CPU Usage...and you.

[Skrypnyk]

So, I'm in a bit of a confused state right now. For the past week or so, I've been noticing sudden slow downs of my computer. Like, really bad slow downs. Either I'd be in FL working and everything will start to sound all glitchy, or I'd be playing Flatout 2 and everything will..just.....run........frame..............by..................frame.

So, I would assume that something is whoring up the CPU when this shit happens and I did a little test.

Test 1:

The little drop in PF Usage was when I closed mIRC, IM, etc. I did not touch anything after I closed the programs.

Test 2:

This was taken after I shut down every single thing I could. Stuff running from the system, stuff running from admin, local service, etc. I did not touch anything after I closed the programs.

Test 3:

This was taken after I restarted my computer. I waited a minute so that everything could load, after which I did not touch my computer.

I noticed that there's a lot of major fluctuation in the green line. I went on my sisters computer (which is quite outdated technology wise), yet her CPU Usage stays from 0-10%.

So I'm curious as to what a solution might be. I've already scanned my hard drive as well as defragged it, yet I don't really see much changes.

[Xerol Oplan]

Go to the processes tab and enable the column for CPU usage, and see what's actually taking up that CPU.

What kind of virus/spyware protection are you running? If it's Norton Anything then that'll consume large amounts of CPU at regular intervals, and some other programs also do periodic scans.

[Skrypnyk]

I do have symantec running in the background, but I've always had that run in the background with no real problems.

Watching the process tab run, most of the time the system idles is at 99 with like, explorer or firefox going from 2-9.

Winamp did jump to like, 30 something at one point though.

[Pyrion]

CPU usage shouldn't inherently matter unless you've got a process prioritized higher than anything else you're running.

Severe framerate drops and glitchiness sounds more like you're running out of physical memory resulting in pagefile HD thrashing.

Also, make note that Task Manager is misleading as fuck. You'd be better served with SysInternals Process Explorer. It's far more accurate and useful than Task Manager.

[Skrypnyk]

CPU usage shouldn't inherently matter unless you've got a process prioritized higher than anything else you're running.

Severe framerate drops and glitchiness sounds more like you're running out of physical memory resulting in pagefile HD thrashing.

Also, make note that Task Manager is misleading as fuck. You'd be better served with SysInternals Process Explorer. It's far more accurate and useful than Task Manager.

Any suggestions on how to fix this 'physical memory' problem? But more RAM I assume?

[Pyrion]

CPU usage shouldn't inherently matter unless you've got a process prioritized higher than anything else you're running.

Severe framerate drops and glitchiness sounds more like you're running out of physical memory resulting in pagefile HD thrashing.

Also, make note that Task Manager is misleading as fuck. You'd be better served with SysInternals Process Explorer. It's far more accurate and useful than Task Manager.

Any suggestions on how to fix this 'physical memory' problem? But more RAM I assume?

No, not necessarily, just kill some services and processes you don't need.

http://majorgeeks.com/page.php?id=12

[Gollgagh]

Yeah, my CPU usage fluctuates that high pretty often without being seriously slow like what you are describing.

[Smoke]

CPU usage shouldn't inherently matter unless you've got a process prioritized higher than anything else you're running.

Severe framerate drops and glitchiness sounds more like you're running out of physical memory resulting in pagefile HD thrashing.

Also, make note that Task Manager is misleading as fuck. You'd be better served with SysInternals Process Explorer. It's far more accurate and useful than Task Manager.

Any suggestions on how to fix this 'physical memory' problem? But more RAM I assume?

No, not necessarily, just kill some services and processes you don't need.

http://majorgeeks.com/page.php?id=12

Disabling services has no measurable effect on your system performance unless you have a REALLY slow system. It's a classic myth, and there's been benchmarks to prove it.

An idling service won't use any CPU power until it's called to action, and will barely use any RAM.

http://www.besttechie.net/forums/index.php?showtopic=3007

There's a link to a no longer existant Anandtech topic in there, which did include benchmarks on various configurations to see how much of a difference it made. The only difference visible was on a pretty much totally outdated system that barely ran XP to begin with.

As for actual advice: Pay attention to what's taking up that CPU power at those intervals. Don't look at the pretty graph because it won't tell you much, look at the Process list instead.

[Pyrion]

Disabling services has no measurable effect on your system performance unless you have a REALLY slow system. It's a classic myth, and there's been benchmarks to prove it.

Really? First time I've read about this, on all of the tech forums I frequent.

I'll take your word for it.

[Smoke]

Disabling services has no measurable effect on your system performance unless you have a REALLY slow system. It's a classic myth, and there's been benchmarks to prove it.

Really? First time I've read about this, on all of the tech forums I frequent.

I'll take your word for it.

It's a popular myth . Most sites just take it from others without doing any actual research, and quite a few people who follow those guides tend to be convinced by the idea of 1FPS more(Even if they don't achieve it, or don't even consider seeing what the difference is)

More fun myths:

http://mywebpages.comcast.net/SupportCD/XPMyths.html

EDIT: Oh hey here's the actual benchmarks:

http://forums.anandtech.com/messageview.aspx?catid=34&threadid=1678445&enterthread=y&arctab=y

[zylance]

Disabling services has no measurable effect on your system performance unless you have a REALLY slow system. It's a classic myth, and there's been benchmarks to prove it.

Really? First time I've read about this, on all of the tech forums I frequent.

I'll take your word for it.

It's a popular myth . Most sites just take it from others without doing any actual research, and quite a few people who follow those guides tend to be convinced by the idea of 1FPS more(Even if they don't achieve it, or don't even consider seeing what the difference is)

More fun myths:

http://mywebpages.comcast.net/SupportCD/XPMyths.html

EDIT: Oh hey here's the actual benchmarks:

http://forums.anandtech.com/messageview.aspx?catid=34&threadid=1678445&enterthread=y&arctab=y

That guy doesnt like Firefox? And he advocates for IE! I'm not sure I would trust him...

[Smoke]

Disabling services has no measurable effect on your system performance unless you have a REALLY slow system. It's a classic myth, and there's been benchmarks to prove it.

Really? First time I've read about this, on all of the tech forums I frequent.

I'll take your word for it.

It's a popular myth . Most sites just take it from others without doing any actual research, and quite a few people who follow those guides tend to be convinced by the idea of 1FPS more(Even if they don't achieve it, or don't even consider seeing what the difference is)

More fun myths:

http://mywebpages.comcast.net/SupportCD/XPMyths.html

EDIT: Oh hey here's the actual benchmarks:

http://forums.anandtech.com/messageview.aspx?catid=34&threadid=1678445&enterthread=y&arctab=y

That guy doesnt like Firefox? And he advocates for IE! I'm not sure I would trust him...

Note that his Firefox section mentions Opera a whole lot. Some of his info is outdated though, and he's a bit in denial about IE's lack of standards support. The text shows quite a bit of bias against Firefox too(Such as the parts on how IE is so wonderful because it works with most websites, and how horrible Firefox is for not passing the Acid2 test, which IE fails completely)

His other info's pretty good though.