I need a program that can monitor my system temps.

[Occulus]

It must be:

1) free.

2) NOT Rivatuner. (it no longer does what I need it to do)

3) Completely comprehensive; i.e., CPU, GPU, RAM, etc.

Does anyone know what I should use?

edit: I've run system stress tests like Prime95 and a couple others, but I keep getting bluescreens during gaming when the system gets hot. I've examined and cleaned the heatsinks, my airflow is good, and the hardware is fairly new. I can't track down the problem and I will not pay to have the system diagnosed.)

[hobbit709]

1. I use Core Temp to monitor my core temperatures.

I don't know of any that will monitor RAM temps without an external sensor.
http://www.alcpu.com/CoreTemp/
there are 32 and 64 bit versions

this one also works
http://efmer.eu/boinc/

[Occulus]

2. Thanks

I swear my PC is mocking me. It only BSODs in games, when I've JUST completed a hard section of whatever game.

This will help. Thanks again!

[Gore1FL]

4. Is it a laptop or a tower?

It sounds like an airflow problem.

[Occulus]

6. A tower.

The airflow seems fine... Maybe I just need a better case.

[Gore1FL]

8. It may be that the heatsink is clogged with dust.

I've had to remove the fan and scrape crap out of min in the past.

[panzerfaust]

15. ME Too

It is stable and without malware

[M0rpheus]

3. Sounds like your Video Card may be dying, or overheating.

I know you said airflow is good, but the issue sounds the same issue I had with my gaming laptop.

2D applications (Desktop, webpage viewing, etc..) were fine. But as soon as I did anything that stressed the video card past those norms I'd get a bluescreen/shutdown.
The culprit was a dead fan.

If the only time you have an issue is during graphics intense applications, I'd put money on some combination of Vcard/Fan/airflow as the issue.

I used CPUID to track down the issue
http://www.cpuid.com/softwares/hwmonitor.html

*Edit: CPUID doesn't include RAM temps (you'd need a probe or some other hardware to get that, as the prev poster noted), but you've mostly eliminated that as an issue with Prime95 (IMHO).

[Occulus]

5. Well, here's a screenie after playing Diablo 3 for about ten minutes or so

Sorry for the large image, if it's too big:



Those CPU cores SHOULD NOT be reading 0 degrees. That much I do know. After I post this I'll open up the case and check to see if I forgot to plug in something like a CPU sensor, or if it got pulled out a bit.

When the PC freezes, the hard drive thrashes around for about thirty seconds to a minute (solid red light), the system going unresponsive during that time, and then BOOM- BSOD. Now, I just recently reinstalled Windows (everything seemed fine afterward), but then, bit by bit, the BSODs are becoming more common, WITHOUT file loss or corruption. I thought it was my hard drive, so I ran a couple utilities on on each to check for bad blocks/sectors, without any luck. My temps seem to be in range, and all my fans are running. The only thing I haven't been able to check is the RAM (Memtest86 won't boot from my stick for some reason), but that doesn't seem to be the problem anyway.

The only thing I can think of is that I also recently reapplied a layer of Arctic Silver to my CPU when I blew out/cleaned the heat sink (cats and cat hair really don't mix with heat sinks!). Could I have damaged/bent a pin on the CPU? Could that be the problem?

It ticks me off because I just built this PC not that long ago, and I should not be getting BSODs at this point. It would make me feel a lot better to know which piece of hardware is becoming ill, because I just can't afford to randomly replace components until the problem disappears... but every utility I've tried to use to track the problem down is just smiling and saying "everything's fine, pal!"

My bet's on a failing hard drive, but that's only an educated guess at this point.

[M0rpheus]

7. Is your Phenom unlocked? that appears to be the issue with the core temps...

If it is unlocked, is it OC'd?

All the temps I can see look to be within range (I think the Phenom tops out @ 65C). The temps above your CPU will likely be sensors close to/under the cpu that you can use for reference.

The hard drive could be the issue,but that's a potentially costly (as opposed to free) fix, in time and $, unless you have a spare kicking around.
I usually start with the free options first. Since you mentioned the arctic silver, is it possible that the heat sink is not seated properly after you replaced it?
I would also try reseating your ram and any cards installed in the box.

My tablet is about to die so I'll cut it here. Let me know how the simple stuff works out before you go throwing parts at it.
Back to check on you in a bit.

[Iterate]

9. Find a S.M.A.R.T. reader

I found a problem with a client's system (XP/single processor) a few years back that sounds very similar. In that case, the HD was failing at a slow rate and corrupting the pagefile.sys and some other often used files first. An OS repair would just postponed the failure by fixing a few files and reassigning the pagefile sectors. Not long after fixing that one, several others at the same place and of the same manufacture and date did the same thing.

In all of those cases a sector scan didn't show a problem.

I use SIW on a stick, just because it's an all purpose tool. There are more compact ones that even sit quietly in the systray so that you can monitor in real-time, but the last time I used one was two years ago with a neighbor kid's laptop. I haven't heard back from him, so I assume the SMART R/W errors that I did see were a one time problem. I remember that laptop well because it was such a week-long freeby mess, but I don't remember what SMART reader I used. It probably avoided an unnecessary HD replacement.

In that case a sector fix wouldn't complete. As near as I could ever tell, he was chat/surfing in bed during a concurrent vista/virus program(he had five of them installed)/adobe update, fell asleep, and woke up with the laptop on the floor. The registry was mashed and he had no backups. Sometimes though you can never know the real source of the failure.

Plus, which BSOD errors are you getting? Apparently random, or the same one?

[hobbit709]

10. Crystal Disk Info works for me.

http://majorgeeks.com/CrystalDiskInfo_d5931.html

[grok]

12. have you considered a power issue?

AMD cpu's suck up alot of power when running hard. But I'd rather suspect the Graphics card. you have a high performance one.

I've had a an instance where i was using Nvidia CUDA cores to do some intense computation and was writing to my dvd drive in tiny chunks all while doing some serious file compression on the hard drive..

Even though I have a 750 watt power supply, it just couldn't take it and shut down briefly. It was in my Linux logs.

Anyways, not all power supplies keep on providing "clean" power when stressed. Might start spiking.

as far as your cpu core temps, is ACPI enabled in the bios?

[panzerfaust]

16. No bent CPU pin - else it would not run (that has been my sad experience)

What I would do (I built my first PC in 1982, my last in 2011, getting ready to build my next now):

Open the case:
Unplug every single connector and board.
Blow out all the connections with compressed air.
Vac all the fans and filters (vac not good on MB - can suck off components).
Make sure the fans are really running!
Reassemble.

No Joy?
Well, you did mess with the CPU so ...
Pull the CPU
Remove ALL of your new heatpaste (99% Isoprop Alcohol works fine)
Put ONLY a SMALL DOT of heatpaste on the CPU.
Reassemble.

Still no Joy?
Try running with the minimal amount of RAM the sys will boot with,
only the boot hard-drive connected,
and ALL add-in cards removed.

Yet still no Joy?
Think power supply.

Simply swap it out with a good one,
OR buy a PS tester (which one should have anyhow)
I use a PS tester called "Ultra" which I bought on Amazon for 15 bucks several years ago (and which has saved me many hours and $$$ since then. There are several for about that price range (search 'Electronics' power supply tester computer)

If that all fails, then try temporarily replacing the boot drive with another one - preferably one running with a different OS (LINUX is free) which - if that works - points to the need to, once again, reinstall windows, or to replace the HD.

If none of this has worked - likely the MB is Mort - though one could try wiping the CMOS, and then, if still inop, replacing the BIOS before tossing the MB.

[Occulus]

20. It was the hard drive

I heard the read head impact the plates twice late last night (I hate hate hate that sound). Then it coughed and died for good.

Put in a new .5Tb Seagate Barracuda HDD today- I've been running for a few hours now, without any BSODs. The old HDD was manufactured in 2006, so I guess it was just at the end of its life.

[davepdx]

11. HWiNFO is another program to try

http://www.hwinfo.com/

Click on the "Sensors" button on the main window and you'll get the obvious sensor data.

I do remember reading a while back where someone cautioned about using more than one sensor program at one time. The caution, as I remember it, was that some sensor programs can interfere with other sensor programs that are run at the same time.

[Occulus]

13. Thanks, everyone. Still tracking down the issue

but all of this is good advice!

[davepdx]

14. Are you using an updated video driver for your video card

by chance? Have you tried a recent, though different version of your video card driver to see if the driver might be the issue? Several years ago I had a problem on my system with a video driver that caused intermittent bluescreens only when I was gaming. Just a thought.

[panzerfaust]

17. Would simply pull any discrete graphics card out

and use the onboard video whilst trying to sort out systems issues.

But, you are right, graphics drivers can wreck havoc - however if the system was stable, and no new drivers were installed recently it is unlikely to be a driver issue.

My experience with drivers of all types is that if I am not having a problem, or missing on performance, for a given component, then I do not update drivers.

[davepdx]

18. Occulus was having the blue screens only while gaming...

In general I'd agree with your point regarding pulling out the GPU and using the integrated graphics in the situation where only 2D work was being done. The OP indicated that the blue screens were only occurring while gaming. Using an integrated GPU (if actually present) might not be feasible for gamers because of the resulting poor graphics performance.

As a gamer I've encountered situations where game A and game B work fine but I want to get the new, just released game, game C, and it requires a new video driver in order to take advantage the great, new video effects in the game. When using the new video driver game C works as expected but coincidentally game A's framerate drops significantly. Using new/updated video drivers that produce better game performance and unexpected side-effects is something that gamers have to deal with when on the "upgrade video driver train" unfortunately.

If the GPU sensor readings in post #5 are accurate then I don't think the GPU is too hot at all. The GPU temps are quite good actually. I am thinking that Occulus probably ran the sensor program while he was gaming (and not after exiting the game) so that the Max temp values are valid. The fairly low GPU temps in my mind eliminated the heat issue and lead me think of the video driver issue. I was reading a series of posts on an overclocking forum today where one particular sensor program was indicating an air cooled CPU temps that were less than ambient air temp (which can't happen). Some of these sensor reading programs can provide flakey results at times so Occulus, imo, was wise in trying to find an alternative sensor reading program. No telling what the CPU temps really were - and those temps are key.

[Occulus]

19. Welp, problem solved. 'Twas the hard drive,

which coughed and died late last night after the read head twice impacted the plates.

I've installed a new one now, reinstalled windows, and all is well again. Thanks, everyone, for all the advice!

(Steam and Diablo 3 are still there. I make it a matter of course to keep my system and my programs and games installed on different physical drives, so this is more of a pricey irritation than the complete disaster it otherwise would have been.)

[Dash87]

21. I use CPU-z hw monitor. Great and simple program.