New CPU-Z Upgrade Lowers Ryzen performance

Discussion in 'Frontpage news' started by Hilbert Hagedoorn, May 4, 2017.

  1. Hilbert Hagedoorn

    Hilbert Hagedoorn Don Vito Corleone Staff Member

    Messages:
    41,043
    Likes Received:
    9,324
    GPU:
    AMD | NVIDIA
  2. Undying

    Undying Ancient Guru

    Messages:
    15,715
    Likes Received:
    4,741
    GPU:
    Aorus RX580 XTR 8GB
    Overall AMD CPU's scores are lowered, not only Ryzen. Some user already posted this :

    [​IMG]
     
  3. ddelamare

    ddelamare Master Guru

    Messages:
    221
    Likes Received:
    3
    GPU:
    Inno3D GTX 1070 X4
    So whats that, an average of 30% or so less performance core v core? Seems a bit heavy handed.
     
  4. Kaarme

    Kaarme Ancient Guru

    Messages:
    2,398
    Likes Received:
    1,022
    GPU:
    Sapphire 390
    I don't belong to the tin-foil hat brigade, but this sounds quite weird. Why would the Ryzen architecture process something efficiently if it was meaningless? I'm quite sure there are things the Intel core architecture is also good at. In fact, many current game engines prove that, making the core architecture still the best for gamers today.

    This is the same as saying that since a Porsche engine allows faster acceleration than a Ford engine, the acceleration factor must be disregarded if those two cars are ever compared. Sorry, CPU-Z developer, but I won't buy this explanation, which sounds totally random.
     

  5. GeniusPr0

    GeniusPr0 Maha Guru

    Messages:
    1,306
    Likes Received:
    28
    GPU:
    RX 6800 XT Black Ed
    There's no point having a benchmark that shows off what one processor arc can do in a 1 in a trillion situation against others that can't.
     
  6. ddelamare

    ddelamare Master Guru

    Messages:
    221
    Likes Received:
    3
    GPU:
    Inno3D GTX 1070 X4
    I just checked a recent validation from an x1800 @3.8Ghz, i guess the 6700k is running at stock. It still looks fine to me, nothing to worry about:
    [​IMG]
     
  7. gianluca

    gianluca Active Member

    Messages:
    56
    Likes Received:
    3
    GPU:
    Sapphire Nitro+ RX480 OC
    The benchmark seems very different compared to the previous version, the scores are way lower than the ones I had in the past.
    Right now my cpu is working at 3.5 GHz base clock with a 4.6 GHz turbo and I got 207 ST and 1348 MT.
    When Ryzen was released my cpu was working at 3.6/4.0 GHz and I had this result: 1050 ST and 6680 MT (in multithreading uses only the base clock when turbo is enabled).
    This is BS. Now I want to see how things changed for the Intel cpus.
     
    Last edited: May 4, 2017
  8. Silva

    Silva Ancient Guru

    Messages:
    1,514
    Likes Received:
    663
    GPU:
    Asus RX560 4G
    All scores dropped and have to be retested, my 2500k lost more than 1000 points in Single Thread.
    What's more important is knowing if this is a fair and real way of measuring performance across different architectures.

    Having unbiased real world benchmark examples is important. No one uses their computer to calculate Pi or something like that.
    PS: looking at those numbers again, something is wrong: AMD dropped more than Intel...wth?

    Those are the old values.
    And for god sake, use spoilers for big images.

    They argued the lower numbers are for comparison sake, as more cpu's with multiple cores will come out and would be difficult to read the high numbers.
    It's fine by me, but everything needs to be benchmarked again...
     
    Last edited: May 4, 2017
  9. DLD

    DLD Master Guru

    Messages:
    886
    Likes Received:
    71
    GPU:
    MSI GTX1060 6GB
    ''The new benchmark uses a new algorithm, and its scores can not be compared with the previous version.''

    I don't see anything weird here.
     
  10. Cplifj

    Cplifj Active Member

    Messages:
    91
    Likes Received:
    22
    GPU:
    290X
    thwarted

    so they found out ryzen was actually very good at something and they just had to obscure that fact in favor of intel.

    I hope intel payed well enough to manipulate the markets yet again.

    There, someone had to say it. Because most of us allready know it or everything that happens just points in that direction allways.

    Everybody knows marketing is lies , but they keep selling em despite.
     

  11. Ryu5uzaku

    Ryu5uzaku Ancient Guru

    Messages:
    7,056
    Likes Received:
    270
    GPU:
    6800 XT
    Yup looks fine to me. If that 1800x was 4ghz it would still be at top?
     
  12. GeniusPr0

    GeniusPr0 Maha Guru

    Messages:
    1,306
    Likes Received:
    28
    GPU:
    RX 6800 XT Black Ed
    What don't you get? AMD results were inflated, of course they dropped more.
     
  13. DARKSF

    DARKSF Active Member

    Messages:
    55
    Likes Received:
    13
  14. chronek

    chronek Member Guru

    Messages:
    184
    Likes Received:
    3
    GPU:
    Geforce 980 GTX 4GB gddr5
    in short - they created benchmark before and they did not think that any procesor could perform better than intel, so when ryzen show up they had to change benchmark so it could favor intel... btw any program compiled with ms visual c++ is optimized for intel...
     
  15. baasgene

    baasgene Active Member

    Messages:
    57
    Likes Received:
    10
    GPU:
    Asus 3080 Tuf OC
    mmm maybe most cpu benchmarks needs to be updated then, maybe it will mimic what we see in games on ryzen.
     

  16. Kaarme

    Kaarme Ancient Guru

    Messages:
    2,398
    Likes Received:
    1,022
    GPU:
    Sapphire 390
    I hope the CPU-Z developer has bought a few lottery tickets if he has a unique 1 in a trillion luck.

    A more plausible explanation is that there's real value in processing it faster and thus the architecture does it. The CPU-Z developer simply didn't like it for his own reasons, so he used his natural privilege as the tool developer to change it to something else, to serve his own ends. It's his right, so it's okay, but he would be more respectable by not trying to invent fancy excuses for doing it.
     
  17. MorganX

    MorganX Member Guru

    Messages:
    132
    Likes Received:
    13
    GPU:
    AMD Ref 6800XT
    This is why I don't trust benchmarks and rely on personal real world experience. Kind of curious this comes after AMDs stock drop. Piling on AMD en vogue now? Anyway, I'm eagerly awaiting my 1700 and Biostar ITX from newegg. I have a feeling it's going to leave my 5820k x99 ITX in the dust. We'll see. If it does, maybe I should email the results to the makers of CPU-Z.

    Right now, benchmarks have no credibility with me.
     
  18. MorganX

    MorganX Member Guru

    Messages:
    132
    Likes Received:
    13
    GPU:
    AMD Ref 6800XT
    This type of behavior, IMO, renders CPU-Z, completely incredible.
     
  19. SC Ghost

    SC Ghost New Member

    Messages:
    1
    Likes Received:
    0
    GPU:
    HD 5850
    LoL. Remember when Nvidia called Oxide ? It was like: We're not good at async compute so async compute is totally irrelevant and disable it in your benchmarks please.
     
  20. GeniusPr0

    GeniusPr0 Maha Guru

    Messages:
    1,306
    Likes Received:
    28
    GPU:
    RX 6800 XT Black Ed
    core for core and clock for clock - almost 30% higher than Intel Skylake

    I still disagree. Even Lisa Su said they were aware it's about ~7% behind Skylake on average, for IPC. If this was the case, AMD would have been all over this special case of processing a specific algorithm, which is quite possibly useless in the real world, or is at least not proven to be useful.
     

Share This Page