Has Intel fixed the Sandy Bridge S3 sleep bug ?

Discussion in 'Intel MoBos' started by n1tr0, Jul 1, 2011.

  1. n1tr0

    n1tr0 n00b

    Messages:
    27
    Joined:
    Jan 21, 2011
  2. notarat

    notarat [H]ard|Gawd

    Messages:
    1,780
    Joined:
    Mar 28, 2010
    I take it all sandy bridges are affected?

    Odd...

    I'm using an MSI P67A-GD65/2600k and it sleeps fine. longest I've let it sleep is from ~11pm to 9pm the next day.

    Same goes for my MSI H67/2100 combo, though it hasn't slept for as long yet (just built it a week ago as a replacement HTPC)
     
  3. you2

    you2 [H]ard|Gawd

    Messages:
    1,191
    Joined:
    Aug 10, 2004
    I have the P67A-GD65/2500K and it sleeps fine (now) but I had to tweak a large number of settings to get it to 'sleep' instead of 'bsoding'. It will still 'bsod' if it is running at 100% when sleep kicks in (i.e, if you run prime95 wtih sleep enabled).

    Try that on your system notarat and see if it doesn't bsod :)


     
  4. lgestle

    lgestle [H]ard|Gawd

    Messages:
    1,663
    Joined:
    Aug 11, 2004
    Gigabyte H67 i3 2100 no problems with sleep. Goes in sleep multiple times daily.
    MSI Z68-E55 i7 2600K@4.4ghz no problems with sleep either. In fact that is my default method as I have it WOL for multiple tasks.
     
    Last edited: Jul 2, 2011
  5. n1tr0

    n1tr0 n00b

    Messages:
    27
    Joined:
    Jan 21, 2011
    According to Asus' previous posts, the S3 sleep bug was a problem with Intel's microcode. The bug never affected "everyone" and there have been plenty of attempts to work around it like disabling PLL overvoltage. Sometimes the fixes have worked, sometimes they haven't. If you're not familiar with it, check the Asus P67 and Asus Z68 support threads.
    I'm just looking to see if it's actually "fixed".

    From back in January
    The thread goes on for pages and pages, months and months, but Gary/Asus has never come out to announce a fix on their end.
     
  6. Raja@ASUS

    Raja@ASUS ASUS Community Support

    Messages:
    2,390
    Joined:
    Apr 18, 2011
    There has been no fix from Intel for S3 resume with PLL OV enabled afaik. I know someone who spoke to ASRock R&D who said that the fix mentioned in that thread was to do with S3 resume on the 2133 memory ratio and not PLL OV as hinted at in that thread. Some boards also need a full AC cycle before the PLL OV setting is engaged after changing it in BIOS (something to bear in mind before running PLL OV experiments and saying that S3 resume works with it enabled).

    -Raja
     
  7. n1tr0

    n1tr0 n00b

    Messages:
    27
    Joined:
    Jan 21, 2011
    How about with the PLL OV disabled ? because personally I had 2 Asus P8P67 Deluxe boards (1 original, 1 B3) that I tested with multiple power supplies, cpu's, and ram and both had the S3 resume bug even with PLL OV disabled and following the rest of the advice on work-arounds.

    Of course at the time, Asus wasn't answering tech support requests on their own forum and was impossible to reach via phone.

    You have to admit, sleep mode is over a dozen years old and has been a pretty standard/useful function; Almost half a year after launch of the new platform and this still isn't fixed ?! :mad:
    Does everyone just wait on a tech handout from Intel these days ?
     
  8. Raja@ASUS

    Raja@ASUS ASUS Community Support

    Messages:
    2,390
    Joined:
    Apr 18, 2011
    WRT to PLL OV the only way is to wait for a handout from Intel - only they know what that MSR actually does and what it's start up requirements are for power good.

    -Raja
     
  9. DBRASCO

    DBRASCO [H]Lite

    Messages:
    65
    Joined:
    Jan 8, 2004
    with PLL OV off and a stable system that would sleep on 1702 bios.The only change I made was to updated to 1703 and now cant get a stable system or sleep to work....... I only updated the bios........ and flashing back to 1702 did nothing the boards broke .....can't over clock stable and no sleep w/out overclock like most I think it's ASUS and it's poor bios that is most to blame....I can't even use XMP profile after the flash the board just locks up now unless on auto....

    Brasco
     
  10. Eldata

    Eldata Limp Gawd

    Messages:
    366
    Joined:
    May 10, 2011
    Did you clear CMOS before flashing back?
     
  11. Eldata

    Eldata Limp Gawd

    Messages:
    366
    Joined:
    May 10, 2011
    For what it's worth, I have a P8P67 Deluxe that has no sleep issues with PLL OV disabled. I suspect that's the case with many, if not most, P67 boards. But you raise a good question...."Is Intel to blame for all P67 sleep issues".

    According to Raja, Asus has been working on custom BIOS code changes that may address the issue and they are looking for end users with sleep issues to test these changes. It's not clear whether there were any takers or where this approach stands at the moment.

    My view is that if Asus was really serious about this then it wouldn't be too difficult to get their hands on one of the many "faulty" boards and do the required testing themselves. In fact, I've seen at least one individual offer such a board to Raja. So while Asus points a finger at Intel, pointing a finger at Asus seems just as appropriate IMO.

    Raja, would you like to respond further?
     
  12. n1tr0

    n1tr0 n00b

    Messages:
    27
    Joined:
    Jan 21, 2011
    I'd avoid updating the bios/uefi then, I had sleep working on the occasional beta bios, only to have it broken again with an update and reverting back couldn't get it to work again. It really makes the Asus boards seem incredibly flakey, however I understand other brands are having similar issues, it's just that I only bought Asus.

    I still suspect it's a hardware/reference design flaw, not a driver issue, because many systems, including my previous 2 systems, will wake up from sleep perfectly fine if it's done within a short time period (under 2 hours), but sitting for a longer time (~4+ hours) causes a host of hardware reinitialization faults. Anyone with the buy and a port 80 LED diagnostic display (like the p67 deluxe boards) knows exactly what I'm talking about. It's as if the capacitors are discharging over time and the board has insufficient voltage to wake all the components at the same time. When you buy a board boasting over a dozen power phases and you're running a large top of the line power supply, this seems pretty BS.
     
  13. Eldata

    Eldata Limp Gawd

    Messages:
    366
    Joined:
    May 10, 2011
    No sleep issues with 1503 (board came with that), 1606 or 1702. Skipping 1703. And the board goes in and out of sleep at least ten times a day, including overnight. No, I'm not afraid to update the BIOS because of possible sleep issues.

    In fact, I've never had an Asus board with sleep issues that I wasn't able to correct (knocking on wood). Even a still networked twelve year old CUSL2 based system , running Win98 SE, has no S3 sleep issues here.
     
  14. n1tr0

    n1tr0 n00b

    Messages:
    27
    Joined:
    Jan 21, 2011
    ... Good for you :rolleyes:

    So what you're saying is that you don't believe this is a bug as has been acknowledged by the Asus reps and that instead there are just a large number of people who got crappy/defective boards ?
     
  15. therat

    therat Limp Gawd

    Messages:
    157
    Joined:
    May 18, 2011
    Well the sleep bug is confined to the P67 boards as the Z68 boards don't suffer the sleep problem at all. As to who to blame, hard to say because almost all manufacturers of P67 boards have sleep problems with some of their boards but not all.
     
  16. DBRASCO

    DBRASCO [H]Lite

    Messages:
    65
    Joined:
    Jan 8, 2004
    I followed Raja's instructions to a "T". On flash procedures. I was stable for over 3 weeks mild over clock and sleep worked flawlessly. I flashed it to 1703 broke something. Even with a flash back. I can now not use XMP or manual memory settings. The board will hang or take for ever to load windows. S3 sleep is totally broke.

    So some one at Asus explain that. My P5 DH Deluxe has been able to sleep for 4 yrs no problems over clocked e6600 @ 3.8
    Either it's Asus and a poor manufacturing or a poor Bios coding that seems to be causing the problems!

    I feel like maybe a class action lawsuit could get something accomplished. It's a feature of the product I bought and would like to use. My choice so all the nay Sayers who say don't use it. That's your choice.

    It would either 1. Make Intel or Asus take responsibility and quit pointing fingers back and forth. 2. Get a judgement against either for reimbursement of some Kind for faulty product.

    Brasco
     
  17. Eldata

    Eldata Limp Gawd

    Messages:
    366
    Joined:
    May 10, 2011
    There are two scenarios here;

    1. Sleep issues with PLL OV not disabled.
    2. Sleep issues with PLL OV disabled.

    The first scenario is what Asus calls Intel's bug. Just about all P67 boards suffer from it. The workaround is to disable PLL OV. However, many boards do not respond to the workaround and hence the second scenario. Note that the wording of the last sentence makes the two scenarios appear to be related. This is not necessarily so.

    My post is suggesting that Asus has not done enough to ascertain the true nature of the second scenario. Instead, they appear to be insinuating that all sleep issues should be blamed on Intel. It may well be that the two scenarios are completely unrelated. Yes, the second scenario may be due to crappy/defective boards and I was lucky enough not to get one.
     
  18. Eldata

    Eldata Limp Gawd

    Messages:
    366
    Joined:
    May 10, 2011
    Well the sleep bug is confined to the P67 boards

    This is not true. PLL OV has to be disabled on Z68 boards too. There are also Z68 boards that do not respond to the workaround. However, complaints are far fewer than for P67 boards.
     
  19. Eldata

    Eldata Limp Gawd

    Messages:
    366
    Joined:
    May 10, 2011
    What's stopping you from RMAing the board? You're mistaken if you believe all P67 boards have the issues you mention.
     
  20. Ocellaris

    Ocellaris Ginger @le, an alcoholic's best friend.

    Messages:
    18,838
    Joined:
    Jan 1, 2008
    I think this is an Asus problem and not an Intel problem.
     
  21. DBRASCO

    DBRASCO [H]Lite

    Messages:
    65
    Joined:
    Jan 8, 2004
    What's stopping me is the scenario that has happened to most, the new board does the same or has more problems and the time and labor I'm down to fix a problem that Asus is doing little to address if another board fixed it to break down the road.

    No I don't believe all boards have the issue as shown by this forum, but most seem too. Hence Asus has not done enough to fix the problem. I won't ever believe they can't duplicate the same issues in house with PLL OV on and with it off. But the excuse we keep hearing is we are trying to duplicate the problem, or it's intels problem.
     
  22. therat

    therat Limp Gawd

    Messages:
    157
    Joined:
    May 18, 2011
    Well if Asus can't replicate the problem maybe they ought to obtain one of the systems that does and pay all the costs of getting it. Maybe then with a faulty system they can finally solve this problem.
     
  23. Disco_Stu_04

    Disco_Stu_04 Limp Gawd

    Messages:
    426
    Joined:
    Aug 18, 2003
    Agreed. I've had an MSI board and an Asus board. Both w/ Sandry Bridge and I've only had sleep problems with Asus.
     
  24. you2

    you2 [H]ard|Gawd

    Messages:
    1,191
    Joined:
    Aug 10, 2004
    I've had sleep/resume problem with p67 msi and z68 asrock. Same symptons. When I did a search for a fix I found folks had problems iwth pretty much every brand mb.
     
  25. Suicidal Insanity

    Suicidal Insanity Limp Gawd

    Messages:
    335
    Joined:
    Jan 6, 2003
    Have an intel P67 board and use sleep daily, never had an issue. No PLL OV enabled tho.
     
  26. Eldata

    Eldata Limp Gawd

    Messages:
    366
    Joined:
    May 10, 2011
    Try the new 1850 beta BIOS for P8P67 Deluxe. You don't seem to have much to lose.
     
  27. DBRASCO

    DBRASCO [H]Lite

    Messages:
    65
    Joined:
    Jan 8, 2004
    Flashed to 1850, What ever broke flashing from 1702 to 1703 is remaining broke.........1850 will still not let the computer use XMP settings and wake....... now when it trys to wake from sleep I get 3 beeps and it trys to reboot..... pushing memok button does nothing.... No overclock right now at all.

    all default settings, except enable all C states and XMP, disable PLL OV and CPU spread spec. No Wake up.....



    So I'm at a loss with this board....
     
  28. n1tr0

    n1tr0 n00b

    Messages:
    27
    Joined:
    Jan 21, 2011
    you can just keep RMA'ing it trying to get a good one like I did until either you run out of thermal paste, sticky pads for the cooler backplate, or patience ... nothing like having to call Microsoft to reactivate your products (a good reason to have bought retail copies instead of oem) :rolleyes:
     
  29. Dan_D

    Dan_D [H]ard as it Gets

    Messages:
    54,825
    Joined:
    Feb 9, 2002
    I've never had issues with S3 sleep functions on any of the ASUS boards at stock settings or even overclocked so long as PLL overvoltage was disabled. With it enabled, it's never worked for me.
     
  30. Eldata

    Eldata Limp Gawd

    Messages:
    366
    Joined:
    May 10, 2011
    Same here and with most P67 users IMO. But this begs the question....Why doesn't Asus change the default Internal PLL Overvoltage setting, in the BIOS, from Auto to Disabled? They can always change it back if and when they or Intel correct the problem. It's as though they want to tie all sleep issues to Intel and this one setting.
     
  31. Dan_D

    Dan_D [H]ard as it Gets

    Messages:
    54,825
    Joined:
    Feb 9, 2002
    I can't answer that. Perhaps the ASUS guys can. What I have noticed is that ASUS tends to do a lot of things design wise, which are different than the way other manufacturers do things.
     
  32. Eldata

    Eldata Limp Gawd

    Messages:
    366
    Joined:
    May 10, 2011
    There's a "corollary" of Murphy's law that suggests the possibility that something else could have failed around the same time you upgraded to the 1703 BIOS. An experienced troubleshooter should always expect the unexpected.
     
  33. ppo

    ppo n00b

    Messages:
    37
    Joined:
    Jan 20, 2011
    The real problem with the S3 bug is that NO COMPANY has yet officially confirmed that the bug exists.
    Therefore the problem is on our side (the users). Try this, try that, what is your hardware, drivers etc...RMA the board, get a new one, re-install, no joy.
    And that has been happening since January.
    And here's the last one: Sent the P8P67Deluxe Rev B3 for RMA with the mention of the S3 bug to ASUS Canada. They send me a new one: same problem. WTF....
    They're stalling the users.
    I have become very proficient at re-installing mb's mind you.
    But after all that time, I have had it. The ASUS P8P67 Deluxe is in the trash can.
     
  34. Eldata

    Eldata Limp Gawd

    Messages:
    366
    Joined:
    May 10, 2011
    I've been using Asus boards since 1995. Always thought highly of their hardware but have a totally opposite view regarding their firmware.

    Asus Hardware == 1st World
    Asus Firmware == 3rd World

    A perfect example is in the BIOS for the P8P67 Deluxe. One of the devices enabled by default is a non-existent serial COM port. This causes Windows to set aside resources for it unless the user disables it in the BIOS. I'm almost certain that hidden sloppier scenarios abound.
     
  35. The Mac

    The Mac [H]ardness Supreme

    Messages:
    4,492
    Joined:
    Feb 14, 2011
    umm....i believe there is a com header on the board.

    allot of people still use serial devcies...

    i have a box of risers with comm and paralel ports with header connectors.

    i dont disable it, and its never caused a problem.
     
  36. Eldata

    Eldata Limp Gawd

    Messages:
    366
    Joined:
    May 10, 2011
    You are mistaken. There is on other P8P67 variants but none is on the P8P67 Deluxe. I have the board and can show you a picture with the vacant header and support IC pads.
     
  37. 123Lanoix

    123Lanoix 2[H]4U

    Messages:
    3,876
    Joined:
    Nov 13, 2005
    sleep is a stupid feature anyways and is counter productive

    just shut your computer off

    are you running a library or an internet cafe?
     
  38. Dan_D

    Dan_D [H]ard as it Gets

    Messages:
    54,825
    Joined:
    Feb 9, 2002
    No, the ports are not non-existent. There are headers on the board's PCB to connect the actual ribbon cable and port. The port is simply no longer included on the I/O panel as they once were in say the 1990's.

    Could be. I haven't looked at all of them that closely, but all of the P8P67 series I've seen has had the ports on them.

    Many people are going to disagree with that. I don't actually turn my computer off or let it sleep. I simply have it turn off the monitor(s) after a certain amount of time. I've never used the sleep functions because in the past, it's been harmful to mechanical disk drives. I've also found that Windows itself (in the past) has handled sleep mode badly and waking up from a sleep state has often been anything but reliable. These days those concerns aren't really all that valid, but old habits die hard. That and I am on my computer so much that sleep mode would generally annoy me more than anything. I let my laptops and HTPC do it but not my server or my gaming rigs.
     
  39. Eldata

    Eldata Limp Gawd

    Messages:
    366
    Joined:
    May 10, 2011
    Repeating....There is no serial COM port header on the P8P67 Deluxe. I'm sure of that. You're guessing.

    It's user ignorance that allows Asus to get away with crap like this. No doubt you also consider the matter trivial as they obviously do.
     
  40. 123Lanoix

    123Lanoix 2[H]4U

    Messages:
    3,876
    Joined:
    Nov 13, 2005
    i am happy with my p67 sabertooth it has worked soildly in this july heat

    sleep function works if you do not overclock at all but if you do then it doesn't work
    maybe an intel thing or an asus thing but iam pretty happy with the lastest bios