Problems with my new AMD64 rig.


Jan 9, 2005
Hey, I'm a long-time reader first time poster here at Tom's Hardware.
Firstly I'd like to apologise if this is in the wrong forum, but right now i'm very desperate.

After looking around, I was in need of a nice new rig to play the latest games on. I eventually decided that it was an AMD64 3500 that I wanted.

The full specs that I chose are;
AMD64 3500+ Sck939 (Winchester)
Asus A8N-SLi
2 x MSI 6600GT (Setup in SLi mode, of course)
160GB 7200RPM, 8MB, S-ATA HDD
1GB (2 x 512MB) PC3200 DDR
400W ATX1.0 PSU

When I first got the componenets everything went together neatly. The first problem that I came accross was that the ASUS motherboard has a 24-pin power input. With only having an ATX1.0 PSU I didn't initially know what to do, after a little searching I found that it was possible to use the power cable that came with my PSU.
Everything from there on in went fine, the rig far exceeded my expectations.

The problem arises when I try to play a game. Just to note, i'm using the latest version of Dx and the nVidia 67.03 drivers aswell as the latest motherboard drivers. The game can play fine for roughly 10 minutes then I get booted to the desktop with an error. It's the windows error with the "Send error report".
The form of crashing happens worse with the 3D Mark benchmark programs. It will often fully lockup and not allow me to do anything.

The games I have tried so far are Half-Life 2, CS:S, Everquest 2, Doom 3. Every one of them crashes (in different time periods). Each of these games runs with blistering performance, it's just these damn crashes.

I'm quite a hardware newbie, I know minor things, enough to build my own rig. I initially thought it was the power, but after checking numerous websites it says that 400W should be adequate for the components in my rig.

If anybody could offer any assistance, I'd be eternally greatful. I'd also like to apologise for any spelling mistakes in the passage above!

Thanks in advance,


Aug 30, 2002
My first guess would be that your case temps are gettign too hot, and your Graphics cards are overheating. A simple test for this would be to open one side of your cae and set up a fan to blow in there. I know it sounds stupid, but more often than not, this seems to be the 'case'.

Hope this helps out...if not check back

If an argument can't be settled in one or two paragraphs, perhaps you're anal and should just let it go...


Former Staff
Sometimes power supplies overheat too. This is especially true with newer components that put a high load on the 12v line, with inexpensive power supplies. If that were causing the crash, you should be able to measure a voltage drop on the 12v line.

But it's probably something else getting hot. I've even seen RAM overheat, and gaming does put more stress on RAM than most other applications.

<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>


Jan 15, 2002
Recent video-cards can suck a lot of juice so having two of 'em in SLI can put a lot of strain on the power supply, make sure that they each get their own dedicated wire and that there is noting else connedted to the same lead.

Watercooled Mobile Barton 2500+ @ 2.6GHz (200MHz x 13)
Abit NF7-S V2.0
2x 512MB of Samsung TCC4
Sapphire Radeon 9700 128MB @ 360/310
2x Maxtor 40GB 7200RPM RAID-0


if everthing else checks out you might try running only one video card. As SLI is a newer technology its possible their might be a bug or two.
Run one card and see if problem reoccurs. if not then put the other card in by itself and rerun all tests.
if you dont get any problems then there is something wrong with the sli part. if the problem happens with the cards running by themselves then the problem is either in the video cards, drivers, motherboard, or power.


Jan 9, 2005
My PSU came with the case, a very cheap case I might add.
This is how i've always done it, bought a very cheap case with a PSU inside. It's a non-branded PSU.
I'm guessing this was a mistake with such a powerful, expensive PC.
Thanks for all your replies guys, I tried it with only one card and it happens.. but the intervals are a little larger.

Would you guys reccommend me buying a decent ATX2.0 power supply from a good brand? If so could you reccommend some from the UK at a reasonable price. I hadn't budgeted for this expense.

I'll try what you said tonight once I get back to college.
I hope that is the problem, it'll save me from buying an expensive PSU!


<P ID="edit"><FONT SIZE=-1><EM>Edited by SeanTraynor on 01/10/05 08:12 AM.</EM></FONT></P>


Jan 9, 2005
The error that I got just recently from Half Life 2 Deathmatch was:


The solution, it says there, is to update with the latest drivers (which i'm on) or replace your graphics card (I think not!!)

Oh and I tried the fan idea, this computer is definetly not overheating.<P ID="edit"><FONT SIZE=-1><EM>Edited by SeanTraynor on 01/10/05 11:25 AM.</EM></FONT></P>


Jun 3, 2004
I would try to troubleshoot the PSU... I have heard of guys in this forum going into there local comp store and asking to test a PSU. You could try that... bring in your system and just ask to test one of the PSU's out. I have never tried this myself but I have heard a few guys say they have tried it. It could save you alot of time troubleshooting. That is a pretty "hefty" system for a generic 400W PSU.


Mar 18, 2004
Someone with this same board, or another Asus PCI-E board, called their tech support on using his older atx PSU with this board, and was told a video setup like this wouldn't work without the 24pin psu(the extra 12v lines, to be specific, at a guess). Take it for what it's worth.