Advisable to disable paging file?


Recommended Posts

If you're asking if I've done laboratory tests on multiple computers then No

But if you're asking if I've just used Windows since 3.11 and adjusted you-name-it ;) then Yes

There is no set rule for the amount of PageFile, because every system is different

50 Meg? Only "Just in case" That way I'm safe :)

  Hitman2000 said:
its really simple.

try a couple days without pagefile and if you feel its faster and experience no lockups or problems then keep going.

That is basically exactly what I was going to say.

The OP said that he experiences stuttering during games, and wondered if turning off the PF would help with this. My suggestion would be, turn it off and see if there is a change. Turning it off will NOT harm your computer, but it does bring up the possibility of 'out of memory' issues if you've got a lot of active memory hogging programs...which you wouldn't have while playing most games.

To the OP, make a note of the settings and values of your current PF and then turn it off. See if you experience any of the stuttering you said you were experiencing. If you don't, and there are no other problems then run it for a while without a PF if the stuttering was really that bad. If you still experience the stuttering with the PF turned off, then turn it back on with the previous settings.

The one thing I'd add:

If you have more than one physical hard drive, it's worth your while to enable a system-managed page file on the extra drive, and disable the one on your c: drive. It will cut down on the access to c: and reduce the chance of stutter.

  Thunderbuck said:
The one thing I'd add:

If you have more than one physical hard drive, it's worth your while to enable a system-managed page file on the extra drive, and disable the one on your c: drive. It will cut down on the access to c: and reduce the chance of stutter.

Nonsense. While a pagefile on a second hardrive may be faster (especially if the 2nd hd is faster than the primary hd), it has nothing to do with "reducing stutter" - whatever that means. Utter nonesense.

As you've undoubtedly seen in this thread, people who have absolutely no idea what they are talking about love to give advice about pagefiles. As I said earlier, these pagefile threads are just silly. The best available advice is to do whatever you think works best for you. Period.

  allan said:
Nonsense. While a pagefile on a second hardrive may be faster (especially if the 2nd hd is faster than the primary hd), it has nothing to do with "reducing stutter" - whatever that means. Utter nonesense.

As you've undoubtedly seen in this thread, people who have absolutely no idea what they are talking about love to give advice about pagefiles. As I said earlier, these pagefile threads are just silly. The best available advice is to do whatever you think works best for you. Period.

Look, maybe "stutter" was not the precise term I should have used (because, yes, that relates more to video performance), if disk IO bottlenecks on the system disk you could certainly see some periodic hangs. Placing the page file on a separate physical disk definitely cuts down on this.

  Thunderbuck said:
Look, maybe "stutter" was not the precise term I should have used (because, yes, that relates more to video performance), if disk IO bottlenecks on the system disk you could certainly see some periodic hangs. Placing the page file on a separate physical disk definitely cuts down on this.

Same response. Utter nonsense.

Disk can cause stutter and even hangs if R/W access becomes unavailable due to a bottleneck. This might mean that it is time for a new HD, or if you have a cheap SSD (like I do), it means that Windows is writing to it too much too often (major amount of write requests will get cheap SSDs stuck, which games should not do - they should use the RAM).

  Udedenkz said:
Disk can cause stutter and even hangs if R/W access becomes unavailable due to a bottleneck. This might mean that it is time for a new HD, or if you have a cheap SSD (like I do), it means that Windows is writing to it too much too often (major amount of write requests will get cheap SSDs stuck, which games should not do - they should use the RAM).

Exactly! To say it is because of the pagefile, however, is (to repeat myself :) ) utter nonsense.

  kimsland said:
I agree, except when some programs explicitly require the PageFile even if you have 8Gig of ram installed (actually I think even Office prefers some PageFile)

I wonder what others have given their PageFile, and what works for them?

No page file since i have 8gb, which is since xp x64. Using office, a RAD, browsers, audio/video manipulation, x264 en/decoding you name it.

Not once ever have i had a insufficient memory error.

But, if you just open everything at once and leave it running i can imagine that you manage to run out of steam. Besides that it's plain silly to leave apps which don't do anything running, windows multitasking prowess is not something to write home about.

In other words, the more programs you run at once the greater the risk of the OS crapping out on you.

The other thing is that Arma 2 seems to be a poorly programmed game to begin with. It runs almost the same speed on a high end SLI rig as it does on a low end single graphics card. I think the stuttering has more do with the graphics then with the RAM. You should hope that that the graphics companies find a driver change they can make to speed it up.

  Udedenkz said:
I had one with Windows 7, stay away from chkdsk in Windows 7 - it requires more than 4GB RAM to run for some reason - probably bug.

That's a load of crap. There is absolutely nothing wrong with chkdsk, nor does it require any certain amount of ram. I would know since I've successfully ran chkdsk /r on both my laptop and desktop which both have 2GB of RAM, and did not experience any issues during or after the chkdsk. Here's a post by Steve Sinofsky about said "bug": http://www.chris123nt.com/2009/08/03/criti.../#comment-11469

  xiphi said:
That's a load of crap. There is absolutely nothing wrong with chkdsk, nor does it require any certain amount of ram. I would know since I've successfully ran chkdsk /r on both my laptop and desktop which both have 2GB of RAM, and did not experience any issues during or after the chkdsk. Here's a post by Steve Sinofsky about said "bug": http://www.chris123nt.com/2009/08/03/criti.../#comment-11469

Just because you can't reproduce it, doesn't mean that is not there. Also the link you have given to, many people complain of this bug - one even says that it will use as much RAM as the largest file on the drive that is checked (in this case I need to create a 16GB PageFile to run CheckDisk). It will probably be fixed in Service Pack One, that or drivers will be fixed (unlikely for older systems?).

But there is no bug. All I've seen are people who reported it using so much memory (which is by design), and nothing more. It also does NOT use as much ram as the largest file that's being checked. In all honesty, I wouldn't care if they revert the behavior back to how chkdsk was before. The fact of the matter is there is no bug in chkdsk regarding its ram usage.

  xiphi said:
But there is no bug. All I've seen are people who reported it using so much memory (which is by design), and nothing more. It also does NOT use as much ram as the largest file that's being checked. In all honesty, I wouldn't care if they revert the behavior back to how chkdsk was before. The fact of the matter is there is no bug in chkdsk regarding its ram usage.

Well it sure crashed Windows 7 for me - I think.... Second time I tried it via CMD, and it boomed to 3.3GB and then I killed it off - after a low memory warning. You are probably right about the largest file thing, but to accept that a simple application that uses almost no RAM in Earlier Windows Operating Systems should be allowed to leech such an immense amount of RAM is ludicrous - by design I should be able to run chkdsk on a 1GB system. If the bug is not in chkdsk, then it is somewhere else and should still be fixed. :)

  Udedenkz said:
Well it sure crashed Windows 7 for me - I think.... Second time I tried it via CMD, and it boomed to 3.3GB and then I killed it off - after a low memory warning. You are probably right about the largest file thing, but to accept that a simple application that uses almost no RAM in Earlier Windows Operating Systems should be allowed to leech such an immense amount of RAM is ludicrous - by design I should be able to run chkdsk on a 1GB system. If the bug is not in chkdsk, then it is somewhere else and should still be fixed. :)

You can run it on a 1 GB and 512 MB system. It'll stop short of hitting the systems physical memory limit. What I observed from my experience of how chkdsk works in 7 was echoed by Steve's post saying it was by design. I knew something was weird about the "bug" when it decided to stop taking up memory at about 92% overall usage. Which made me think it wasn't a bug due to how it behaved. Now, If someone's system does crash or freeze, then the way chkdsk now works could be bringing out a bug elsewhere. I even let chkdsk finish the job, and once it was done all of the memory was freed and the program exited.

One thing I am curious about is whether chkdsk really does finish sooner, as Steve said, than it would have on Vista or earlier.

Interesting, if you use chekdisk (Explorer) on a SSD - no RAM increase. On a HD partition, If you checkdisk without telling it to fix errors / sectors - explorer will reach 600MB at most. This time, I reran checkdisk gui, with both options checked it stopped at around 3.3GB leaving 3% RAM empty. No memory warning so far!

So I am going to see if W7 is stable, as I am running without a PF at the moment, I am going to see what will happen when firefox grows and gobbles up the rest of RAM.

EDIT: It started to remove GUI elements, no thumbnails, missing icons. lol. :D

EDIT2: Actually this is kinda bad, for chkdsk didn't free up memory for other applications to use. Meaning that you can't multitask when running chckdsk. Whos bright idea was that?

Edited by Udedenkz
  kimsland said:
The why does a system run faster without it?

That is anecdotal, based on your opinion. Probably a placebo. Due to being misinformed about what the pagefile does, your subconscious wants to believe that disabling it will produce a performance effect, therefore you see a performance increase that isn't really there. Have you actually ran the numbers to check your own gut feeling? Because I have a feeling the numbers will surprise you.

  kimsland said:
Wait I'm sure it does

This is just by looking at a 512Meg Ram computer without Pagefile

That is most likely because the computer has a slower hard disk bottlenecking its performance. If you tried anything intensive, you will encounter problems, for example I disabled the paging file on an old system with a P4 and 768MB of ram, and then I received an out of memory error running UT2004 (even though my machine was well above the minimum spec).

Disabling the pagefile does not infer a performance benefit and anyone who infers that it does hasn't actually ran the numbers and put their computer under an intensive test. It doesn't matter how much RAM you have, Windows is designed to always use your memory and paging file as efficiently as possible, and disabling it just interferes with the balance. If you have a slow system, then it will be slow even with no page file.

People really need to stop spreading nonsense because the fact is DISABLING YOUR PAGING FILE INFERS NO PERFORMANCE BENEFIT. I have tested it on a number of computers, and never found it to help.

Don't disable your page file.

  Frank Fontaine said:
People really need to stop spreading nonsense because the fact is DISABLING YOUR PAGING FILE INFERS NO PERFORMANCE BENEFIT. I have tested it on a number of computers, and never found it to help.

MS even states to increase performance make your PageFile the same as Min and Max

This is because the PF can become fragmented

Many forums state place your PF on a separate Hard Drive

This is to reduce "C" drive from continuously writing (and basically Hard Drive head movement reduction on "C" System Drive)

In both cases disabling the PF fully on "C" System Drive, can produce better performance, in respect to hard drive activity, especially when other parts of the Hard Drive are being read/written to at the same time

Is that a performance increase or just hogwash!?

  yxz said:
i disabled the pagefile because i want my files at the beginning of the hdd

Where your files are has zero to do with page being on or not. Sorry but you are ill-informed there. Page file hasn't even needed to be in contiguous file location since the death of the Win9x kernel. And it never has been forced to be at the beginning of the drive.

To the original question: I have 8GB or ram and two page files. On of 8 GB on my C: drive 3 striped SATA drives (450GB total). And another 8 GB on D: 2 striped SATA drives (975 GB total). But under Vista and Win7 even that is probably over kill on the complexity. And I have a 4 GB dual channel Ready Boost drive as well.

IMO, Allan is 100% right. There is one small way however to get an increase in performance. If you have two or more PHYSICAL drives, not partitions, move the page file there. I have 3 drives and two page files. Both are on the 2nd and 3rd hard disks. Nothing on C: drive. I cannot see nor feel any difference BUT I know its there. There is less work for C: now. Its not going to show up in any benchmark IMO though. But its there.

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.