Sign in with
Sign up | Sign in
Your question

X1800XT Can't finish 3DMark05 demo

Tags:
  • Graphics Cards
  • Graphics
Last response: in Graphics & Displays
Share
December 1, 2005 1:08:00 PM

it runs fine, till you get to that airship and water dragon. It runs it fine the 1st time but then tries to run it a second time at 1 frame per second and just freezes the whole computer. as best I can tell the vid card is sending a temperature reading of 81 celsius to the ATI control panel I have in the background running. When 3Dmark loads a game I can sneak a peek at the temp of the video card if I position the window just right.

More about : x1800xt finish 3dmark05 demo

December 1, 2005 1:10:00 PM

Post the temps, I would be interested.

Assume you have good airflow in case, what is cpu temp?
a b U Graphics card
December 1, 2005 1:10:33 PM

It would seem that you are running into an overheating issue...

How well is the HSF seated on the GPU? Fan running?
Related resources
December 1, 2005 1:25:15 PM

my case is uber cooled, fanwise. I have the Antec P180 case. It has 2 180mm fans bringing air in and 2 180mm fans taking air out. And to boot, it has an 80mm fan blowing right onto the graphics card itself. The heat sink and fan on the card itself appear to be seated firmly.

If you put your hand behind the graphics card's vent you can feel the hot air rushing out, and rather fast. So the fan is working and running. It is getting air. And it is getting hot. Apparently too hot. Something here is amis.

p.s. idle temps for the vid card itself are 57-60 celcius. CPU = 29 and the mobo 30.
December 1, 2005 1:36:43 PM

ran the test again, but when it repeated the airship test (why does it repeat that one?) and it was going at 1 FPS I aborted it. Within a second I recorded the temp on the vid card at 71 celcuis. not in the hot zone. The CPU was 40 celcius. The mobo was still at 30 celsius the whole time because of all my fans. I can't figure out why it runs thhat ONE test twice.

?
a b U Graphics card
December 2, 2005 6:07:30 PM

Are you running the benchmark or the demo?

The demo wouldn't be a VPU issue as it's a movie of the benchmark.

The benchmark portion you are talking about is the CPU tests.

Check your system RAM.
December 3, 2005 1:31:43 PM

i did check the ram. in fact, my two sticks show errors on test 5 of memtest. I am quite upset about it. So I took one stick out and ran memtest and it seems fine. With that single stick I was still getting the same problem. It happens whether I hit benchmark or demo.
December 3, 2005 4:07:10 PM

it could be your cpu heatsink seated incorrectly.
December 3, 2005 4:23:51 PM

but the cpu temps are showing within tolerance. I mean, no higher than 40 celcius. Also, I put my finder on the heat sinks, after grounding myself, to feel if anything is too hot. nothing is. I can only touch the underside of the vid card . . . the metal X-bracket that supports the HSF on the other side and it is warm to the touch, but at no point did it get hot enough where you would have to pull your finger away. My old Radeon 9700 pro DID get that hot.
December 3, 2005 4:33:37 PM

I have the free trial version. it is the newest one and doesn't nee the patch. I hit "Run 3D Benchmark.

What I am wondering is why it runs that particular benchmark twice in a row? The airship and dragon? It runs fine the fisrt time then it reloads itself in a crappy resolution and then goes a blazing 1 frame per second!!!! It's called the Canyon Flight test and listed as a "game test" as opposed to a "CPU" test.
December 4, 2005 3:14:33 AM

the 2nd test is using software rendering thats why it looks so ugly.
December 4, 2005 12:42:31 PM

why do YOU think my system is freezing/hanging at that point???
December 4, 2005 12:50:07 PM

I would think 1 of 3 things


1. PSU, can you play games? have you tryed to play games?

2. RAM, try putting your ram in different sockets...find a mix that works (thats how you fix memtext test#5...lol try the other Dual channel slots)

3. What do you have running as background tasks? list EVERYHING you have running? someimes weird prgrams will stall....i have had problems with F@H and some Norton crap messing up my games
December 4, 2005 5:13:45 PM

Quote:
I would think 1 of 3 things


1. PSU, can you play games? have you tryed to play games?

2. RAM, try putting your ram in different sockets...find a mix that works (thats how you fix memtext test#5...lol try the other Dual channel slots)

3. What do you have running as background tasks? list EVERYHING you have running? someimes weird prgrams will stall....i have had problems with F@H and some Norton crap messing up my games


1. I can play Morrowind and Battlefield 2 till the cows come home, though at times I do get freezing in Morrowind, but that game is notorious for that. My particular PSU was very powerful for its day and since it is not powering 2 graphics cards it still is. I have only added a RAID 0 array and 120 mm fans to my system.

2. I am down to one stick of RAM that passes memtest86. It is in slot 1.

3. taskmanager has no running applications when I try and run the benchmarks.


if I can scrape up $150, I can test the RAM by itself. I can probably test the CPU too, but I cannot test the video card per se because this is the only mobo I have that has PCI-e.
December 5, 2005 1:12:35 PM

Quote:
3. taskmanager has no running applications when I try and run the benchmarks.
Not applications. List the processes.
December 5, 2005 1:16:49 PM

My thoughts:
1) That graphics card sounds like it's running way too hot. I'm thinking RMA.
2) If this is an upgrade, did you fully clean the old drivers, switch to VGA drivers, and then install new drivers?
3) Are your mobo chipset drivers the latest and greatest?
4) Have you flashed BIOS?
5) Do you have a dualcore CPU?
6) Are you sure that your PSU is doing a good job?
December 5, 2005 2:34:43 PM

Quote:
My thoughts:
1) That graphics card sounds like it's running way too hot. I'm thinking RMA.
2) If this is an upgrade, did you fully clean the old drivers, switch to VGA drivers, and then install new drivers?
3) Are your mobo chipset drivers the latest and greatest?
4) Have you flashed BIOS?
5) Do you have a dualcore CPU?
6) Are you sure that your PSU is doing a good job?


1) I am not sure it is way too hot. I have to call ATI and ask them what is too hot. RMA is going to blow because I doubt they send me a brand new replacement.
2) since I got two new raptors, I can tell you that XP is a fresh install.
3) the chipset are the newest ones so far as I can tell. I got them from Asus's website for my mobo, but it would be a lie to say there isn't some confusion in looking at their downloads.
4) I have flashed the BIOS from DOS to one that is just below the only one that carries a warning. I flashed to (not that it will mean anything to you) 1006, but not 1007 because it says to update the Chipset drivers to a specific version which does not appear to correlate with the version numbers offered for my mobo.
5) no
6) PSU's do go bad, but mine should have enough juice to run my system. is it going bad? Tough to tell. At idle, I do not like some of my voltages. Here they are

3.3 = 3.18 idle
5.0 = 4.57 idle
12 = 11.52 idle

those seem low, but mu PSU carries 24 amps on the 12+ rail. Not great but not bad.
December 5, 2005 4:07:46 PM

Well it sounds like you've done everything right. :)  So that's always good.

(Personally, I'd have probably Ghosted my old HD contents onto the new HDs, but that's because I hate reinstalling a million programs. M$ should really have a utility to copy an install over from an old HD.)

That really only leaves two concerns. It seems like your card is running hot, and your PSU doesn't look too peachy keen. Maybe the card is allowed to run that hot though, I'd check with the manufacturer. But your PSU ... if those are idle voltages, then it's possible that with the right load your system might not have enough power. Do you have any software to log your voltages so that you can see if they reach some critical point right before your system fails? Have you tried torture testing your PC with software like Prime95? The PSU should be enough, I'd think, but maybe it's defective. You never know. (Especially when quality companies like Antec are lately having disturbing quality control issues.)
December 5, 2005 5:37:15 PM

Quote:
Well it sounds like you've done everything right. :)  So that's always good.

(Personally, I'd have probably Ghosted my old HD contents onto the new HDs, but that's because I hate reinstalling a million programs. M$ should really have a utility to copy an install over from an old HD.)

That really only leaves two concerns. It seems like your card is running hot, and your PSU doesn't look too peachy keen. Maybe the card is allowed to run that hot though, I'd check with the manufacturer. But your PSU ... if those are idle voltages, then it's possible that with the right load your system might not have enough power. Do you have any software to log your voltages so that you can see if they reach some critical point right before your system fails? Have you tried torture testing your PC with software like Prime95? The PSU should be enough, I'd think, but maybe it's defective. You never know. (Especially when quality companies like Antec are lately having disturbing quality control issues.)


i think I am going to try running Prime95. I'm interested in this ghosting concept you mention. Know of a good prog that does it? How much space you need? Will it prevent you from having to do a clean install of XP? I've got plenty of space for a Ghosting I think.

I have to call ATI today.

As for the quality issues of Antec. Boy tell me about it! I bought their top of the line PSU (the HE 550 watt) and the think just kept shutting off. Come to find out that many are having this issue too! It's so bad that I am sticking to another brand.

One curious thing I noticed today: My CPU should be running at 1.5 volts but it stays well below that . . like 1.28. I have Cool n Quiet enabled, but I Have yet to see those voltages go higher even though my CPU clock speed may reach 2.2 Ghz.
December 5, 2005 5:58:16 PM

Quote:
i think I am going to try running Prime95. I'm interested in this ghosting concept you mention. Know of a good prog that does it? How much space you need? Will it prevent you from having to do a clean install of XP? I've got plenty of space for a Ghosting I think.
Norton Ghost. Oh, sure, there are other programs that do the same thing, but I've been using Ghost for years. It's so nice to just copy over an old drive onto a new one, plug the new one in, and boot right up. :)  Plus it's good for backing up your hard drive.

Quote:
As for the quality issues of Antec. Boy tell me about it! I bought their top of the line PSU (the HE 550 watt) and the think just kept shutting off. Come to find out that many are having this issue too! It's so bad that I am sticking to another brand.
Yeah, I'm super pissed. I used to love Antec, but they're just royally f'ing up lately. It's gotten to where I don't know if I could even trust them with a new purchase. :o  But then so many people don't have any problems. :? It's so confusing.

Quote:
One curious thing I noticed today: My CPU should be running at 1.5 volts but it stays well below that . . like 1.28. I have Cool n Quiet enabled, but I Have yet to see those voltages go higher even though my CPU clock speed may reach 2.2 Ghz.
It could just be that your sensor isn't very accurate. Or that it's supposed to do that. But it could also be that your PSU is just undervolting your whole system. (Which it kind of looks like, since every one of your rails is under.)
December 5, 2005 6:20:35 PM

prime95 kept the CPU voltage to just below 2.5 so it must have been Cool n' Quiet dropping the voltage that low. so the CPU appears to be fine, though I never really doubted anything was wrong with it . . it's a real workhorse.
December 5, 2005 7:20:47 PM

ATI tech cupport says that 80 celcius is not dangerous. so it can't be the card.
a b U Graphics card
December 5, 2005 7:59:28 PM

Really I don't think it's VPU temps that are the issue. The point you are having trouble with is the CPU rendering stage of the test.

I would start running a few other tests (prime error tells you that's not good) like PCMark, etc.

I know the Antec PSUs were causing issue in the GF5800/R9700 era but that was for far lower power. However if the power draw is fluxating that may be enough.

As for morrowind, yeah annoyingly enough Morrowind runs smoothest on W98SE and Win2K, but is unstable with XP and ME (isn't everything unstable with ME?).

If anything's out of 'stock' settings then roll those back first before testing/tweaking for a fix.
December 5, 2005 8:11:18 PM

well, then I'll run more CPU benchies. as for the Antec, they make great PSU'sm but their 550 HE 20+4 pin PSU is very unstable. So much so that I am frankly shocked that Antec hasn't recalled them.
December 18, 2005 6:00:20 AM

hmm so what was teh problem?? I seem to have the exact same problem as you.

here are my specs.

AMD x2 3800+
XFX GeForce 7800GT
DFI LANPARTY nF4 SLI-DR
OCZ EL Platinum Revision 2x 512mb (1gb)
Antec TruPower 550W PSU
Western Digital 320gb SATA
Windows XP x64


mine also feezes at the second test of Canyon Flight.
December 19, 2005 4:48:57 PM

so far ATI has agreed to RMA it if I so choose, but I am not sure what they will do. I simply want them to run 3DMark05/05 on their setup and see if it runs. They felt, however, that since it ran games fine I shouldn't worry about it. My response was that it may point to a possible problem down the road.

Just curious, are you the guy that has been emailing me about this same issue? Also, what part of 3DMark05 does your card stop working?
a b U Graphics card
December 19, 2005 5:04:08 PM

Quote:
as for the Antec, they make great PSU'sm but their 550 HE 20+4 pin PSU is very unstable. So much so that I am frankly shocked that Antec hasn't recalled them.


Well I agree that they make 'good or great' PSUs depending on the scale. However PC Power & Cooling and Fortron make AWESOME PSUs and while the ANTEC ones are good, for enthusiasts I find they do have power issues in their past. I liked Enermax alot, but their quality has gone from Fortron level towars the upper middle of the pack. They still have nice features. Surprisingly enough OCZ has made an effort to get good filtering on their PSUs and they've taken the place of Enermax on my personal list.

It may or may not be an issue, but it ios enough of a concern to troubleshoot that as well. After running some other tests you may be able to narrow down the source (RAM, etc) hopefully that will do it. However considering the issue it could be so many things that it will take a bit of time to trouble shoot it. It doesn't sound like the graphics card, but it could even be something messed up with the mobo that's doing it.

Let us know how the tests go, hopefully they'll reveal something but I'm worried about those RAM results.
December 19, 2005 7:16:16 PM

Quote:
as for the Antec, they make great PSU'sm but their 550 HE 20+4 pin PSU is very unstable. So much so that I am frankly shocked that Antec hasn't recalled them.


Well I agree that they make 'good or great' PSUs depending on the scale. However PC Power & Cooling and Fortron make AWESOME PSUs and while the ANTEC ones are good, for enthusiasts I find they do have power issues in their past. I liked Enermax alot, but their quality has gone from Fortron level towars the upper middle of the pack. They still have nice features. Surprisingly enough OCZ has made an effort to get good filtering on their PSUs and they've taken the place of Enermax on my personal list.

It may or may not be an issue, but it ios enough of a concern to troubleshoot that as well. After running some other tests you may be able to narrow down the source (RAM, etc) hopefully that will do it. However considering the issue it could be so many things that it will take a bit of time to trouble shoot it. It doesn't sound like the graphics card, but it could even be something messed up with the mobo that's doing it.

Let us know how the tests go, hopefully they'll reveal something but I'm worried about those RAM results.

i have switched everything out and tried everything. the ONLY thing I cannot switch out is the MOBO because one is AGP and the other PCIe. Oh, and I even bought a power PC cooling 510 Watt PSU . . . nearly their flag ship model . . paid $219 for it. It is truly special. The old 550 antec is now running another machine. should be plenty. I have narrowed everything down to either the video card or the mobo. but as you have mentioned before, the point that the test crashes is what appears to be a CPU intensive test. Strange, I know.
a b U Graphics card
December 19, 2005 9:49:09 PM

Yeah that's weird, nice Buy on the PCP&C 510 they're better than most 650+ because the 510 is rated at 40+C not the 20C most other are rated at.

Yeah I don't know what else to tell ya' , it's a tough troubleshoot for sure, and the point at which it's causing problem is the reason why I doubt the VPU, since it's all software/system rendered there.

Very strange, and annoying I'm sure. See if you can get a store to pop the X1800XT into one of their PCIe rigs to test with 3Dmk and see what happens, that's the only way I can think of of ruling that out without another MoBo.
December 19, 2005 10:36:44 PM

yeah, I am so phyched about that PSU. I can't wait to push it in CrossFire mode someday.
December 20, 2005 1:34:10 AM

Quote:


Just curious, are you the guy that has been emailing me about this same issue? Also, what part of 3DMark05 does your card stop working?



no, i just found this thread a couple of days ago. i didn't email you.

I'm gonna go ahead a overclock my new machine and after thats done, i'll run the test again and see if it gets any better. I'll let u guys know how it turns out.

oh ya and also, have u ran 3dmark03?? if so, does that freeze too when u it gets to the cpu testing part?
December 20, 2005 7:28:56 PM

Quote:


Just curious, are you the guy that has been emailing me about this same issue? Also, what part of 3DMark05 does your card stop working?



no, i just found this thread a couple of days ago. i didn't email you.

I'm gonna go ahead a overclock my new machine and after thats done, i'll run the test again and see if it gets any better. I'll let u guys know how it turns out.

oh ya and also, have u ran 3dmark03?? if so, does that freeze too when u it gets to the cpu testing part?

yes.
December 23, 2005 10:20:44 PM

If the only thing your setup can't do is one part of 3dMark05 I'd suspect 3dMark05 is the problem. I've always had trouble with ATI cards when they were brand new like the 1800 is now. In a few months to a year all the drivers will be mature and all the games and benchmarks will have patches and everything will work just fine. You are on the "bleeding edge" with your 1800, expect some bumps in the road. I doubt you have a hardware problem and really doubt a heat issue if it is dying in the same spot every time.

My 7800 is running perfectly btw :twisted:

Sorry for the 7800 comment. I've had several cards form both manufacturers including nVidia: 4200, 6800 and 7800, plus on the ATI side: the original Radeon (which gave me trouble when it was new), a 9000-AIW (which was stolen), a 9600 pro turbo (which ran perfectly, no problems ever), a 9600-AIW and a x800-AIW.

Anyway just relax and enjoy your awesome card on the titles that it works perfectly on and in time when the get the bugs out 3dMark will work too.
December 23, 2005 10:37:37 PM

Ok it looks like 3dMark is dying during a CPU test that does not stress your video card one little bit. I think your video card is fine.
January 1, 2006 2:24:54 PM

I just ordered an MSI MOBO to see if the MOBOI is the problem. Cause now two things are happening that really concern me:

1) C&C Generals demo will crash shortly after starting a game.

2) Morrowind Propolyn Chamber swirly lights cause me system to crash.

#2 worries me most because it never used to happen. So it's like if something is failing, it is getting worse.

Frankly I think it could be the mobo because MY model was notorious for being shipped out with a broken North Bridge fan. I waiting to see if that was the issue for the wierd activity. So I got my new NB fan from Asus (which is a different model from the stock one) and installed it. That is not the issue.

The new MOBO I am getting is XPress200 chipset because I want to try a totally different chipset vs. the NForce4 chipset, as part of the troubleshooting process.
January 1, 2006 2:34:57 PM

Quote:
Ok it looks like 3dMark is dying during a CPU test that does not stress your video card one little bit. I think your video card is fine.


well, this problem happens with BOTH 3DMark03 and 05. So as part of the trouble shooting process, I already have 2 mobos to run the tests on but one is AGP but 3DMark05 will not run on it because it is a very slow ATI card (Radeon 9250 I think). Still, it however will run 3DMark03. So I stuck the CPU you are referring to into that computer and ran 3DMark03 fine. I can only conclude that I'd see the same results if that board could run 3DMark05.

I really don't think it is the CPU. Also, I do not (and never) OC. As a general rule. So if something is wrong with something OTHER than the vid card, the MOBO perhaps, then it could be a matter of voltage regulators or the like. It is NOT the PSU that is the problem.
January 2, 2006 9:15:30 PM

Quote:
prime95 kept the CPU voltage to just below 2.5 so it must have been Cool n' Quiet dropping the voltage that low. so the CPU appears to be fine, though I never really doubted anything was wrong with it . . it's a real workhorse.


below 2.5 volts from 3.3??? Problem there ... Cool n quiet drop voltage from CPU 1.45(or whatever you cpu version has) to something like 1.3 volt... There is nothing cool and quiet can do about PSU rails..

So, your PSU is failing to deliver what it should. Watt is not everything. Get something else to test, like a fortron or even an enermax..
January 2, 2006 11:01:21 PM

perhaps I made a mistake. AMD's dashboard shows that an nearly idle, the CPU voltage is 1.3v. When maxxed out, it is 1.5 volts. never the less, I don't think this is a cause of my woes. BUT, I shall disable CC and try 3DMark tests again.
January 2, 2006 11:26:05 PM

C&C disabled. test still failed.
January 3, 2006 12:43:02 AM

I knew that.. CnQ is part of the CPU DRIVERS.. it has nothing to do with PSU..

Anyway, try another PSU just for fun..
January 3, 2006 1:46:56 AM

Try checking your catalyst driver steeings. Did you overclock the card? I had the same problem when I overclocked it. Also, did you plug the video card to the power supply? There should be a special cable that came with the video card which plugs in to the top of the video card and goes to the power supply. This needs to be plugged in. Good luck!
January 3, 2006 10:56:23 AM

Quote:
Try checking your catalyst driver steeings. Did you overclock the card? I had the same problem when I overclocked it. Also, did you plug the video card to the power supply? There should be a special cable that came with the video card which plugs in to the top of the video card and goes to the power supply. This needs to be plugged in. Good luck!


it is not overclocked as I am not in the habit of doing that . . yeah I know I am a dork. And it is connected to the PSU by the Aux port you mention on the card. My PSU is SLI-ready, so the ATI card gets one of these plugs . . four or six pin . . I can't remember. It's not the same 4 pin plug that gets plugged into the MOBO near the CPU, but it looks similar.
January 3, 2006 11:00:14 AM

Quote:
I knew that.. CnQ is part of the CPU DRIVERS.. it has nothing to do with PSU..

Anyway, try another PSU just for fun..


well, I did that already . . which sparked me to buy the one you see in my sig. I tried it with an older Antec 550W. It was only a 20 pin, and to be even-steven, I bought the PSU you see in my sig. But the old Antec had 24 amps on the 12v rail. Pushing it, yeah I know, but . . .

anywho, it's been tested with another PSU.
January 3, 2006 11:12:09 AM

did you try to run with only one stick, take the memory voltage to 2.7 or 2.8, try to unplug either the array or the individual drive(keep only the bootable drive) tor give some slack to the PSU and whole system?
January 3, 2006 12:39:09 PM

actually, I am using only one stick. lol (*I'll edit my sig)

this MOBO gives me errors with both sticks in memtest, so I use only one stick. On an older board, the same two sticks and CPU pass memtest . . go figure.

also, I believe the voltage on the Ram is 2.7.

I can't unplug the Array because WinXp is on it. I can unplug my other 2 IDE drives though. But, let us assume 3DMark ran, then what would I assume? That my PSU is too weak to handle 2 SATA drives and 2 IDE drives? For a top of the line PSU that would blow chuncks!!!
January 3, 2006 8:21:22 PM

My cousin has a computer. With one PATA drive and one SATA drive. then, few day after he put his SATA HDD, the computer refuse to post anymore. Then, he removed the last hardware parts he put in, the SATA hdd.. the computer started right on.. We tested the SATA hdd on an other computer and it was working good.. then, a few day later, the cmputer wouldnt post at all.. even with all hdd removed..

The PSU was failing.. put a new one and back in business..

Did you try it out of the case?
January 3, 2006 10:59:54 PM

try what out of the case? which component?
January 3, 2006 11:03:35 PM

All that it is inside and necessary for running 3dmark05
      • 1 / 2
      • 2
      • Newest
!