580gtx causing system slowdown

S

shiro321

n00b
Joined
May 10, 2008
Messages
6
i just got an evga 580 gtx sc (sent the first one back to evga for the same issue). when i put it in my system i get massive cpu spikes and slowdown, it literally takes 20 minutes for windows to boot up. i have been working with evga tech support and gone through everything in the book troubleshooting this. Its has come down to be the memory or the motherboard, i am not sure which. if i have one stick of ram in the pc it runs fine no lag and no spikes, but if i put another stick in the lag is back, it does not matter which sticks i use its always one is fine 2 or more not. i have run memtest86 and the ram went through 2 passes with out error. the really weird thing is if i put my old 9800 gtx+ in i can have all the ram in and it runs normal.

I am not really sure how to proceed from here, i do not have any extra memory on hand or know anyone i can borrow some from and i really don't want to go buy ram i can't return for it to turn out to not be the problem. any help on how i should proceed would be greatly appreciated.

my rig:
Foxconn Renaissance
core i7 920
6gb super talent WA133UX6G8 ddr3 in tri-channel
ultra x4 1050w power supply 87a on +12
windows 7 x64 (fresh install)

what i have done so far in no particular order:
-reset bios to default
-manually set memory timing
-tested ram with memtest
-tried another power supply
-clean install of windows
-tried different pci express slots
-tried ram in different slots
-drivers that came with card
-newest drivers from nvidia
-banging head against wall repeatedly
-removed unnecessary components and put them back one at a time until the problem reappeared.
 
Last edited:
Sounds like the motherboard, can't handle multiple memory channels at the same time as a high-end graphics card being used. I don't think the high-end foxconns got great reviews for reliability anyway.
 
With the help of caveman-jim from Rage3d.com, I found the fix and posted it in the thread in the nvidia section. Basically, some x58 chipsest have PCI-E problems, maybe even a bad batch of chips that had issues previously unknown to anyone (What? Intel having weird glitches in their chipsets...noway...).

WORK AROUND: Simply install a non GF 4/5xx series card in a spare PCI-E slot, then the new 5xx cards will work perfectly, all devices, no performance issues.

You can use an older ATi card, older nvidia card, etc... It would be preferrable to use an older nvidia card you can dedicate to physics, because at least then it's worth having something sitting in the otherwise wasted PCI-E slot.
 
Last edited:
BTW: I can tell you first hand: EVGA tech support is FULLY aware of this workaround...but they refuse to acknowledge it.
 
Specifically referring to delays on bootup? Interesting. I don't think that solves the bigger 4/5 series card issues with other chipsets.
 
thanks for the help so far.

yes i update the bios to the newest version

i tried putting in my 9800 gtx+ along with the 580 gtx the machine will not post, i get no image what so ever, i also tried an old 7950 gt instead of the 9800 and the same thing happened.

also with the second card in when you push the power button the pc starts up for a second then shuts its self down and starts it self back up.
 
Well, hate to say it but with pulling in an additional card it turns off as soon as you power it, then I would suspect either insufficient power supply or too much CPU overclock as the problem or a combination of the problem.
 
there is no overclock on the cpu and it is doing it with two different power supplies, an ultra x4 1050w and a corsair x850tx. Both of which are more then sufficient to run a 580 and 9800. i believe it is a bug in the bios, because it starts up for about 1/2 second then shuts off then starts up 1 second later like nothing happened but with no image or post. it give an error of 85 on the led screen on the board but that is pretty meaning less because even foxconn support does not know what the codes mean.

i found a fix for the slowdown. it turned out to be a memory mapping bug in the bios an explanation and fix can be found here:

http://foldingforum.org/viewtopic.php?f=38&t=15472&sid=ed7b9e677e15791d8d967d79a3115e32&start=15

i lost 704mb of ram but its better then only being able to use one 2gb stick.

i am going to email this to evga and foxconn as well.

thanks for your help though
 
BBA said:
Well, hate to say it but with pulling in an additional card it turns off as soon as you power it, then I would suspect either insufficient power supply or too much CPU overclock as the problem or a combination of the problem.
Click to expand...

No, it's just hard-rebooting, not powering off due to overload. This is normal with a hardware configuration change (and even just a power supply switchoff to the system) with most modern PCs.
 
after a bit of playing i got my system to boot up in to windows with 580gtx and the 9800 gtx+ in. it will display just fine when it boots up in to windows but displays nothing at post or up until the welcome screen.
 
samuelmorris said:
No, it's just hard-rebooting, not powering off due to overload. This is normal with a hardware configuration change (and even just a power supply switchoff to the system) with most modern PCs.
Click to expand...

No, its not normal except for failed configurations. Sounds like he figured it out with two cards.
 
BBA said:
No, its not normal except for failed configurations. Sounds like he figured it out with two cards.
Click to expand...

Nope, it is normal (one reboot) for hardware config changes. Failed hardware configs would be if it rebooted twice.
 
samuelmorris said:
Nope, it is normal (one reboot) for hardware config changes. Failed hardware configs would be if it rebooted twice.
Click to expand...

Did he not say it turned off as soon as the power switch was pushed? That's NOT normal and that is the point I am making.

Why do you feel the need to point out a normal reboot after installing hardware is anywhere near related to the same thing?
 
It is normal. The reboot process of Intel chipsets at least, at least as far back as the P31/P35, if not the 965, when hardware configs has changed turns the entire system off, fans and all. It is then powered back up automatically. Only if this happens twice (unstable hardware, BIOS settings reset), or if the system does not automatically power back up is it a problem.
 
samuelmorris said:
It is normal. The reboot process of Intel chipsets at least, at least as far back as the P31/P35, if not the 965, when hardware configs has changed turns the entire system off, fans and all. It is then powered back up automatically. Only if this happens twice (unstable hardware, BIOS settings reset), or if the system does not automatically power back up is it a problem.
Click to expand...

The problem here was resolved.

About the reboot: You are wrong, don't even act like you are not, just admit it and let it go. The ONLY time systems power up and instantly reboot is because of a config ERROR or hardware incompatibility/failure. You DO NOT know what you are talking about. Get over it already.
 
I'd get over it if I was wrong, but er, no I'm not? I've run Core 2 systems on Intel chipsets for almost four years. I know how they work!
 
I have been having random system lock-ups with my sig rig. I have tried 3 sets of corsair dominator 1600 etc. nothing seems to right the problem. I might be able to play BC2-Vietnam for 4 hours then poof lockup. Isn't heat related either as my 580's don't go over about 65C nd my CPU is showing 33C. I wonder if this is my problem.

In the event viewer there are PCI-E errors but nothing I was able to pinpoint as to exact cause.
 
PCI-e errors? I'm not sure they'd ever be there on a properly working system, but IIRC the bigbang boards all use nf200s? or is it just the P55 versions? Could be something to do with that.
Do you have the same issue with only one 580 installed?
 
I have removed one from my system and will try running just one for a bit and see what happens. I did a fresh windows install on a fresh disk and everything was fine until I installed the nvdrivers.

Here is the exact error I just had with both installed:


A corrected hardware error has occurred.

Component: PCI Express Root Port

Details:
- System

- Provider

[ Name] Microsoft-Windows-WHEA-Logger
[ Guid] {C26C4F3C-3F66-4E99-8F8A-39405CFED220}

EventID 17

Version 0

Level 3

Task 0

Opcode 0

Keywords 0x8000000000000000

- TimeCreated

[ SystemTime] 2011-02-20T17:32:34.483275600Z

EventRecordID 1166

- Correlation

[ ActivityID] {040748DB-5354-49C4-BA73-E8D8DF7670E9}

- Execution

[ ProcessID] 1288
[ ThreadID] 1824

Channel System

Computer JT

- Security

[ UserID] S-1-5-19


- EventData

ErrorSource 4
FRUId {00000000-0000-0000-0000-000000000000}
FRUText
ValidBits 0xdf
PortType 4
Version 0x101
Command 0x10
Status 0x507
Bus 0x0
Device 0x3
Function 0x0
Segment 0x0
SecondaryBus 0x0
Slot 0x0
VendorID 0x8086
DeviceID 0x340a
ClassCode 0x30400
DeviceSerialNumber 0x0
BridgeControl 0x0
BridgeStatus 0x0
UncorrectableErrorStatus 0x0
CorrectableErrorStatus 0x1000
HeaderLog 00000000000000000000000000000000
Length 672
RawData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
 
Well I popped one out and had a crash about 5 min into my system usage. Swapped out to the other one by itself and will see what happens.
 
Been running nearly 10 hours without a hitch on this one card. Gonna RMA the other I guess.
 
Last edited:
samuelmorris said:
I'd get over it if I was wrong, but er, no I'm not? I've run Core 2 systems on Intel chipsets for almost four years. I know how they work!
Click to expand...

No, aparantly you do not, but hey, why not step into current hardware while you are still learning. Core2 is old news, I've been doing this since the mid 90's, the OP has an I7 and you are irritating, so I am going to add you to my ignore list.
 
It doesn't hurt to read a signature once in a while, jesus christ, get over yourself.
 
You must log in or register to reply here.
Back
Top