Inadequate PSU from factory?

I would say you issue is related to the PSU, as most OEM manufacturers don't use Tier 3 or higher PSU, that run at their rate strength. Since you are stressing the system with your games, it is drawing more power than your PSU can handle.

Another avenue to monitor is temperatures. I would use a program such as HWMonitor (free download) to make sure you are not running on the hot side for your CPU or GPU, which can cause similar effects as a faulty PSU.

Thanks, I'll try downloading that tonight. Will it record sys temp info while i'm running the games?

No, they don't use power supplies based on any tiered list, they do however use good quality oem unit's that are true to thier ratings.

He's right. They're usually good quality PSU's (apart from Gateway/Emachines' PSU, those things seriously give out quickly).
The only beef I have with them is they barely pack the bare minimum for systems. So say for example you wanted to add a mid range to high end dedicated GPU, you would need to change the power supply or risk blowing it in a few minutes. When pushing your system to maximum load you can very easily put the power supply at hand to the loads they were probably tested under which is fine but a clear sign that they would need to be swapped out if your adding components with high power consumption.


@Original poster:

I was trying to look at your specifications online to see if the HP Pavilion Elite e9180t came with a HD 4850 pre installed.

I may be mistaken as this is taken from the HP site but did you add that HD 4850 yourself because it's not listed as one of the upgrade options?

@Atheist:

I ordered it August'09 and I came with the HD 4850. Now on their site it looks like they changed it to an HPE 180t. Looks like the same deal, but maybe their graphics card options have changed?

sorry, IT came with the 4850.

As mentioned earlier, monitor the temperatures with HW Monitor. The temperature your looking for is CPU. Run Prime95 under maximum FFT, monitor the temperatures for a good 15 minutes. Let us know the temperatures.


During the tests I post here, use the same components that came with the PC, apart the PSU. Do the tests in order, if you havent found the problem causing it, proceed to the next step.

1. Use a different power supply to test. If you don't have another to test with go to Fry's and pick one up. Return it once your done. If the PC does not crash, then you need a new PSU. Contact HP if your still under warranty, otherwise buy one your self.

2. Remove half of the ram, and test the first half. Then the second half. (So test 4gb, then another 4gb). If it crashes with one set of the RAM modules, and not the other then you need to find out which exact module it is. You can do this by testing 1 memory module at a time. Contact HP if your still under warranty, otherwise buy a new module or remove that particular memory module and run with 6gb of ram.

3. Remove the HD 4850. Run Prime95 blend test while watching a movie/copying files from the hard drive. (This will help test everything). Do so for the amount of time it usually takes to crash on you, and then add 30 minutes to that time. If the PC does not crash it could be the HD 4850 and I would recommend using it another adequate system with a good enough Power supply to test it.

Thanks guys. I'll try all of the above and see what happens.

Idle core temps for the CPU range from 39-42 C over the 4 cores.
Ran the 64 bit v25.8 of Prime95 torture test at 4096FFT and about everything under. Max core temp reached was 69 C. GPU reached 62 C.

Power consumption of CPU peaked at 147w and held this number for the duration without fluctuation. Used 43-44w idle.

Airflow in/out of case seemed good.

System passed all Prime95 tests that it ran in the last hour.

I don't really have the know-how to attempt removing memory or graphics cards, but I assume the results of Prime95 test eliminated the PSU as the culprit. Would that test have stressed the system as hard or harder than running top end games?

Prime95 will not stress your PSU like a 3D application, since it puts no real stress on the GPU. You can download 3dMark Vantage to stress your GPU to see what happens. If you have the same issues, it points more to PSU or GPU.

You are correct...  

What I wrote didn't come out as the same as what I was thinking. AsAnAthiest explained better what I met to say, when quoting the same post...

The first thing I wanted him to do was check out the temperatures of his CPU (someone beat me to it but I decided to reinforce that advice).

To come as close as to stress testing your power supply apart from building a homebuilt PSU-stress equipment that will deliver the wattage/amperage to the maximum of the PSU's specifications.
1. Prime95 with maximum FFT/maximum memory (requires a custom profile)
2. Run GPUTool to stress the HD 4850, use the "check for stability" option.
3. Run disk transferring to stress the hard drive, Copy/paste like 20gb+ of files.
4. Play a DVD.

Apart from this, very few 3d application can actually stress test every component. The only games I can think of that can come close might be Crysis or GTAIV with max settings.

@Original Post/er:

The temperatures are fine for a i7, the Puget system recommendation is below or at 80c as the i7 has a damn well designed manufactured process. No the stress test did not rule out the PSU, only thing that can RULE out a PSU is to use a different PSU of equal or higher performance or to stress your CPU with the above mention method (this is not recommended in your case as sometimes when PSU's blow they can take some other components with them)

Please proceed with the list of procedures to figure out what the malfunction is do not stress your PSU with my method above as if it is the PSU it may blow (no not explode but blow capacitors, or fuses).

I did run the HWMonitor during gameplay to see if I could look at the data points when the malfunction actually occured. During gameplay, when the stuttering began I quickly minimized and looked at temps. GPU was 89 C. This same malfunction has occurred with GPU temps in the low 80's as well though so not sure that overheateing of the GPU was the cause.

I do sometimes (maybe 60-70% of the occurrences) get a blue screen error message about a clock interrupt on a secondary processor not being recieved during the allocated interval. Any ideas about this? As possible solutions it suggests disabling BIOS memory options like caching and shadowing and uninstalling any recently installed hardware and or software. As this problem has occurred since I first bought the computer, I don't think it was anything that I added.

Also, trying to access setup during start up to look at BIOS settings has proven to be a bit of a challenge...

Thanks for all of the help guys! I work for a living and don't have time to sit at home on the phone with tech support very often, so all of your suggestions are heartily appreciated!