Upgraded RAM now CPU is running at or close to 100%

[rakosnik]

I upgraded the RAM in my HTPC from 4 GB to 16 GB and now the processor runs at or close to 100% at times for no easily understood reason. Before the upgrade the processor rarely ran over 50%. I've stared at the task manager and I don't consistently see any particular process monopolizing the cpu. Then this morning everything seemed fine again. The cpu wasn't being used much at all. Now again this afternoon the cpu started running high again. There really isn't a difference in the way the computer is being used when the cpu is being maxed out and the way the computer is being used when everything is fine.

Any idea why this is happening? This is driving me crazy!

 

[Minstedmaz]

There has to be a process/program using your cpu, if you don't see anything running you should try to search your computer for malware.

 

[InvalidError]

Try using Process Explorer instead of Task Manager, it often shows stuff Task Manager misses and gives you tons more detail about each process.

 

[Natos]

InvalidError does have some good advice. Process Explorer makes Task Manager look like it's blind.

Also, as Minstedmaz suggests, scan your system for viruses. If nothing shows up, try using a bootable disk to scan. This will see anything that has a cloaking device that starts before windows.

 

[Minstedmaz]

Did you try what we suggested you? Did you use procces explorer? Did you search for virus or malware?

 

[rakosnik]

I ran Memtest86 and it didn't find any errors after 7 passes.

I had Windows Security Essentials installed already. I ran a scan and it didn't find any viruses.

I ran AVG rescue disk and it didn't find any viruses.

I took out the new RAM and installed the old RAM and my computer is working much better.

Could there be a setting in the BIOS that I needed to change for the new RAM that would explain my symptoms?

Could I have bought the wrong kind of RAM somehow that would still allow Windows 7 to boot up and function, but with the increased CPU usage that I've previously mentioned?

 

[InvalidError]

Even the worst RAM with the worst timings that otherwise works fine would not cause drastically lower performance so there likely is something more involved in there.

 

[Natos]

What CPU are you running, and what motherboard are you running?

It would take a switch from DDR3-1600 to DDR-266 to make that kind of a difference. Even then, that might not have as large of an impact.

The only BIOS settings that should apply to your problem are RAM frequency control, and CPU frequency control. Most of what the BIOS does is only during POST, after which it hands most of it's task over to the OS.

 

[rakosnik]

Ok, I swapped in the old RAM and at first everything seemed to be fine again. But now its not. So now I'm wondering if I may have accidentally done something when I originally installed the new RAM. Inside the case the components are packed very closely and it was a pain to switch the RAM.

Is there a chance that I could have damaged the CPU or something on the motherboard, but the computer is still able to boot up and function, but just with the CPU symptoms I'm describing?

I also noticed that there was a lot of dust on the CPU fan. I'm actually out of canned air, so I just blew across it, but I'm afraid I didn't get a lot of the dust off of the fan. Could overheating or something else related to the dust be causing the CPU symptoms I've described above?

 

[InvalidError]

If you had damaged something, you would be more likely to get crashes and reboots than performance issues.

The CPU overheating due to a cooling issue could cause it to throttle down to very low clocks but this is rarely becomes an issue until there is some really severe dust accumulation in there.

 

[rakosnik]

Motherboard: ASRock H67M-ITX/HT or H67M-ITX (the manual is for both).

The processor is a Core i3-2100T at 2.50 GHz

The RAM I just installed and then unistalled is PNY XLR8 8GB Desktop Memory Module - DDR3, 1600 MHz, PC3-12800, CL 9, 1.5V - MD8192SD3-1600-X9 Item#: P56-4510 | Model#: MD8192SD3-1600-X9
http://www.tigerdirect.com/applications/SearchTools/item-details.asp?EdpNo=8045512&srkey=P56-4510

 

[rakosnik]

I cleaned off the fan and the rest of the inside of the case and it has been working fine since with the new RAM. It has gone through about 12 hour stretches before where it seemed to work fine, but then the CPU would spike again. So I'm hoping that it was just over heating related to the amount of dust on the CPU fan. Maybe when I just blew on it after opening the case the first time and initially installed the new RAM I compacted the dust by blowing on it and that caused the over heating issues to suddenly begin.

 

[InvalidError]

If you had enough dust in there to cause overheat leading to a crash, you probably want to give your whole case a thorough cleaning including blasting the dust bunnies out of your PSU.

 

[rakosnik]

That's just the thing. Its not crashing. The processor starts running a lot harder than it should for what its doing. It just went through about an 18 hour period where it worked fine and then suddenly the CPU started maxing out again. This time when I quit what I was doing it backed down faster, but it basically is still having the same issue.

As I said in my post just prior to this last night I cleaned out the whole case, especially the CPU fan and the power supply.

If it is a heat issue with the CPU, are there good monitoring programs to see the temperature of the CPU? Maybe I'll see a connection between when the CPU is acting up and the temperature of the CPU. And if I don't I will have eliminated that as a possibility.

 

[InvalidError]

If it is a heat issue with the CPU, are there good monitoring programs to see the temperature of the CPU?

All that system monitoring tools do is scan the SMBus for sensor data and report it so all utilities should be reporting pretty much the same values as long as they have the right sensor description for what your motherboard uses. Pick one and as long as the values are within expected ranges, it is most likely fine.

 

[rakosnik]

I've installed Core Temp and I've been monitoring the temperature for the past few hours. I haven't used my computer for much, but the temperature started out at 50 C and now its in the 70s. Core Temp says the max temperature is 85 C.

 

[InvalidError]

85C peak while "not doing much" could be due to a HSF mounting issue - very common with Intel's stock HSF after 3-4 years. You may have nudged the HSF while mocking around the DIMMs and broken the thermal paste "bond" between the CPU and HSF. If this is the case, you probably need a new HSF.

 

[rakosnik]

Does HSF stand for Heat Sink Fan?

And would this explain the CPU working hard than it normally needs to.

 

[InvalidError]

Knocking the HSF would explain the higher than expected temperature but not the CPU load in Task Manager.

 

[rakosnik]

Is it normal for the CPU to get close to its max temp? We just finished watching a movie, which didn't stress the processor that much. But when we were done the CPU temp was around 80 degrees C. Its max temp is only 85.

If the CPU reaches max temp, will the system automatically crash and shutdown or is there risk of the CPU damaging itself?

Can the CPU go beyond 85 C?

 

[InvalidError]

Can the CPU go beyond 85 C?

Sandy Bridge CPUs are rated up to something like 92C or 95C but they are not supposed to get anywhere near that hot under normal use.

My i5-3470 (Ivy Bridge) is rated for up to 105C but it rarely goes beyond 65C while operating at about 70% load using the stock HSF.

The CPU has thermal limits that cause it to first throttle to reduce its TDP and hopefully drop its temperature before anything bad happens. If that fails, there is a second trip point that triggers an immediate shutdown to hopefully prevent damage. So in theory, it should be relatively difficult to accidentally kill modern Intel CPUs through overheating when everything else is stock.

 

[rakosnik]

I've been watching WTV video files made by Windows Media Center and that seems to increase the temperature more and faster than MPEG-4 videos I make with Handbrake or MCEBuddy. 82 C seems to be the magic number. Once my PC hits 82 C the video begins to get sluggish and the CPU starts to work harder.

 

[InvalidError]

Once my PC hits 82 C the video begins to get sluggish and the CPU starts to work harder.

The CPU is not "working harder;" what you are seeing is probably thermal throttling kicking in: the CPU reduces its clock and voltage to reduce its heat output and prevent its temperature from going any higher.