Sandy Bridge: Problems with overclock after coming out of sleep

[996GT2]

I've noticed that my i5-2500K will only clock up to 3.4 GHz (yes, 3.4, not 3.3) after coming out of sleep mode, even though I have it set to 42x multiplier in BIOS. When I do a full restart, it will run at 4.2 again without issues.

Has anybody else had this occur with their SB systems? What's the solution?

 

[slag]

yep, same problem, same board.. I just dont let it sleep anymore.

 

[996GT2]

Yeah, I'm going to contact Gigabyte about it and see what they can do. You should try doing that as well since more voices always helps.

 

[iCyborg]

Funny that my Asus X58 had a similar problem whenever bclk was set >170MHz. They released a BIOS fix for this specific problem - that's the only time I've ever updated MB BIOS. I guess you'll have to wait for a bios fix too, don't really get what's so hard with sleep, as I still have some problems with my webcam that sometimes hangs after I try to start a video call after sleep. Very annoying since I can't kill the process, can't even shutdown, have to cold reboot it...

 

[Edrick]

Not only overclock, but LLC and vCore values. (See other thread)

 

[Dadofamunky]

Bummer, sleep mode comes in really handy. I think a BIOS fix is the only solution, like iCyborg said. Fortunately this is a new platform, so I think you can expect to see one in a month or two. Till then, looks like your machine will have to go without naps for awhile!

I wonder if this will happen with my Asus board.

 

[mclaren777]

I'm not willing to sacrifice Sleep Mode for OCing so it looks like I'll be leaving everything stock.

 

[exar333]

Sleep mode is lame anyway; just disable it.

OC ability >>>>>>>>>>>>>>> Sleep mode

Hopefully Gigabyte issues a fix though...

 

[aigomorla]

i have never seen a first generation intel platform ever sleep properly while overclocked.

Sorry pal, wait for a bios update to fix it in about 4-6 months.

I'm not willing to sacrifice Sleep Mode for OCing so it looks like I'll be leaving everything stock.

my computer is broken if it ever goes though a sleep cycle, so i rarely use them. :X

 

[mclaren777]

The Dell computer I'm using right now has been put on Standby (XP) twice a day on average for the past eight years. It's still going strong and I've never had any issues with it. Fully shutting down a computer is a strange concept to me.

 

[PreferLinux]

I've found Vista's and 7's "Hybrid Sleep" takes like 2-3 times as long to resume after being switched off at the wall as it takes to resume from hibernate (available after disabling hybrid sleep). And it takes somewhere around 5 times as long to go to sleep compared to when hybrid sleep is disabled.

 

[MrTransistorm]

A fix is on the way. There are now beta BIOS w/S3 fix available for the following:

GA-P67A-UD7 - F7e
GA-P67A-UD5 - F6d
GA-P67A-UD4 - F7f

Get them here. More info here. I would assume that the other P67 models will have fixes soon, too.

These are beta, of course, so flash at your own risk. Also, apparently it's recommended to use @BIOS (within Windows) to do the first flash from stock. Info here.

 

[996GT2]

A fix is on the way. There are now beta BIOS w/S3 fix available for the following:

GA-P67A-UD7 - F7e
GA-P67A-UD5 - F6d
GA-P67A-UD4 - F7f

Get them here. More info here. I would assume that the other P67 models will have fixes soon, too.

These are beta, of course, so flash at your own risk. Also, apparently it's recommended to use @BIOS (within Windows) to do the first flash from stock. Info here.

Thanks! Hopefully the UD3 gets a new BIOS soon as well.

I have an SSD so fully rebooting isn't a huge issue, but it's still 30 seconds to cold start vs. 5 seconds to resume from sleep

 

[996GT2]

A fix is on the way. There are now beta BIOS w/S3 fix available for the following:

GA-P67A-UD7 - F7e
GA-P67A-UD5 - F6d
GA-P67A-UD4 - F7f

Get them here. More info here. I would assume that the other P67 models will have fixes soon, too.

These are beta, of course, so flash at your own risk. Also, apparently it's recommended to use @BIOS (within Windows) to do the first flash from stock. Info here.

Hmm, apparently these new BIOS releases may not fix the problem fully:

GA-P67A-UD7 - F7e doesn't fix the S3 problem completely for me. My turbo MPs are set to 45 for all 4 cores and after wake up the turbo behaves like the default turbo function, like the mp are set 45,44,43,42 for the cores. At 100% load CPU runs at 4.2 and increases to 4.5 with the load going down. That's exactly how the default turbo function works.
Not 100% fixed from my point of view.

 

[MrTransistorm]

Hmm, apparently these new BIOS releases may not fix the problem fully:

Hmm. That's unfortunate, but at least we know that they are working on it.

 

[ClockHound]

Thanking all the P67 early adopters! We trailing edgers salute you and will thank you again in 3 months when production ready BIOS versions are available for our purchase.

Oh, wait, we late arrivals to the 32nm party need you to upgrade to Z68 systems before we dive into the future shark-free waters.

Please accept our congrats for paying ASUS and Gigabyte to beta test their products - You make computing better for the rest of us.

 

[Edrick]

Hmm. That's unfortunate, but at least we know that they are working on it.

Still has the LLC and vCore issues as well. But yes, they are working on it. I am sure this will be fixed shortly.