Can not compare fps to a x1900 card.. The x1900 is a beast in comparison in some games.

meljor wrote on 2020-07-01, 18:14:

Can not compare fps to a x1900 card.. The x1900 is a beast in comparison in some games.

The point was not to compare a 7900GS to an X1900, but to see the performance difference between the D820 and the X6800 CPUs in an environment that was not bottlenecked by the video card which, in this case, happened to be 2 X1900 in CF .

ENunn wrote on 2020-07-01, 06:13:

EDIT: Did a fresh format and clean install (needed one anyways), no difference. Something's wrong. My PSU is more than enough to power my GPU, my CPU should be fine. I'm stumped.

I was going to suggest that there might be something running in the background causing problems, but I suppose if you did a clean install, that wouldn't be the case.

Have you tried disabling other things in your device manager, like your network card? Or do you have any kind of power management settings enabled by default?

candle_86 wrote on 2020-07-01, 16:18:

Check your gpu speeds with Everest and check fan speeds, if the 7900gs has a fan failure or can't ramp up it defaults to 2d clocks, also change the 14 year old tim.

Those cards will stick to 2d clock if they overheat, don't get enough power, fan doesn't spin right.

Nothing is showing up in Everest, it's completely blank. My GPU fans are pretty loud though so I can't tell if it's the fan or not.

About the tim, it's the thermal paste yeah? Last time I changed it was a couple months ago. I just changed it again just to be safe, barely any difference. maybe 3-5 fps? Still some major dips in the 40s. Most of the time there's no difference though.

Jorpho wrote on 2020-07-01, 18:50:

I was going to suggest that there might be something running in the background causing problems, but I suppose if you did a clean install, that wouldn't be the case.

Have you tried disabling other things in your device manager, like your network card? Or do you have any kind of power management settings enabled by default?

Don't know how the built in network card can cause any problems but I've disabled that and there's no difference.

The best power option for XP is always on yeah? I can't find anything else about any other power settings, BIOS doesn't really help either.

(there's another setting but I ran out of attachments 😜)

Here's maybe a dumb question... did you install the chipset drivers for the motherboard?

Download ntune and use it to check temps and clocks. Maybe it is sticking in 2D clock speed which would explain it.

Maybe force it to 3d clocks and see if that helps.... unless it is overheating and then you don't want to do that.

cyclone3d wrote on 2020-07-01, 21:02:

Here's maybe a dumb question... did you install the chipset drivers for the motherboard?

Yep.

cyclone3d wrote on 2020-07-01, 21:02:

Download ntune and use it to check temps and clocks. Maybe it is sticking in 2D clock speed which would explain it.

Maybe force it to 3d clocks and see if that helps.... unless it is overheating and then you don't want to do that.

Where do I change that? I can't find anything about that.

Just tick the "Custom Clock Frequencies" and then the Test Button. I am pretty sure that by checking the Custom Clock Settings, it should lock it at that speed but it has been a really long time since I overclocked one of those cards.

Seeing those pics, it makes me wonder if the 7900GS even has a 2d clock.

You could also up the fan speed to 80% to see if that helps. idle temps look ok to me but unless it has more than the stock style cooler, it may not stay that way when running a game.

Maybe an older version of RivaTuner would be better.

Really just trying to verify that everything looks like it is running at the appropriate clock speeds.

cyclone3d wrote on 2020-07-01, 21:44:

Just tick the "Custom Clock Frequencies" and then the Test Button. I am pretty sure that by checking the Custom Clock Settings, it should lock it at that speed but it has been a really long time since I overclocked one of those cards.

Seeing those pics, it makes me wonder if the 7900GS even has a 2d clock.

You could also up the fan speed to 80% to see if that helps. idle temps look ok to me but unless it has more than the stock style cooler, it may not stay that way when running a game.

No difference 😒
I even OC'd to 500 core/690 memory and upped my fan speed to 100%, literally no difference

Also just tried Far Cry , 1024x768 high, framerates not that hot, around the same as HL2 when outside + I'm getting missing textures sometimes but it might be because I'm loading an old save.

Wait grab gpuz, Everest home is to old to read your hardware, the latest home release is 2004, your hardware is newer, all of of it is. Check in gpuz. I dunno why I didn't think of that, I use aidia64 for XP and up because I have a license for it.

candle_86 wrote on 2020-07-02, 00:57:

Wait grab gpuz, Everest home is to old to read your hardware, the latest home release is 2004, your hardware is newer, all of of it is. Check in gpuz. I dunno why I didn't think of that, I use aidia64 for XP and up because I have a license for it.

Ah, here you go.

Woah 100c on the CPU? I hope thats wrong, the PC would probably be on fire by now haha. I can't trust the CPU readings on here either, it hasn't gone down or gone up, not even a by 1 degree. I wonder if it's another one of those 0c type stuff with the other programs.

Oh... is the CPU fan going max speed? In that system it would be really, really loud. This could happen if the heatpipes in the cooler died. Is the base of the cooler really hot and the rest cool?

Could also happen if the cooler is not seated properly.

If it really is running that hot, then it will be throttling like crazy as well.

Download the Dell diagnostics, burn to CD and boot to it.
https://www.dell.com/support/home/en-us/drive … ?driverid=y1rk6

The Dell diags will tell you the correct temps... I think.

If the CPU temp is not the problem:
What is the CPU load when in HL and Far Cry? If it is 50% or more, I bet it is maxing out one of the CPU cores. The only other thing I can think of is that you have v-sync and triple buffering enabled.

Rivatuner should be able to log what is going on.

Nothing else makes sense unless the PCIe slot is a gimped slot that is really only a PCI slot or if the PCIe slot is only running at x1.. but GPUz says it is running at x16 so that shouldn't be the problem.

Also check in CPUz.

100c would definitely cause the CPU to throttle back to stop itself from burning up, I would get that sorted out and try it again

cyclone3d wrote on 2020-07-02, 02:11:

Oh... is the CPU fan going max speed? In that system it would be really, really loud. This could happen if the heatpipes in the cooler died. Is the base of the cooler really hot and the rest cool?

Nah, it's pretty quiet. The only thing that's louder is the GPU fan. The CPU fan isn't dead, it's getting plenty of cool air. Wouldn't I smell burning? Because I can't smell anything.

cyclone3d wrote on 2020-07-02, 02:11:

Could also happen if the cooler is not seated properly.

It's pretty tight so I doubt it.

cyclone3d wrote on 2020-07-02, 02:11:

If it really is running that hot, then it will be throttling like crazy as well.

Wouldn't be suprised if that was the case.

cyclone3d wrote on 2020-07-02, 02:11:

Download the Dell diagnostics, burn to CD and boot to it.

The Dell diags will tell you the correct temps... I think.

I can't find anything about temps, all I see is Thermal State - Safe. So I guess that's good? No errors either.

cyclone3d wrote on 2020-07-02, 02:11:

If the CPU temp is not the problem:
What is the CPU load when in HL and Far Cry? If it is 50% or more, I bet it is maxing out one of the CPU cores. The only other thing I can think of is that you have v-sync and triple buffering enabled.

Rivatuner should be able to log what is going on.

I doubt it'll show anything. No temps in Rivatuner.

I don't have either v-sync or triple buffering on.

cyclone3d wrote on 2020-07-02, 02:11:

Nothing else makes sense unless the PCIe slot is a gimped slot that is really only a PCI slot or if the PCIe slot is only running at x1.. but GPUz says it is running at x16 so that shouldn't be the problem.

Also check in CPUz.

Don't see anything about any CPU temps in CPU-Z but here's the CPU tab.

ninkeo wrote on 2020-07-02, 02:27:

100c would definitely cause the CPU to throttle back to stop itself from burning up, I would get that sorted out and try it again

I wonder if it says it's reporting 100c even though it isn't really 100c, which is causing throttling? If it's true I don't know how to fix that.

Ok, check the memory tabs, and also run the bench test on the bench tab.

Looks like HWMonitor is supposed to support temp monitoring on Pentium D... or at least that is what somebody said when another person asked about temp monitoring with CoreTemp.
https://www.cpuid.com/softwares/hwmonitor.html

cyclone3d wrote on 2020-07-02, 03:18:

Ok, check the memory tabs, and also run the bench test on the bench tab.

cyclone3d wrote on 2020-07-02, 03:18:

Looks like HWMonitor is supposed to support temp monitoring on Pentium D... or at least that is what somebody said when another person asked about temp monitoring with CoreTemp.

Nope.

I was cleaning the CPU cooler and I found this. I don't know if this is a bad thing or what.

ENunn wrote on 2020-07-02, 03:45:

2020070122422286.png 2020070122430881.png […]
Show full quote

cyclone3d wrote on 2020-07-02, 03:18:

Ok, check the memory tabs, and also run the bench test on the bench tab.

2020070122422286.png
2020070122430881.png

cyclone3d wrote on 2020-07-02, 03:18:

Looks like HWMonitor is supposed to support temp monitoring on Pentium D... or at least that is what somebody said when another person asked about temp monitoring with CoreTemp.

Nope.
2020070122445058.png

I was cleaning the CPU cooler and I found this. I don't know if this is a bad thing or what.
IMG_20200701_223107.jpg

Hard to say from the photo, but to me the plastic looks warped .

darry wrote on 2020-07-02, 04:00:

Hard to say from the photo, but to me the plastic looks warped .

I don't feel any warping, and if there is I'm not 100% worried about it, I'm more worried about the copper heat pipes. Some of them look like they've broke off on the top.

edit: nevermind, seems to be a common normal thing

ENunn wrote on 2020-07-02, 04:03:

darry wrote on 2020-07-02, 04:00:

Hard to say from the photo, but to me the plastic looks warped .

I don't feel any warping, and if there is I'm not 100% worried about it, I'm more worried about the copper heat pipes. Some of them look like they've broke off on the top.

edit: nevermind, seems to be a common normal thing

The heat pipes look normal to me too . What I saw in the plastic is probably just an artifact of the injection moulding then .

Ok what I'm starting to think is your cpu temp sensor is broken, can you swap in another cpu.

Yeah.. that is just where they crimped them closed at the ends.

Ok, so me thinks that the CPU is the problem and it is running way throttled.... or the cache on the CPU is not working or something.
Did you ever check to see if the base of the cooler was really hot when the computer is turned on for a bit?

Looking at the CPU-Z bench listings, a Pentium D 3.0 Ghz is supposed to score about 10x as much as you are seeing on single thread and about 15x as much as you are seeing on multi-thread:
http://valid.x86.fr/bench/uize5m/1