Page 1 of 5 123 ... LastLast
Results 1 to 10 of 42

Thread: Report: Not all PSUs support Haswell's low-power states

  1. #1
    Join Date
    Nov 2012
    Posts
    92
    Thanks Thanks Given 
    6
    Thanks Thanks Received 
    0
    Thanked in
    0 Posts

    Default Report: Not all PSUs support Haswell's low-power states

    Saw this little gem today... any thoughts or opinions on how widespread an issue this might turn out to be?

    Quote Originally Posted by TechReport
    Indeed, according to the guys at VR-Zone, Haswell may not play nice with some existing "bargain basement" PSUs. That's because the chip's C6/C7 power states apparently require a minimum load of 0.05 amps on the 12V2 rail. The site says not all PSUs have the appropriate circuitry to let the processor sip that little power without the system becoming "unstable" or shutting down. Current Sandy Bridge and Ivy Bridge processors suck a lot more juice—0.5 amps, to be precise—in their low-power states, and meeting that requirement supposedly requires less complex circuitry inside the PSU.
    Source

  2. #2
    Join Date
    Sep 2009
    Posts
    316
    Thanks Thanks Given 
    20
    Thanks Thanks Received 
    0
    Thanked in
    0 Posts

    Default

    D2D solution for minor rails (3.3V/5V) could meet total 12V minimum load at 0A easily

  3. #3
    Join Date
    Nov 2010
    Location
    Serbia
    Posts
    2,242
    Thanks Thanks Given 
    126
    Thanks Thanks Received 
    8
    Thanked in
    5 Posts

    Default

    And yet again, a non-issue is blown way out of proportion by the hastily self-taught, self-proclaimed know-it-all journalists. Yay.

    So this will only affect those systems whose PSUs have low load/no load protection while not having minimum load resistors present, and the problem will manifest itself as a system shutdown when said system enters a deep stand-by state, from which it has to all but reboot anyway. C6 means all RAM, all cache levels and all register states are purged, and no content is dumped onto non-volatile memory. Only the EFI boot state is kept, and that means that the startup procedure can be skipped, but OS resume will take as long as a cold boot would.

    Who gives a flying fuck?
    Careful what you wish for... You just might get it.

  4. The Following 4 Users Say Thank You to McSteel For This Useful Post:

    cypherpunks (04-29-2013), ehume (04-29-2013), Jon Gerow (04-29-2013), ssshjp (05-01-2013)

  5. #4
    Join Date
    Jul 2011
    Location
    Germany, Berlin
    Posts
    1,559
    Thanks Thanks Given 
    919
    Thanks Thanks Received 
    1
    Thanked in
    1 Post

    Default

    Probably not relevant, but I'm asking around (Intel, Asus, Gigabyte, some psu companies)

  6. #5
    Join Date
    Aug 2010
    Posts
    2,700
    Thanks Thanks Given 
    116
    Thanks Thanks Received 
    0
    Thanked in
    0 Posts

    Default

    One of the French hardware portals supposedly got to question some PSU makers and got this response (listed series are compatible):

    Seasonic: G-series, X-series, P-series
    CM: V-series
    BQ: L8, E9, P10, S7
    Corsair: GS, TX, HX, AX (CX and VX awaiting confirmation)

    Given that L8 and S7 are group regulated, looks like a storm in a teacup. Any reasonably modern PSU should do.

    Also, according to be-quiet's documentation, P10 has minimum load requirement of no less than 0.1 A on 12V1 (answer I got from a rep a while ago). Probably goes further to show issue may be overblown.

  7. #6
    Join Date
    Jun 2012
    Posts
    377
    Thanks Thanks Given 
    5
    Thanks Thanks Received 
    0
    Thanked in
    0 Posts

    Default

    This is what makes me look snazzy for testing zero, as in zero, load ripple / regulation.
    So far I've had one (1) unit that freaked out. It stayed running, but ripple was Not Good.
    (This does not include the two Super Bargain PSUs)

  8. #7
    Join Date
    Oct 2006
    Location
    jonnyGURU forums, of course!
    Posts
    15,706
    Thanks Thanks Given 
    512
    Thanks Thanks Received 
    120
    Thanked in
    91 Posts

    Default

    And.. of course I'm sure we'll be able to disable the C7 state in the BIOS just like when Intel implemented C6 state (which was never enabled by default when it first came out).

    Right now I can't say for sure because everyone's board is running a beta BIOS before launch, but I'm sure we'll see a solution to this already non-issue.

  9. #8
    Join Date
    Dec 2007
    Posts
    558
    Thanks Thanks Given 
    57
    Thanks Thanks Received 
    0
    Thanked in
    0 Posts

    Default Antec PSU C7 Support

    Hi,

    For Antec, the current assessment is stated below. We are working on verifying all series' & sub-series', but our PD department also has other units to test (Computex coming with a couple of launches):

    - HCP Platinum
    - TP Gold
    - EA Platinum
    - HCG M

    I will also prepare some official statement as a press relelase in ~2 weeks and add confirmed series here.

    Joerg @ Antec
    P.S.: Any media honcho, who wants to stay updated, email me. (Competitors exluded )

  10. #9
    Join Date
    Nov 2012
    Posts
    92
    Thanks Thanks Given 
    6
    Thanks Thanks Received 
    0
    Thanked in
    0 Posts

    Default

    Quote Originally Posted by jonnyGURU View Post
    And.. of course I'm sure we'll be able to disable the C7 state in the BIOS just like when Intel implemented C6 state (which was never enabled by default when it first came out).

    Right now I can't say for sure because everyone's board is running a beta BIOS before launch, but I'm sure we'll see a solution to this already non-issue.
    Looks like that is what motherboard manufacturers are going to do, just disable it by default. TechReport posted a follow-up: http://techreport.com/news/24738/few...w-power-states

  11. #10
    Join Date
    Oct 2006
    Location
    jonnyGURU forums, of course!
    Posts
    15,706
    Thanks Thanks Given 
    512
    Thanks Thanks Received 
    120
    Thanked in
    91 Posts

    Default

    Quote Originally Posted by Kougar View Post
    Looks like that is what motherboard manufacturers are going to do, just disable it by default. TechReport posted a follow-up: http://techreport.com/news/24738/few...w-power-states
    I didn't see anything in that article that said motherboard manufacturers are definitely going to allow users to enable/disable C7 in the BIOS. Right now all of the Haswell boards out there for testing have beta BIOSes and none of them are horribly stable except for the Intel reference board.

    That said, I do agree with Robert and would expect the motherboard manufacturers to allow the ability to enable/disable that sleep state. Otherwise there will potentially be a lot of pissed off users.

Similar Threads

  1. Seasonic To Increase Power Supply Prices In United States Next Month
    By awesomegamer919 in forum PC Power Supply Discussion
    Replies: 23
    Last Post: 11-12-2018, 06:29 AM
  2. Corsair CX450M & Haswell (& Skylake?) Low Power States Support
    By GDILord in forum PC Power Supply Discussion
    Replies: 4
    Last Post: 11-22-2016, 07:55 AM
  3. D5100 Status Report
    By Oklahoma Wolf in forum Off Topic
    Replies: 0
    Last Post: 06-23-2015, 12:35 PM
  4. Eleven PSU's tested at Tech Report.
    By Adamantine in forum PC Power Supply Discussion
    Replies: 20
    Last Post: 10-06-2007, 02:42 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •