bug fixing efforts on our Ryzen test systems.
  2. Is the benchmarking function also made for vm's or only for physical machines? thanks in advance.
  3. gppu test fail on 1950x CPU On second run CPU test is passing but GPU benchmark looks like values are only for 1 gtx1070 instead of 4 .
  4. Vlask

    GPU basic benchmarks

    i'd like to see simple GPU benchmark for testing basic values of graphic cards like: Pixel Fillrate (MPixel/s) - think can be found in some 3Dmarks Texel Fillrate (MTexel/s) - think can be found in some 3Dmarks Triangle/Vertices Rate (M/s) - don't know how gpu makers measure this, cant remember any benchmark for this one... Best will be based on Opengl 1.1 so also very old profi cards can be tested (3DLabs, Intergpraph and similar stuff) and test won't be tied to some version of DirectX. Fillrates can be measured by really old, outdated and no longer supported SPECglperf - https://www.spec.org/gwpg/pastissues/Feb2_02/opc.static/glperf.htm , but results have too many variants of results (see attached result file as warning how it should not look) and wont work on modern systems. This kind of benchmark would be usefull for diagnostics between theoretical values of cards and real scenario. Also for detecting of number texture units and ROPS on cards with unknown specs - getting often mixed info from various sources about these at mobile chips or OEM cards (GPU makers ignoring them mostly and won't release any closer info). FillRate.log
  5. The list of comparison systems in the benchmarks is very interesting. Here it is not known what kind of graphics card is installed in the systems. Is there a list of the installed GPUs?
  6. Hi everyone. I'd like to know if there is any chance to export results from "Manage User Results" i've run 10s of benchmarks on different configuration but i can only remove them from right click and there are no other options. is there something i missed or maybe 3rd party addon that allows exporting these results?!
  7. First of all, I want to thank you for building such a wonderful tool on Windows Phone. AIDA64 is one my favoriat hardware detection tools on Windows PC, I'm thrilled that it's now on Windows Phone as well. The version on my phone is 1.1.0.5 On System page, I hope the storage info can be more specific, I wonder if you can add these info as follows: which class (Class 10 or U-1)? what type (eMMC5.0)? what kind of chip (TLC or MLC)? Manufacture(Intel Samsung)? etc, On CPU (GPU)Page, As we all know, Windows Phones are all based on Qualcomm SOC so far, I think it's not too difficult to add a build-in benchmark to this page,as well as a ranking system, I always want to knows how my 2yr old flagship phone still holds its place among newer gen CPUs. Thus, it can help me decide to buy a new phone or holding on the current one. (Plus, an OC buttom,maybe?) On Network Page, The Current version has only 2 Blocks, CoreSystem & Wi-Fi it bothers me that I have to lookup the Qualcomm database to find out which network my phone support at least you can show me if my phone is GSM/WCDMA/FDD-LTE(better with Band details) enabled or not. On Battery Page, Charging cycles counts, Health indecator,Manufacture,etc I want to knows my battery is healthy or not. It should remind me to by an replacement if the battery poorly functions. Really hope to see more changes in the next build! Love you guys!
  8. I'd like to compare the results 1: of different settings of 1 PC and 2: from different PC's. Ad 1: I now generate a report (to be able to look up the settings). After completion I can see the benchmark results. I'd like to add all results to the user list, however I can only find a way to do it for each result seperately. How can I add all results in one step? Ad 2: How do I export/import the results of benchmarks ran on other PC's (my own reference results)? I also like to remove some reference results to reduce the amount of results. How?
  9. ( The picture of testing result can also be seen here:http://imgsrc.baidu.com/forum/w%3D580/sign=5d8024603ac79f3d8fe1e4388aa0cdbc/0eb30f2442a7d9335c2885ecac4bd11372f001a8.jpg ) This is the testing result of my 4-way AMD Opteron 6147 machine, but some people keep telling me this result is impossible, which makes me a little doubtful now! Could this be an error of Aida64, or it’s just a fact that some people refuse to believe? Looking forward to your reply! Best regards!
×