lawtq

OH MY GOD!!!

S3 Sleep works flawlessly with this bios, rev 1.1 board, it has never ever worked before, finally, a 2 second bootup!!

And this is with a 4.4ghz OC!

  You flash a 760 board with a 758 bios? 

This so called "Double Pump" happens when you have Virtualization Technology Enabled in the BIOS, it may also happen when you make certain changes in the BIOS that require a restart.

Disabling VT should stop this from happening, its disabled by default now in the new BIOS so that should stop it happening automatically, if you enabled VT then it will continue to happen.

lam_das

When O When Are We Going To See The Working S3!
When O When Are We Going To Realize Full Potential Of i7920!
Please Fix S3, EVGA, We Beg Of Thee!

LOL +1

I hope everybody that complains here about S3 is also emailing EVGA regularly like I am.

They are polite about telling me how sorry they are and how soon they will fix the issue... but they never do.

This was my first EVGA board and probably my last.  I don't hate it... it does MOST of what I need it to do.  I'll use it for 2 more years probably.  But it (EVGA) certainly didn't live up to the hype.

timmeh84
hey i was just wondering as i am in the progress of returning my corsair memory due to major stability issues.. which ram is generally pretty stable.. OCZ????

I've been using OCZ Platinums since ~May 2009 (3 x 2GB running to spec 1600MHz, 1.65V, 7-7-7-24) and have not had a single problem.

christopher_c_williams

i see lots of complaints about the S3 problem along with mention of the mobo revision either rev 1.0 or 1.1; can someone please explain in specific detail what the differences are between the revisions in the X58 SLI boards because i have one of each and i'm starting to see differences in the machines (when all hardware is literally identical).  i don't put the machines to sleep much, but it would be nice to do so eventually. 

if there is already a different post that explains the differences, can someone kindly point me in the right direction.

thanks.

I would be interested to see if EVGA would state the differences.  I have never found anything concrete about the differences.  I had a rev 1.1 board that I had to RMA and got sent a 1.0 one and I see no difference.  One person (I forget who) in a post back at that time had a theory that it was a different clock generator chip, but when we compared, my 1.1 had what he expected on the 1.0 and vice versa.

Also, I have no sleep issues with my 1.0 board, but I do not have a really high OC.

Is also could be possible that they took the 1.0 board and made changes to it before sending it to me as an RMA replacement.

Not trying to start any debate, but my feeling is that the issues some people are experiencing with 1.0 boards and S3 might not be simply an issue with 1.0 versus 1.1 versions.

If you do find something concrete somewhere, please do post it here.

jasonanderson

christopher_c_williams

i see lots of complaints about the S3 problem along with mention of the mobo revision either rev 1.0 or 1.1; can someone please explain in specific detail what the differences are between the revisions in the X58 SLI boards because i have one of each and i'm starting to see differences in the machines (when all hardware is literally identical).  i don't put the machines to sleep much, but it would be nice to do so eventually. 

if there is already a different post that explains the differences, can someone kindly point me in the right direction.

thanks.

I would be interested to see if EVGA would state the differences.  I have never found anything concrete about the differences.  I had a rev 1.1 board that I had to RMA and got sent a 1.0 one and I see no difference.  One person (I forget who) in a post back at that time had a theory that it was a different clock generator chip, but when we compared, my 1.1 had what he expected on the 1.0 and vice versa.

Also, I have no sleep issues with my 1.0 board, but I do not have a really high OC.

Is also could be possible that they took the 1.0 board and made changes to it before sending it to me as an RMA replacement.

Not trying to start any debate, but my feeling is that the issues some people are experiencing with 1.0 boards and S3 might not be simply an issue with 1.0 versus 1.1 versions.

If you do find something concrete somewhere, please do post it here.

 
Qoute from the article. It is from January 16 2009.
 
"Speaking of those problems, the first one is the S3 resume bug that occurs when the board is clocked with a high BCLK setting. While EVGA has worked on improving the resume features when the board is overclocked, it does not always work right. We have found with the latest BIOS that the board typically resumes from S3 state when the BCLK is kept below 166 or so, not always, but generally nine out of ten times in our testing. Anything higher and it is a crapshoot if the board will resume properly. This is a hardware limitation based on the Clock Generator utilized by EVGA."
http://www.anandtech.com/weblog/showpost.aspx?i=538 Read half way through and you will see. Also I can tell you when I owned that board that I had problem on stock settings also not just OC'ed. There are still problems with this board and resume from S3 such as loss of devices USB or Hauppauge TV tuner or if you are trying to shut it down after S3 it will get stuck. How ever things have gotten better but they will not be completely resolved. Talking about revision 1.0 of the X58 3xSLI
 
I did not own 1.1 revision of the EVGA board so I cannot comment on the S3 but many forum members report that it works the way it should.
 
I am not trying to put EVGA boards down or nothing. As a metter of fact I am not planing to own any other boards other than EVGA due to good CS and these forums and their constant atempts to make us happy with BIOS revisions every 2 weeks however if we do not complain they will not do anything about it.

Agreeing with findings of anandtech.com is same as admission of guilt. However they did not disagree with it either. I do not blame them for it. If they would agree with it then everyone would want another board even the guys that do not know what the S3 is or what it does. Other X58 chipset board manufacturers had S3 issues but so far they managed to resolve those issues through BIOS updates.

I have the same problem with s3 resume,I have a rev 1.0 board and sz2x bios.
I did a test and changed nothing in the bios but set the vcore to 1.88,on the third resume from s3 mode it failed and I had to restart :(
I bought the board 3 month ago and was hoping to get a rev 1.1 :(
I am considering to sell my board and get a gigabyte.

jasonanderson

christopher_c_williams

i see lots of complaints about the S3 problem along with mention of the mobo revision either rev 1.0 or 1.1; can someone please explain in specific detail what the differences are between the revisions in the X58 SLI boards because i have one of each and i'm starting to see differences in the machines (when all hardware is literally identical).  i don't put the machines to sleep much, but it would be nice to do so eventually. 

if there is already a different post that explains the differences, can someone kindly point me in the right direction.

thanks.

I would be interested to see if EVGA would state the differences.  I have never found anything concrete about the differences.  I had a rev 1.1 board that I had to RMA and got sent a 1.0 one and I see no difference.  One person (I forget who) in a post back at that time had a theory that it was a different clock generator chip, but when we compared, my 1.1 had what he expected on the 1.0 and vice versa.

Also, I have no sleep issues with my 1.0 board, but I do not have a really high OC.

Is also could be possible that they took the 1.0 board and made changes to it before sending it to me as an RMA replacement.

Not trying to start any debate, but my feeling is that the issues some people are experiencing with 1.0 boards and S3 might not be simply an issue with 1.0 versus 1.1 versions.

If you do find something concrete somewhere, please do post it here.

That was me :)

Yes it seems atleast some EVGA X58 3x SLI 1.0's have the same clock gen as the 1.1's and more surprisingly the clock gen on my 1.0 is still being used on the EVGA X58 P55 boards so i am not sure what going on here, possibly the started adding these new clock gens to the boards before they  could get rev 1.1 printed instead of 1.0 or maybe they had a bunch of 1.0 PCB's made and the clock gen should just be a drop in change.

The weird thing is my 1.0 with the same clock gen as on 1.1's and even new P55 boards has problems resuming from S3 sleep and i even use a 965 which i overclock using the multi but once the MHz gets up there to around 3.80-4.0GHz the board starts having problems resuming from sleep so it seems this problem is more than just a BClock/Clock Gen based issue.

We have been working with Nvidia to integrate dual Nvidia [ with DHIC card into a E758 MB. Using this BIOS version. Currently we can not boot due to video memory limitations. Nvidia is recommending a BIOS change to support this configuration.
 
Nvidia Engineer we are working with (they are available to support):
 
Doug Traill | Senior Solution Architect | NVIDIA Corporation |
dtraill@nvidia.com| Tel: 321 610 8018 | Cell 408 646 2166
 
Would it be possible to get this modifcation done?

Moltenlava
Yes it seems atleast some EVGA X58 3x SLI 1.0's have the same clock gen as the 1.1's and more surprisingly the clock gen on my 1.0 is still being used on the EVGA X58 P55 boards so i am not sure what going on here, possibly the started adding these new clock gens to the boards before they  could get rev 1.1 printed instead of 1.0 or maybe they had a bunch of 1.0 PCB's made and the clock gen should just be a drop in change.

The weird thing is my 1.0 with the same clock gen as on 1.1's and even new P55 boards has problems resuming from S3 sleep and i even use a 965 which i overclock using the multi but once the MHz gets up there to around 3.80-4.0GHz the board starts having problems resuming from sleep so it seems this problem is more than just a BClock/Clock Gen based issue.

I design and support some PCI boards.  If I change a chip on a board that uses the same socket or footprint as the old chip, I may or may not change the board revision on a sticker or printed on the board but I don't usually change the rev number that's etched into the copper - since the board's artwork hasn't actually changed.

[edit] "Artwork" - there's an old-school term.  We used to hand tape the layers of the board on giant mylar sheets and photo reduce to working artwork.  Two colors of tape if it was a double -sided (!) board and photograph through color-separating filters.  That was back in the 70s before most of you over-clocking twerps were born.  Now where did I put my tri-focals and my teeth?

If I change the traces or holes on the board - different pins connected, different parts with different footprints, or even just relocating a trace due to cross talk, impedance, or delay issues - then I change the rev number etched in the copper.

Maybe EVGA does something similar.  That could explain the chip vs rev level puzzle.  Or not.  Doesn't really matter anyway...

Just a WAG,
Z.

timmeh84 hey i was just wondering as i am in the progress of returning my corsair memory due to major stability issues.. which ram is generally pretty stable.. OCZ????

I'm using a 12GB mixed lot, 6GB Mushkin Redlines 1600 and 6GB of Geil Ultra1600 GU36GB1600C7TC, at 1636MHz, 7-7-7-24; 4.34GHz cpu clock 24/7.

kiran otter My system is running a bit hotter with this BIOS.  CPU is 5-7C higher at idle, and the VREG temps are nearly 10C higher.  Why???  After an hour of just sitting and surfing the web, the fans start to crank up to compensate.

Didn't happen to me. It quite usual that web pages will run things that crank up the CPU usage of your system without you noticing. There's about a 80% chance that that is what is happening to you. 

geppwnd  What you mean full potential of 920 "needs explaining"?

The only way to get S3 working (partially, because even with this, you get 5 successful resumes and then tempratures of VREG and SYSTEM go haywire) on Rev 1.1 board is to use Auto vcore, which limits the overclocking ability of the board. Using a manual vcore to get to 4.2 or 4.4 Ghz, you get no S3 love. Luckily, my D0 is a good one and give me 4Ghz with Auto vcore. But with HT ON, the difference between 4Ghz and 4.4Ghz is 15-20% in heavily threaded apps like compiling and encoding. So, there you go: can't enjoy the 4.4Ghz love just because S3 don't worky!