Announcement

Collapse
No announcement yet.

Intel passing off bogus Xeon Server 'metrics'

Collapse
X
Collapse
  •  

  • Intel passing off bogus Xeon Server 'metrics'


    I am a bit of an old guy, I now realize. I've been at this for several decades.




     




    I have been following Charlie Demerjian for some time now since when he was at The Inquirer.




     




    Today he put up a tweet condemning some performance metrics put out by Ryan Shrout:









     




    If you read through that tweet you will find that at some point in responding to Shrout's tweet he says:




     




    "Ah screw it, I should just write an article."




     




    It's in this tweet comments:









    The article was written in five hours, and is very well written.




     




    In the article he shows how Intel and Shrout are now blatantly misleading with bogus numbers. Comparing $20k parts which use more power than one $7K part with cooked benchmark numbers which the FTC had told them that they weren't allowed to use as benchmarks without some form of disclosure.




     




    https://www.semiaccurate.com/2019/11...its-a-new-low/




     




    Quote





    The ‘blog’ in question is one of the newest ways Intel is trying to put out a message around the press. The company has all but destroyed the credibility of their messaging of late and rather than fix the problem, they are trying to work around it. Rather than being honest or explaining their dubious and at times outright misleading claims, they are attempting to go directly to the public who might not be aware of how things are being skewed.






     




    Intel has a long history of misleading consumers, the last one to get widespread coverage was that Principled Technologies debacle in 2018.




     




    Serve the Home also did a response:









    The above t.co link references the STH article that they did on the 5th. It can also been seen here:




    https://www.servethehome.com/intel-p...ng-benchmarks/




     




    In STH's article they dissect one result used in the Shrout graph called GROMACS. They show that the 'performance strategy' team used an non-Zen2 optimized version 2019.3, when an optimized version 2019.4 that had been RTP for a month before they put out the article.




    Quote





    One can only conclude that Intel’s “Performance at Intel” blog is not a reputable attempt to present factual information.






     




    If you want @Demerjian has been tweeting throughout the afternoon and this night with other things about this.




     




    Here's the Shrout team's response:









     




    https://medium.com/performance-at-in...e-b16c47b11a01




     




    And Charlie's response to their article:









    EDIT: this will probably go on and on and on. TH3 is out today so, expect some sniping and bad benchmarks the other way too.



    More...
      Posting comments is disabled.

    Categories

    Collapse

    Article Tags

    Collapse

    Latest Articles

    Collapse

    Working...
    X