Intel Gags Customers from Publishing Performance Impact of Microcode Updates

Discussion in 'HardForum Tech News' started by DooKey, Aug 23, 2018.

  1. DooKey

    DooKey [H]ardness Supreme

    Messages:
    8,035
    Joined:
    Apr 25, 2001
    Just when you thought the optics for Intel couldn't get any worse than they are Intel updates their license terms for the latest microcode update and shoots themselves in the foot. The latest license agreement says you can't "publish or provide any Software benchmark or comparison test results." I really hoping this was an honest mistake and not deliberate. If deliberate, the person who made this decision needs to be fired. All I can do is SMDH.

    The company has hence updated the license terms governing the microcode update distribution to explicitly forbid its users from publishing comparative "before/after" performance numbers of patched processors.

    UPDATE: We reached out to Intel and this issue has seemingly been addressed. This is what Intel had to say to us.

    “We have simplified the Intel license to make it easier to distribute CPU microcode updates and posted the new version. As an active member of the open source community, we continue to welcome all feedback and thank the community.” – Intel spokesperson
     
    Last edited by a moderator: Aug 23, 2018
  2. readeh

    readeh Limp Gawd

    Messages:
    276
    Joined:
    Apr 1, 2013
    Oh snap! they dun goof.
    Not buying anything Intel till they fix their shit which probably won't be for a long time and then I'd likely go with AMD anyway.
     
  3. trick_m0nkey

    trick_m0nkey Moderator Staff Member

    Messages:
    3,997
    Joined:
    Oct 11, 2005
    This will definitely prevent people from doing benchmarks.
     
  4. Grimlaking

    Grimlaking 2[H]4U

    Messages:
    2,849
    Joined:
    May 9, 2006
    I really hope this is a mistake. But to be honest those rules for the distributors don't apply to end users I would think. Someone better at legaleese might want to give this a gander.
     
    undone likes this.
  5. undone

    undone n00b

    Messages:
    18
    Joined:
    Jun 29, 2018
    How does that work if I get my bios from a motherboard vendor?
     
  6. Basically it's a tacit admission that the patch is going to impact chip performance in a very real way...
     
  7. ChristianVirtual

    ChristianVirtual [H]ard DCOTM x3

    Messages:
    2,524
    Joined:
    Feb 23, 2013
    "last week I had this performance .... this week I had this performance ... can someone explain what this could cause by ?" -> Problem solved or legally disputable ?
     
    GDI Lord likes this.
  8. Diverge

    Diverge Gawd

    Messages:
    536
    Joined:
    Nov 22, 2005
    I'm not buying a new CPU until these issues are resolved without performance hits, and there is transparency. Intel's loss...
     
    Nukester likes this.
  9. Hagrid

    Hagrid [H]ardForum Junkie

    Messages:
    8,326
    Joined:
    Nov 23, 2006
    I guess the crapload of microcode and OS fixes is really tanking performance? Or maybe there is more bugs to come since it just seems non stop and they need to do a preemptive strike?
     
    Darth Kyrie and Vader1975 like this.
  10. mkrohn

    mkrohn 2[H]4U

    Messages:
    2,330
    Joined:
    Apr 30, 2012
    AMD isn't without blame/dirt. the AMD/Intel argument is very similar to the Dem/Rep line in politics. The Reds are the clear evil in your face right now but the blues have plenty of evil of their own. We have always needed a viable third option in both instances.
     
    tetris42 and Aronjlove like this.
  11. Jim Kim

    Jim Kim 2[H]4U

    Messages:
    3,434
    Joined:
    May 24, 2012
    Wait till Nvidia add's GPP and the new 5 Year NDA to their drivers.
    Want the latest driver, sign here.
     
    Chupachup, Darth Kyrie, Madoc and 6 others like this.
  12. shpankey

    shpankey Limp Gawd

    Messages:
    157
    Joined:
    May 27, 2005
  13. notarat

    notarat [H]ard|Gawd

    Messages:
    1,697
    Joined:
    Mar 28, 2010
    You will not, and will not allow any third party to (i) use, copy, distribute, sell or offer to sell the Software or associated documentation; (ii) modify, adapt, enhance, disassemble, decompile, reverse engineer, change or create derivative works from the Software except and only to the extent as specifically required by mandatory applicable laws or any applicable third party license terms accompanying the Software; (iii) use or make the Software available for the use or benefit of third parties; or (iv) use the Software on Your products other than those that include the Intel hardware product(s), platform(s), or software identified in the Software; or (v) publish or provide any Software benchmark or comparison test results."

    Intel is 1st party
    Mobo Mfgr is 2nd party
    You are 3rd party (so it applies to you, as well.)

    Do not update your Microcode until the license terms are changed or you are in violation of the license terms (if you bench and publish the results)are open to legal issue. It's that simple.

    Edited to clarify - (if you bench and publish the results)
     
    Darth Kyrie, Baenwort and undone like this.
  14. ecuador

    ecuador Limp Gawd

    Messages:
    206
    Joined:
    Dec 29, 2008
    I started a thread yesterday about this, as Kyle seemed to have missed it, but the shocking thing is that this is intended for end users (which is why Debian refuses to include the patch) and as written does not only apply to before/after tests, but any kind of benchmarking you do with your (?) Cpu!
     
    Darth Kyrie, Baenwort and readeh like this.
  15. readeh

    readeh Limp Gawd

    Messages:
    276
    Joined:
    Apr 1, 2013
    How can you throw this on AMD lol.. Right now Intel has 32 lawsuits going on right now because of their security issues and microcode/performance impact. From what I know AMD cpu's has had no impact or barely any from all this. And no I'm not defending AMD.. I'm running Intel cpu's in all my systems.
     
  16. GoodBoy

    GoodBoy [H]ard|Gawd

    Messages:
    1,401
    Joined:
    Nov 29, 2004
    The third party is not you. That would be if ASUS gave the microcode and documentation (the important part) to MSI or someone else.
    see section "(iii) use or make the Software available for the use or benefit of third parties"
    If the third party was the end user, then they would be restricted from providing bios updates.. there's no way that is the case.

    The "publish benchmarks" applies to the mobo manufacturer, not to review sites or end users.

    Seems like some new mobo performance reviews are warranted...

    From what I've read, this second round of microcode adds another 7 to 10% performance hit, but it wasn't clear if that was across the board or specific workloads like the first microcode updates..
     
    Baenwort likes this.
  17. ecuador

    ecuador Limp Gawd

    Messages:
    206
    Joined:
    Dec 29, 2008
    What are you talking about, Intel has been one of the shadiest companies, abusing their market share, "fixing" their compilers, paying off vendors to drop AMD's products when they were vastly supwrios (early 2000s). I mean they had to pay billions as penalties and even that was considered a slap on the wrist given the extend of their behavior.
     
    Darth Kyrie, tetris42, Madoc and 6 others like this.
  18. MaZa

    MaZa 2[H]4U

    Messages:
    2,707
    Joined:
    Sep 21, 2008
    What? If I understood correctly what this is about there is no way in hell this is legally enforceable. End user can benchmark his rig the way he damn pleases and Intel has no power over it at all. What can they do if someone publishes numbers about their own system? Sue them? "Rrrrright..." would say any judge and leave the room.

    *Edit* and if someone tries to pull the "you do not own it, only have a license to use it" I will virtually punch you... I OWN my rig, not Intel.

    *Edit2* I really must be understanding something wrong because otherwise this thing Intel is saying is so stupid.
     
    Last edited: Aug 23, 2018
  19. Hagrid

    Hagrid [H]ardForum Junkie

    Messages:
    8,326
    Joined:
    Nov 23, 2006
    Motherboard manufacturers do cpu benchmarks as well? I did not know this. Any links?
     
    Baenwort likes this.
  20. MaZa

    MaZa 2[H]4U

    Messages:
    2,707
    Joined:
    Sep 21, 2008
    Skyblue and trick_m0nkey like this.
  21. Lifelite

    Lifelite 2[H]4U

    Messages:
    3,537
    Joined:
    Aug 26, 2015
    Fucking hell. Intel is just one screw up after another lately. This is what happens when you are practically the only choice for so long.

    Great timing on AMD's part to step up there game. I still think they need to go a bit further so they can get back to the point of having near as much performance as Intel Processors for a much lower price, like they used to be in the 2000s.
     
  22. Lifelite

    Lifelite 2[H]4U

    Messages:
    3,537
    Joined:
    Aug 26, 2015
  23. MaZa

    MaZa 2[H]4U

    Messages:
    2,707
    Joined:
    Sep 21, 2008
    Probably, though this is pretty much equivalent to testing if water is wet and fire hot. What else did they expect but a backlash?
     
  24. PaulP

    PaulP Gawd

    Messages:
    776
    Joined:
    Oct 31, 2016
    I'll bet their large (server farm) customers are asking what the performance hit is with these new patches. So don't think of "publishing" only in terms of blogs and review sites, but also any information provided to customers by the BIOS and/or motherboard manufacturers, like technical bulletins, etc.
     
    Baenwort likes this.
  25. Jaymzkerten

    Jaymzkerten Limp Gawd

    Messages:
    197
    Joined:
    Jul 30, 2012
    Looking at doing a SFF gaming build, guess all this pretty much cements that I should go with AMD to power the thing.
     
    Darth Kyrie likes this.
  26. Hagrid

    Hagrid [H]ardForum Junkie

    Messages:
    8,326
    Joined:
    Nov 23, 2006
    Quote = Intel's latest round of microcode patches addresses the L1 Terminal Fault. The company has already posted its performance measurements with its new patches in the Windows operating system. Intel's results show little performance impact in many scenarios, but up to a 31 percent performance loss in some virtual environments that house untrusted guest operating systems.

    They say it shows a little, so it is probably cherry picked stuff. 31%?! Wow. Too many to remember, but wasn't there another that was about the same % in high I/O operations?
    Maybe you will need their special chilled water system to OC high enough to overcome all the fixes......
     
    Darth Kyrie and jnemesh like this.
  27. Hagrid

    Hagrid [H]ardForum Junkie

    Messages:
    8,326
    Joined:
    Nov 23, 2006
    Ahh, I see what you mean. Yeah, that does make sense.
     
  28. Lifelite

    Lifelite 2[H]4U

    Messages:
    3,537
    Joined:
    Aug 26, 2015
    Usually they know there will be a backlash, they are just measuring how much of one there'll be.

    There's also the thing where they do something like this to spark a huge outrage, then do something much quieter/lesser, which will slip by because people are focused on the big thing.

    They then say "lol jk" about the big thing and everyone forgets about that and the little thing too. See Bethesda and paid mods. There are paid mods in Skyrim now, btw.
     
    The Worst Player, Madoc and Nolan7689 like this.
  29. Disco_Stu_04

    Disco_Stu_04 Limp Gawd

    Messages:
    427
    Joined:
    Aug 18, 2003
    Can't publish bench's?

    Well now I really want to see them.

    Intel's own Streisand effect.
     
    Mr. Bluntman and EODetroit like this.
  30. GoodBoy

    GoodBoy [H]ard|Gawd

    Messages:
    1,401
    Joined:
    Nov 29, 2004
    Was dumb of them to try and restrict benchmark results, especially since they published their own:
    https://www.intel.com/content/www/us/en/architecture-and-technology/l1tf.html

    Majority of workloads are unaffected or show +-1%
    postgres database takes 18% performance hit.
    virtualization takes up to 31% hit.
    webserver workload took a 7% hit.
    There were a few others I didn't recognize what they were, that were about 10%.

    I doubt most of us will see any impact from these changes.

    What will suck is if this shit just keeps continuing, and 10 years worth of past CPU's can't be secured, or if doing so makes them perform poorly. Soon as that performance hit lowers them below AMD's performance, time to jump ship.
    AMD is affected by some of these vulnerabilities but not all. Pretty sure they haven't taken any significant performance hit due to these.
     
  31. ChadD

    ChadD 2[H]4U

    Messages:
    4,068
    Joined:
    Feb 8, 2016
    Transparency is possible.

    No performance hit is not.

    They cheated security design for speed... and now it will bite them.

    New chips fixed in hardware won't take as big a hit... but they will still take one compared to the same chip without the fixes. Main issue for Intel is they can't get their next gen fab process up and running. That would solve their issues. They could fix their flawed design which would incur some performance loss and hide it in the gains from the shrink.

    Everything for Intel is now riding on their fab division.
     
    Baenwort likes this.
  32. DeathFromBelow

    DeathFromBelow [H]ardness Supreme

    Messages:
    7,176
    Joined:
    Jul 15, 2005
    If any of you Intel guys need an upgrade I have a few Bulldozer CPUs left. ;)
     
    Darth Kyrie, Verado, jeffj7 and 4 others like this.
  33. ChadD

    ChadD 2[H]4U

    Messages:
    4,068
    Joined:
    Feb 8, 2016
    The benchmarks I would love to see would be Bulldozer vs the then current Intel core chips with all the meltdown / specter / Foreshadow fixes. lol

    Something tells me AMDs non cheating chips wouldn't have looked that bad. I know I still have a core duo system around and the Meltdown fixes just tank performance on that chip.
     
  34. p.i.m.p

    p.i.m.p Gawd

    Messages:
    583
    Joined:
    Jul 18, 2004
    than i urge every intel user to do so
     
  35. jardows

    jardows [H]ard|Gawd

    Messages:
    1,615
    Joined:
    Jun 10, 2015
    Well that sours me from the new 9000 series. I'm going to be suspicious of benchmark results. I trust the [H] review will be fair.
     
    Darth Kyrie, Baenwort and jnemesh like this.
  36. alxlwson

    alxlwson You Know Where I Live

    Messages:
    6,046
    Joined:
    Aug 25, 2013
    I really wonder when the class action is coming.
     
  37. Delicieuxz

    Delicieuxz Gawd

    Messages:
    959
    Joined:
    May 11, 2016
  38. Betaboy1983

    Betaboy1983 [H]ard|Gawd

    Messages:
    1,264
    Joined:
    Dec 31, 2008
    You don't even have to benchmark then.. It's already obvious the patch will wreck performance talking crap like that.
     
  39. FrgMstr

    FrgMstr Just Plain Mean Staff Member

    Messages:
    48,317
    Joined:
    May 18, 1997
    UPDATE: We reached out to Intel and this issue has seemingly been addressed. This is what Intel had to say to us.

    “We have simplified the Intel license to make it easier to distribute CPU microcode updates and posted the new version. As an active member of the open source community, we continue to welcome all feedback and thank the community.” – Intel spokesperson
     
  40. ThatsAgood1jay

    ThatsAgood1jay [H]ard|Gawd

    Messages:
    1,025
    Joined:
    Mar 21, 2006
    *Sniff*

    Ya'll smell that? it's desperation.

    Intel leadership needs some bloodletting, this is ridiculous.
     
    Darth Kyrie likes this.