Does leaving your computer on cause a performance hit?

At this point i'd be inclined to say it's well within the margin of error. Will be interesting to see what happens after a day or two. Certainly a drop of 18 points in PC Mark would be no reason to restart frequently.
 
yeah, a good think would be to run restart the machine and run the test, then repeat this process as many times as possible, from this you can pool together all the results and determine a standard deviation from the mean score as a base comparison to see if subsequent results from the actual tests are just anomalies
 
bumblebee_tuna said:
This is just a suggestion but I think incoming dust is going to be a factor..... A computer that is running 24/7 is going to gather more dust than one that isn't..... And doesn't dust casue insulation and therefore heat?

Unless that is part of the experiment....... then never mind......
Click to expand...
My PC has been on for a while, dust isn't a huge issue in my room.

apj101 said:
Hi Omega, i really like your idea. Its god to approach these question with a scientific mind set, i like how you controled your test as best as possible...

the only issue i can see that PCMark 04 can give quite varied results, would have been best to run it a few times at each sample point and them take the average... 3 times would have been enough, 5 would have been better (time permitting). The drop of 18 points could just be due to the varience in the program
Click to expand...
Thanks for your input. I will re-test the experiement, and I will run PCMark04 3 times, just to get an accuate score. I will also run the CS:S stress test, and Lost Coast test a few times.

I'll report back with the initial results. But first i'll need to clear everything out again, so i'll have the results within an hour posted here.
 
Ok, I cleaned out my computer and rebooted. APJ made a good point, that running each test many times would help reduce the small margin of error that occurs each time you run the test. So I ran several tests 3-5 times, to get an accurate measurement.

PCMark04:
Number of times ran: 5
Individual results: 3658, 3635, 3638, 3604, 3648
Average from total tests run: 3637

Performance Test v6:
Number of times ran: 3
Individual results: 317.4, 316.6, 316.8
Average from total tests run: 316.9

CS:S Stress Test:
Number of times ran: 3
Individual results: 90.21, 89.69, 90.69
Average from total tests run: 90.20

HL2 Lost Coast Stress Test:
Number of times ran: 3
Individual results: 51.59, 51.01, 50.85
Average from total tests run: 51.18


Since these tests take much longer to perform, I will only conduct a test every 24 hours. Once I get a few days worth of data, I will create several graphs.
 
[-0MEGA-];445705 said:
Since these tests take much longer to perform, I will only conduct a test every 24 hours. Once I get a few days worth of data, I will create several graphs.
Click to expand...

Talking about having too much spare time here.. :rolleyes: :D
well,.. keep it up, im really ancious to see the results :)
 
Archangel said:
Talking about having too much spare time here.. :rolleyes: :D
well,.. keep it up, im really ancious to see the results :)
Click to expand...

hehe, I usually have it run a test and then read a magazine or something while it's running :P

I'm anxious as well. This time the test scores should be more accurate.
 
my goal is to get my computer to get to my desktop screen in 10seconds or less which its at like 12-13secs so... i have to get msn off startup and iam good
 
it depends on the amount of ram you have and what you do within that uptime...if you play a lot of games but only have 512 MBs of ram you're gonna see some slow downs after i'd say about 2 days...people with 2 Gbs of ram likely won't see any performance hits after a week
 
Maddhatter said:
it depends on the amount of ram you have and what you do within that uptime...if you play a lot of games but only have 512 MBs of ram you're gonna see some slow downs after i'd say about 2 days...people with 2 Gbs of ram likely won't see any performance hits after a week
Click to expand...

I'm trying to even it out, mostly what i'm doing on the computer is chatting with MSN and AIM, surfing the web, listening to music, looking at photos, videos, ect. I play games on ocassion, but not too often.
 
Here are the new results after having the computer on for ~14 hours:

PCMark04:
Number of times ran: 3
Individual results: 3482, 3472, 3476
Average from total tests run: 3476

Performance Test v6:
Number of times ran: 3
Individual results: 317.6, 315.7, 315.7
Average from total tests run: 316.3

CS:S Stress Test:
Number of times ran: 3
Individual results: 91.16, 90.10, 89.60
Average from total tests run: 90.29

HL2 Lost Coast Stress Test:
Number of times ran: 2
Individual results: 50.90, 50.14
Average from total tests run: 50.52



Previous Scores:
PCMark04: 3637
Performance Test v6: 316.9
CS:S Stress Test: 90.20
HL2 Stress Test: 51.18
 
leaving it on only hurts your energy bill
Click to expand...

Hardly, when a computer goes into sleep mode it consumes up to 99% less power than it does while in use. Leaving it on doesn't hurt anything.

Disk fragmentation, is really a thing of the past for normal users. Older file systems that were disigned differently had this issue because they would not overwrite the original file when modified. It could create a new file randomly and just mark the old one for deleting. While most linux/unix file systems are not really effected by disk fragmentation (tho not immune) other ones are. FAT had lots of problems with it.

So, yes it could happen, but really its unlikely unless you are a slave driver on your system, ie audio or 3d rendering.

Memory leaks are a factor as well. If you run applications developed at a sub par standard they can cause huge memory leaks and really cause system preformance hits on your computer. Rebooting helps clear that stuff up typically, and patching your application to the newest version also helps.

OS memory management, and memory controllers also play a factor. Crappy chipset drivers or OS corruption can cause performance hits as well. Things like not utilizing virtual memory correctly, not being able to read/write to system ram properly, the OS having problems allocating proper resources, etc.

In my professional opinion a well built and maintained system, with a solid (stable) up to date OS and software will not have huge performance hits in any way shape or form in basic or somewhat advanced useage. If you start editing and composing 100s of Gigabytes of digital film or audio and modify it constantly, then yeah memory problems may occur in your system.

I would like to see the same test done on OS X, a few versions of Linux, Windows, and then list the applications used as well. Video games are known to have memory leaks and could easily help the cause of systems losing performance.
 
tlarkin said:
Hardly, when a computer goes into sleep mode it consumes up to 99% less power than it does while in use. Leaving it on doesn't hurt anything.
Click to expand...
You completely contradict yourself.
Disk fragmentation, is really a thing of the past for normal users. Older file systems that were disigned differently had this issue because they would not overwrite the original file when modified. It could create a new file randomly and just mark the old one for deleting. While most linux/unix file systems are not really effected by disk fragmentation (tho not immune) other ones are. FAT had lots of problems with it.
Click to expand...
Fragmentation is still a big problem in modern hard drives


I would like to see the same test done on OS X, a few versions of Linux, Windows, and then list the applications used as well. Video games are known to have memory leaks and could easily help the cause of systems losing performance.
Click to expand...
Then why don't you run those tests? :)
 
Bobo said:
You completely contradict yourself.
Fragmentation is still a big problem in modern hard drives
Click to expand...

No its really not a huge deal at all, unless you are manipulating large amounts of data. Modern file systems keep it to a minimum so its not an issue.

Certain file systems exhibit a greater susceptibility to fragmentation than others, for example, a FAT file system becomes fragmented much more quickly than NTFS. Many file systems on Unix-like platforms do not require defragmentation at all.[citation needed] These systems attempt to keep fragmentation below a certain point so defragmenting is not necessary. This fragmentation resistance works well as long as the file system has a fairly large amount of space free.

On systems without fragmentation resistance, fragmentation builds upon itself when left unhandled, so periodic defragmentation is necessary to keep disk performance at peak and avoid the excess overhead of less frequent defragmentation.
Click to expand...
Source: http://en.wikipedia.org/wiki/Defragmentation

NTFS, HFS+, ext3, etc are fragmentation resistant file systems.




Then why don't you run those tests? :)
Click to expand...

I already know its not going to effect system performance a whole lot unless there is some other underlying problem, ie memory leaks, os memory management issues, defective hardware, etc. Or if you fill your Hard drive to the brim with downloads, data, and programs.
 
[-0MEGA-];446737 said:
Yes, just for kicks i'll run one PCMark04 test with all apps closed right now, and see what the result is.
Click to expand...
I just ran PCMark04 with no unnecessary apps open (just like the first test), and I got 3489.

tlarkin said:
I already know its not going to effect system performance a whole lot unless there is some other underlying problem, ie memory leaks, os memory management issues, defective hardware, etc. Or if you fill your Hard drive to the brim with downloads, data, and programs.
Click to expand...

I'm not filling my hard drive up with data, in fact i've actually deleted a large video that I dont need anymore. The point of the test is to see whats actually causing the performance hit. With PCMark04, I have the pro version so it gives me detailed information for multiple tests. When I run the test later today, I will compare the results with that of the first, to see what is actually performing worse.
 
[-0MEGA-];446745 said:
I just ran PCMark04 with no unnecessary apps open (just like the first test), and I got 3489.



I'm not filling my hard drive up with data, in fact i've actually deleted a large video that I dont need anymore. The point of the test is to see whats actually causing the performance hit. With PCMark04, I have the pro version so it gives me detailed information for multiple tests. When I run the test later today, I will compare the results with that of the first, to see what is actually performing worse.
Click to expand...

I wasn't making my statement directly at you omega, it was directed towards Bobo who says fragmentation is a big problem with todays computers and it is really not.

There are lots of factors that could cause this and if your system is health disk fragmemtation is probably not one of them. I leave my PC at home on all the time and I play some First person games and it performs nearly the same across the board, and I probably reboot the thing once a month.
 
tlarkin said:
I wasn't making my statement directly at you omega, it was directed towards Bobo who says fragmentation is a big problem with todays computers and it is really not.

There are lots of factors that could cause this and if your system is health disk fragmemtation is probably not one of them. I leave my PC at home on all the time and I play some First person games and it performs nearly the same across the board, and I probably reboot the thing once a month.
Click to expand...

I know you weren't talking to me, I just wanted to point that out.
 
You must log in or register to reply here.
Back
Top