NF4 COLD/WARM Boot Problems...Read this.

[Forsa]

ok got it to boot into windows now it just freezes. Put ram into slot 2 and kept reseting and it worked after a few tries . Now to get windows fixed or something.... thanks alot and yes i finally updated sig.

[RGone]

Forsa, have you got enough voltage on that VX stuff. At least 3.2Volts? I am seeing posts before your situation of users of VX running fine for months and then have problems and have to got to TCCD to get it to work well or at all. I am not sold myself on UTT and high volts yet.

 

Elsewhere I saw you make mention of flashing the 3/16/05 bios and maybe you should because it has come to my "understanding" that "high volts" and UTT/BH-5 should perform better; possibly in slots 1 and 3. SO you need a bios that allows for the sticks to be used primarily in slots 1 and 3 and the first bios for that IS the 3/16/05.

 

RGone...

[Quickmcj]

RGone and Angry>

 

Are we making some kind of human error or is there an issue between the motherboard/chipset and the videocard? Can you confirm?

 

The last piece of evidence I can figure to show that I am not doing something wrong, is to try the Leadtek 6800GT in another motherboard with a different chipset. But that is easier said than done.

 

So now we are all holding our breath until ozandangel returns.

[Zar]

I have been having the same Problem with the cold boot. I thought I would let everyone know the only way I could get my comp. to boot was to turn on the comp. then off and back on real quick with hitting the power button right after. I bolive a few others have said this also.

[dsumanik]

Hello all this is my first post and it is going to be lengthy. I have the hard reboot issue described here to the tee, and i understand everyones frustration. All components are BRAND NEW in this rig.

 

here is a thread in which i tried to work through the problem with the OCZ reps unfortunately to no solution:

 

http://www.bleedinedge.com/forum/showthread.php?t=9573

 

 

WHAT I HAVE DISCOVERED THROUGH TESTING:

 

1. No bios revision cures this problem.

2. Putting your vidcard in different slots doesnt help. I havent tried a pci card, or different pcie card yet.

3. Resetting cmos, or booting with "safe settings" jumper doesnt help.

4. I RMA'd ram, still have same problem...more on this later.

 

 

THE TEMPORARY CURE:

 

If you do a hard reboot, and the system wont post...3 lights on your diagnostic indicator should show up, these lights are located on the lower left of the board. You can also hook up your frontx led indicator on the SLI-DR

 

This means that the board is having trouble detecting the memory, or starting it up.

 

To get the system running (tested using 310 or 414-2 bios):

 

1. Reset cmos. (this isnt the cure only a part of it)

 

2. You will have to boot in single channel in order to get the system to post from here so put one stick into slot 2 (top orange slot) and try to post.

 

If it works go into bios..... if it doesnt swap sticks into the top slot and try again.

the system should post eventually, repeat these steps until it does.

 

NOW

 

once it posts in single channel, go into bios:

 

up voltage on the ram to a little higher than its recommended spd rating. I use ocz vx4000 and crank it to 3.2 volts.

 

then MANUALLY select the spd timings of your ram in genie bios.

 

for my VX i use: 2-3-3-8

 

config HDD, RAID, peripherals, Boot priority etc. as you normally would.

 

save these settings to one of the CMOS RELOADED slots. Label it "jump start"

 

save and exit.

 

Turn off PC (Dont unplug power cord..more on this later)

 

 

FROM HERE LABEL THE "BAD STICK" SO YOU DONT HAVE TO FIGURE OUT WHICH ONE IT IS AGAIN.

 

 

Now put in the bad stick (slot 4, bottom orange) and the system should post.

 

From here you can enter bios and overclock away.

 

 

 

KEEP IN MIND:

 

Soft reboots are safe, so once you get it up and running leave it alone if you want to save yourself frustration. If you have alot of fans and need silence at night, just set up in windows for the sytem to go to standby, it is silent then.

 

If you need to turn power off/move PC/power cuts out/ and a hard reboot is inevitable or unecessary:

 

Do a soft reset first, go into bios SAVE YOUR OVERCLOCKED SETTINGS into another bank in CMOS RELOADED....then load from the "jump start settings".

 

save and exit, let system post.

 

Now you can cut power to the system without going through an indian rain dance afterwards.

 

When you want to go back to performance settings from here...boot up, enter bios and reload from the saved bank, do a soft reset and you are good to go.

 

 

 

ODDITIES NOTICED/MY THOUGHTS ON THE PROBLEM

 

- Initially my new system worked fine for 1 or 2 days before the problem arose.

 

- After RMA ing the memory to OCZ, same sequence of events occured. Fine for a couple days, then reboot issue reared its ugly head.

 

- using a non UTT mem module (corsair pc4000 pro), the system posted first try everytime

 

- if system is left alone for a period of time (can be a random interval) with power cord plugged in and power applied to mobo, sometimes it just magically fixes itself and will boot.

 

- occasionally the system will post from a cold reboot on its own...however, pulling the power cord from the PSU and letting the mobo sit with no juice at all DEFINATELY results in a 3 light, no post situation afterwards (unless jump start settings are loaded)

 

- the led's on my pc4000 pro show some intial activity, then go out after about 10-12 seconds when power first hits the board after the cord has been pulled out from the PSU or power switch turned off. Perhaps this is when the board detects spd timings? Trying to power up after that time interval helped in some instances when it wouldnt post.

 

 

My personal thoughts:

 

Obviously an issue between the lan party board and UTT mem modules. May be fixable in a bios update.

 

OR

 

The dfi board doesnt have enough juice to power up the modules initially after a cold reboot, especially ona UTT module which requires high voltage in order to run at rated speeds......this would explain why the system will magically boot if left alone for awhile..because capicitors and mosfets etc have a chance to charge up fully, before an attempt to post is made.

 

 

It seems that after a good burn in something happens to the UTT VX that makes it hard to detect by the mobo, or require more initial voltage to post.. My old sticks wouldnt post and when i got the new ones it booted first try after i deliberatly did what i knew would DEFINATELY cause 3 light syndrome...but it posted with no dinking around, and i could power down as often as i liked for the first day.

 

After 2 days of hard burn-in/testing to find the limits of my OC, problem returned, occasionaly at first, but now... every time. Same cure fixes it now.

 

The sticks "slowly died."

 

HOWEVER

 

i dont think it is likely i got two sets of defective sticks, because i was told they were tested on a dfi board before they were sent. However id like those same testers to run the mem at 3.5 volts for a week in memtest and see what happens.

 

Also another thing to think about is that they work fine once jump started, no errors, and performance is astounding. 2-2-2-5 @ 258 rock solid.

 

So if they were defective....wouldnt they generate errors and blue screens up the yin-yang?

 

After reading this thread and seeing the vidcard problems, perhaps pcie solts could be related as well I have a 6800gt...i will go read the thread on the evga forums and see if the problem can be cured via an RMA from them if it seems like that is the source.

 

I dont think its the vidcard though...the common factor here is the board...and it definately is in need of some bios work still (as all new boards do)

 

I hope this helps solve some problems, if we work together i think we will be able to figure it out guys, dont give up.

[dobbs]

I also have cold/warm boot problems with my SLI-DR. Here's my experience thus far:

 

First off, the system was shipped to me from Monarch Computer in Single mode with both cards in and connected via the SLI bridge. I had problems from the get-go with the thing not "posting" , as it were. I use a HP L2335 23 inch widescreen monitor via the DVI adapter, FWIW.

 

The very first thing that I noticed was that it would just not display an image on screen in Windows for anything. I could hear the sucker boot into Windows via the startup sound, except I would get no picture. It was as if the video card was not sending a signal to the monitor to display the picture, or that it was sending a signal that was out of the scan range of the monitor. Huhhhhhhh...

 

After shutting my PC down a few times I then noticed that sometimes it wouldn't even display the POST screen when I cold booted. This also weirded me out. I removed the second video card and left in the card in the PCIe 16x slot (the one closest to the CPU in single mode) and that resolved any problems I ever had with posting or not displaying signals in Windows. This is still via the DVI connection, mind you.

 

I troubleshooted some other problems that I had with my memory timings and eventually got around to giving the second card a go; once again I had issues once SLI was back in the mix. The most prominent of these was during the playing of Half Life 2, I could play the game for about 5 minutes, and then depending on what kind of drivers I was using, I would crash in a different way.

 

Using the 71.89s, I crashed to a black screen that immediately put my monitor out of range (once again, acting like there's a signal loss of some kind betwen the card and monitor) but what was extremely troubling about this was the fact that I could not simply hit reset to clear it. ANY time after I had this crash, my PC would not POST without a cold restart. Pressing reset did nothing -- it acted like it the machine was going to post (as the CPU fan speeds would readjust and the CD-ROM would blink, typical things when you reset your PC) but it never did or at least I never saw anything on screen. Only shutting the system off (via a 4 sec press of the power button) could restore it to "life".

 

Using the 76.45s was a bit different; now when I crashed in HL2 SLI mode, I would crash to a screen that look like a multicolored test-pattern as my sound looped over and over again. More of a stock game crash -- my video stayed on the whole time and never went out of sync, but once again same issue -- only a cold start would bring the machine back.

 

I then called up BFG to see if there was any potential issues they knew of with the cards and this type of behavior or if my core temps were too hot. They said no, but to try running HL2 with each vid card separately to see if that cleared up any of the issues. Naturally it did -- I knew that running in single card mode I would have no cold/warm boot issues and sure enough, I never crashed in Half Life 2 with either card during 1 hour + gameplay sessions.

 

This has led me to the following conclusion: there is definitely something flaky between running 2 cards in SLI mode and the DFI NF4 BIOS. Yeah yeah, that's not new news to anyone who's been reading this thread. But what's interesting is that there seems to be some commonality in between the SLI / DVI users; everything works OK in single card mode even with DVI (even when there's an acknowledged problem with DVI widescreens and Nvidia drivers but as far as I know these only pertain to widescreen resolutions only) but as soon as you hook up that second card, there's instant problems.

 

Right now what I've done is change my motherboard permanently to SLI mode and put my single vid card in the second slot. As others have noted, if the SLI-DR board is in SLI mode via the jumpers and there's a vid card in the slot closest to the CPU, the system will not POST. At least that's what I've experienced. The system WILL run with a card in the 2nd slot though, and that's where I'm at now. Prior to this, I had always been plugging the monitor / DVI into that top video slot. I'm going to get home and put my second card in the top slot and see if anything's different while driving the monitor out of the "bottom" slot. I'll test cold boots and see how that goes. Then I'm going to give it the old Half Life 2 crash test and see what happens.

 

Important to note that my system's been through the battery of Memtest/PiFast/Prime95 and passed it all, so it seems like I have a pretty stable CPU/memory situation... its just that once you enable SLI, all stability on this board goes to .. It makes me wonder if memory timings have to be readjusted once SLI is added to the equation, but that really shouldn't be.

 

If anyone's had experiences similar to mine, I'd be anxious to hear about them and if (and how) you cleared them up. I'm getting ready to try the 414-3 BIOS but my fear is that it's a beta and will simply add another variable in a situation where I've spent the last 4 days trying to eliminate variables.

 

Anyhow, my 2 cents. Any feedback welcomed.

[OzSnoal]

Great, informative post dsumanik.

 

Let's hope the problem can be solved.

 

I also believe it can be solved with a bios update - we'll just have to hope and see.

Or keep on RMA'ing

 

Angry_Games or Rgone posted about trying VX in slot 1/3 with the later bios. (414-2/3). Bios was updated to fix using of those slots with VX and he recommended trying those slots.

 

GL, holding thumbs for a fix.

[dobbs]

Well, no dice thus far after a few more hours of testing. I put the second card in my system and the thing would no longer display the POST screen on the bottom video card. It seems that when the motherboard is in SLI mode WITH two cards in the machine, it'll only display video through the top card. What was really strange about this is that Windows somehow gets confused and thinks that I have two monitors. The bottom card was the one I was using prior to installation so my POST info will only display with the monitor connected to the top card, but when it gets into windows, that card blanks and Windows will only display video when the monitor is hooked into the bottom card. Weird, eh?

 

I sorted out this mess by simply uninstalling the Nvidia drivers, rebooting, and then reinstalling them. The system now displays the POST screen as well as Windows from the top card and acts for all intents and purposes like the primary video adapter of the machine.

 

One weird note -- when I was having these problems and tried to enable SLI mode, Windows complained that I could still enable SLI mode but it recommended that I use the SLI connector for better performance. Trouble is, the connector was already on. I took care not to touch the cards or the connectors and when I simply reinstalled the drivers, Windows acted the normal way when I enabled SLI (asking you to reboot, etc.)

 

However, despite my fresh build and funky install procedure, it still crashes in Half Life 2 when running in SLI mode using the 71.89s. So, SLI is still clearly not functioning correctly.

 

I'm trying out the 414-3 BIOS to see if this has any kind of better recognition between SLI / DVI and hammers out some of the problems. Maybe making HL2 work would be a nice side bonus but I'm not betting on it.

[nagini]

Well, I have been away from this thread for a few days but I am back and have good news to report.

 

My booting problem ended up having to do with my monitor. As many have said (including me) we were noticing the system booting but no image on screen. Using a VGA adapter works but no DVI.

 

It turns out that the LCD input autoscan mode is the culprit. By setting the LCD's signal input manually to DVI-D I get video every time.

 

That only took a week and a half and swapping every component I have.

 

So, if you are having this problem - give this a try! :-)

[Quickmcj]

I use a regular Samsung 957DF VGA CRT monitor. No luck so far.

 

My computer works fine with the 6600GT from POV. Boots every time.

 

So I am waiting for my 6800GT card to return from RMA.

[needmorstuff]

nice to see you are back up and running nagini. :nod:

 

Well, I have been away from this thread for a few days but I am back and have good news to report.

 

My booting problem ended up having to do with my monitor. As many have said (including me) we were noticing the system booting but no image on screen. Using a VGA adapter works but no DVI.

 

It turns out that the LCD input autoscan mode is the culprit. By setting the LCD's signal input manually to DVI-D I get video every time.

 

That only took a week and a half and swapping every component I have.

 

So, if you are having this problem - give this a try! :-)

[Scoobie Dave]

Im having a small problem with my pc. When I go to restart my pc in windows it seems to get stuck where it would normally restart and go 'beep' where the bios writing starts up etc. To get around this problem I turn off the psu and then back on again, press the power button and it boots up no problems. It has only been doing this for the last couple of days but im not sure why it's just started? Anyone else had this problem?