With Windows Vista x64 and 6GB of RAM do you need a page file?

Some enthusiast does, and he's willing to sacrifice a bit of reliability to have some fun with his engine.
Nobody is arguing against your element of chioce. You can do what you want.
It's just that you are not getting a 6X increase in power. But you are still increasing your risk, which the point we are trying to give you, is it is not enough (if at all) a performance increase to actually take the risk to change it.

To the people who say don't do it cause it'll cause problems, have you ever tried it?
Yes, actually. However not with a machine with this much RAM in it.
In one circumstance, Windows simply re-enabled it to prevent my system from crashing. In another, my system got very jacked up (more of a test, so nothing bad came of it... all my stuff backed up).



Like I said before: read up on how this stuff actually works.
Just having the ability to "allocate" RAM to the paging file (a very good chunk of which is actually used) means more of your physical RAM gets put to use.
In a more accurate way, you are actually decreasing the physical RAM your actual apps have to run.

But I'm done with this thread.... Do what you will with your system, just don't come crawling back when it doesn't work, as you've been advised against it! No excuses at that point. Isn't Microsofts fault or anyone elses.
 
So you think running programs should be put to hd on virtual memory but at the same time fill the ram with superfetch, containing crap that is not being run and should not be in the memory at that time?

ROFL! Small controversy there eh?
 
I never had any problems with it, but I turned the page file back on.

I honestly can't tell a speed difference one way or the other. The only thing I notice is that my hard-drive is louder/on much more with a page file.

I'm currently doing a lot of work with four virtual machines. Critical work so I enabled it for the time being to ensure I don't have any problems. That being said, I did work with those VM's for a bit with no page file and had no problems.
 
its a bunch of nonsence saying that disabling your page file will mess up your system. if you dont agree then tell me why a system with 4gb ram and 4gb swap file is going to be more stable than a system with 8gb ram??? well there is no difference in stability is there!! now if a system has 8gb ram and a swap file there will obviously be more memory available IF ITS NEEDED! now if i have 8gb ram and i really only need like 2gb like most users doing gaming photoshop ect then there is no way dissabling the page file will messup the system if you have 8gb ram when its not even being used.
 
Search for posts from me, on why disabling a pagefile is bad. I've described this hundreds of times. Other Microsoft people have also posted why disabling the page file is bad, hundreds of times... These threads always snowball into 100's of posts with it works for me/or is breaks for me. Running without a safety net is not advisable. But it's your computer, you can do with it what you want...

This posting is provided "AS IS" with no warranties, and confers no rights.
 
The fact that it's possible to do it doesn't mean that it's worth doing. If you have no page file on your 8GB system and it all works fine, then that's okay. But it won't be giving you performance improvements; the only thing it will mean is that there's hundreds of megabytes of that 8GB being used up by useless pages which could be put to better use by SuperFetch/cache. On a system with 8GB RAM it's not likely to make much difference either way because there's still plenty of RAM free, but even if performance doesn't decrease, it's not going to increase either.
 
The fact that it's possible to do it doesn't mean that it's worth doing.
Thank you! I love how many people hear of these uber-tweaks from someone, and just assume that because it can be changed, it must be better! Can and should are two very different words.
 
if you dont agree then tell me why a system with 4gb ram and 4gb swap file is going to be more stable than a system with 8gb ram???

Day to day it probably won't.
But there is no noticeable increase in disabling the swap file, putting on another drive, or all this other stuff.


So since there is no difference, why on earth do you want to take this failover mechanism away from the OS?
 
I have 6GB and still use the pagefile. I don't see the point in disabling it. But maybe I should perform some benchmarks to see if it makes a significant difference in anything.
 
Windows always has a pagefile; it is an absolute impossibility to "disable" such a critical component of the virtual memory subsystem. People have been perpetrating this myth of disabling it or setting it to 0/0 for what, over 15 years now? And it's just as true as it was when it first appeared: it's not true, and as such it would be great to see people get past it.

Aside from getting some ridiculously fast hard drives, or excessively massive amounts of RAM and putting the pagefile on a RAMdisk (entirely possible, and wicked fast these days), there's really nothing the end user can do with some Registry edits or variable adjustments that will make any noticeable difference outside of running a benchmark that specifically tests for such things. And really, who the hell sits around running benchmarks 24/7 and doing nothing else with their PC?

Right...
 
^ It's quite possible I'm wrong --- but I believe you can in fact nearly disable the page file - at least if explorer isn't lying to you.

Important - - - - Set your page file to the lowest amount first. 2MB for XP. Then Reboot. Then Change it to no page file and then reboot. If you change it to no page file without first minimizing it and rebooting it seems the size of the page file gets locked and you lose that much hard-drive space (If I remember correctly.) Maybe that's why you think you can never disable the page file?

After the reboot look at the pagefile.sys on your C: drive (typically)

It should be 2MB ---

That for all purposes isn't much of a page file.

I can't try from work because it's locked down, but you can for all practical purposes get rid of it.
 
Yeah, in XP (and earlier versions of Windows) you can definitely disable the PF, delete it and it's not recreated. I haven't tried in Vista yet.

I used to do that before ghosting my drive since it didn't get zeroed out on shut down. I know it's possible, but I'd rather have a faster shutdown the other 99%+ of the time.
 
That's because you're looking in the root directory of the system partition for a pagefile.sys file - and that's not where Windows will create the necessary file when you set the options for "No paging file."

Takes time to learn the ins and outs of Windows - you can't just read some tweak guide in an hour and know it all. And setting the pagefile to 2MB will end up being rather detrimental in the long run - if and when Windows needs more pagefile to work with, you lose performance drastically when it wastes all that time expanding the entirely-too-small pagefile the user decided was 'enough' and then has to populate it with the data it needs to page.

Three little words are still the best possible advice after Windows is installed, any version of Windows:

Leave it alone.
 
That's because you're looking in the root directory of the system partition for a pagefile.sys file - and that's not where Windows will create the necessary file when you set the options for "No paging file."

Takes time to learn the ins and outs of Windows
Well, then enlighten us.

I'm just amazed you can say that and sit at the same time. That's like ventrilliquism. :D
 
Yah well, you do this as long as I have, fix as many machines as I have, install Windows and tweak it obsessively for decades, teach people how to do it right thousands of times... then we'll talk. :)

You cannot disable the pagefile in Windows, period. You can do whatever you think you're doing, you can set whatever you think you're setting, and Windows will recreate it as necessary as a temporary file in the Windows\System directory with a random filename as required. The pagefile is a major component of the virtual memory subsystem, as has been noted and explained countless times over the years. Disabling such a major component could/would have serious repercussions on the operating system, typically causing it to grind to a screeching halt.

Modern OSes have been designed to work with the virtual addressing schemes since the 80386 processor brought the technology into the light many years past. To think that even now, with today's high powered machines and gobs of fast high speed RAM, that you can simply "turn off" such an integral part of an OS... it's ludicrous.

As far as people figuring it out, well... I guess it's up to them to do the research and learn how Windows (and other OSes) really work.
 
Yah well, you do this as long as I have, fix as many machines as I have, install Windows and tweak it obsessively for decades, teach people how to do it right thousands of times... then we'll talk. :)
Oh, the "trust me, i'm an expert response" with vague details. Nice. I thought that went out of fashion already. :p

I'll put your shens to the test later when I get home.

Don't get me wrong, i never said disabling the PF was a good idea, the whole topic here (in the last couple of exchanges) is whether you can disable it. From a programming perspective, I can understand keeping and using some of the VM structures (in memory) by the OS may happen whether or not a PF exists. Your seemingly invented on the spot explanation (phantom pagefile) isn't very convincing. But is is very easy to test. :p
 
All you guys with 4gb or more of ram, and multiple 500gb drives, jeesh, you worry about having a pagefile of 1-2 gigs? LOL.....

Don't disable it, the pagefile does more than you think, especially in Vista. I have been setting my pagefile to 2048mb ever since I started getting to 2gb ram, and now with 4gb in my Vista 64 rig, I still leave the pagefile at 2048mb.

Microsoft is smarter than the average guy, and if having 4gb or more of ram negated the pagefile, they would know about it....

Leave it on.
 
You cannot disable the pagefile in Windows, period.

The pagefile is a major component of the virtual memory subsystem

Modern OSes have been designed to work with the virtual addressing schemes since the 80386 processor

Of course, you can disable the paging file. Virtual memory and paging is something else. Paging is when the processor needs to lookup the physical address in a page table. The page can be in physical RAM or it might have been paged out.

This kind of paging cannot be disabled, of course, but the paging file can (when the paging file is disabled, the OS cannot page anything out, of course).
 
Joe Average....

Most, (not all) of the people on this forum have been "doing" computers for years.

You’re making a pretty extravagant claim. I have access to Microsoft premier support and I think I'll just ask our premier field agent exactly what he thinks about your claim.
 
Nothing a little research at TechNet can't fix... I learned it, so can others. When and if Windows needs to page, it's going to do it, regardless of whatever settings the user has chosen, I don't think anyone will argue that point - and that's exactly what I'm saying.

Hell, some software will not even function without a pagefile enabled - Photoshop, anyone?

The basic gist is again, as always, leave it alone. Windows works just fine and dandy as long as you leave it alone and use it. Doing things that are outside the norm is when the problems occur, and they do, and we all know it, even those with less experience than some of us.
 
Joe, you need to do more research, if you still say that the OS cannot live without the paging file. Instead of just reading through KBs, I can recommend intel.com and amd.com. They have some good PDFs on how the CPU works with virtual memory, paging and all that
 
Don't get me wrong, i never said disabling the PF was a good idea, the whole topic here (in the last couple of exchanges) is whether you can disable it.
Actually: look at the subject and the OP again... it's both.


NWhen and if Windows needs to page, it's going to do it, regardless of whatever settings the user has chosen
Yes!
I think some people are talking from their ass on this one.

Windows can enable the page file if it has the chance and needs it.
The flip side of that is that I've seen it where Windows DIDN'T do that and the system had a pretty bad crash.
 
LOL

OK, I'm home and I tested before and after disabling the page file on my XP system (HTPC in my sig). The system has 2GB installed. I hope you don't mind that I actually tested it and didn't just pretend to be some kind of "expert." The results were a real nail biter.

1. I captured the entire list of files in the \windows folder and all sub-folders to use as a base line with the PF enabled.

2. I turned off the page file. I checked the \windows folder and compared to the base line. No new files were created. Rebooted.

3. The PF was automatically removed from the root directory during the reboot, as it normally does when disabling the PF. The were 2 file differences after the reboot: the perflib_perfdata* performance log files had new filenames, like they do on every reboot. Those were the only changes from base line in the \windows folder and all sub-folders.

I'd like to pause for a moment to completely dispell what Joe Average claimed. Windows does not "recreate it as necessary as a temporary file in the Windows\System directory with a random filename as required"

Just to put that silliness to rest, I used a 2 line utility that allocates a 50MB of memory at a time to fill up all memory until it threw an out of memory exception.

Guess what, no phantom swap file was created. :rolleyes: No new files were even created in \windows and sub-folders compared to the base line before disabling the PF. You were 100% wrong on this. If you don't know, just don't make up stuff to pretend you do.

Goodbye and PLONK. You'll be in "good" company.
 
Can I ask a question?

What is the point of arguing about this, if it makes ABSOLUTELY no affect on performance?
 
What is the point of arguing about this, if it makes ABSOLUTELY no affect on performance?
The problem really is about bad information being passed by a couple of "experts" who have obviously never tested it with a huge amount of memory. It's OK to be cautious and just say leave it enabled, but some of these people start making ludicrous claims that are testably false.

The forum is a place to pass around information, not misinformation. I'm actually curious now if applications and games (including DX/OpenGL texture management) can gracefully handle out of memory situations. I think i'll turn off the swap in Vista and see how long it takes to get errors with 6GB installed. Not only would it satisfy my curiousity, it could also provide a data point to answer the thread starter's question. Now that is information. ;)
 
Just to put that silliness to rest, I used a 2 line utility that allocates a 50MB of memory at a time to fill up all memory until it threw an out of memory exception.

I've seen this happen.

I've also seen Windows completely re-enable the page file as well...

Your both right, I'd say. I can care less one way or the other, it's just that I HAVE seen Windows re-enable a page file after disabling it (took it a couple days before it did that, but it eventually did).



And FWIW your memory exception errors (Which get worse as time goes on, especially if you run large apps like Photoshop) are exactly why you need to keep a pagefile enabled.
Yes, with 6GB of RAM, you shouldn't run out of RAM.
I also don't forsee myself crashing anytime soon, but I won't be neglecting to put on my seatbelt, either.

Just a fallback mechanism... Enabling and Disabling makes no difference on performance. All you end up doing in the end is removing a safety measure, so WHY BOTHER?!?!?!?!?
 
Hell, some software will not even function without a pagefile enabled - Photoshop, anyone?

Photoshop will give a little message on startup saying there is no paging file, but it will run just fine. Photoshop also has its own internal virtual memory system, the scratch disk
 
The PF is recreated if it is deleted and it is still enabled. If you disable the PF, it is not recreated automatically. I just demonstrated that above.

BTW, of course I got an out of memory exception. I did that on purpose to demonstrate the fact that Windows doesn't just recreate the swap file by itself after it has been disabled. That was the silly claim you enthusiastically embraced a few posts up.
 
Photoshop will give a little message on startup saying there is no paging file, but it will run just fine. Photoshop also has its own internal virtual memory system, the scratch disk
LOL

bbz_Ghost was 0wned again.
 
Going on the posts above, Joe Average said some software. Photoshop is one example of software that uses it's own swap file implementation.

Do we really have to be this ridiculous to call him owned because one piece of software doesn't fit his comments? He didn't say ALL....he said some. This is turning into a childish game, and isn't helping anyone out. pxc, you gave me a lecture above about providing information, and I couldn't agree more....but please tell me how your last post is providing anyone with anything useful?

Once again, I'll pose the question...who the hell cares? If it doesn't give any benefits, then I think we can all safely agree there's no reason to do so....so WHY still argue about it?

If you are going to lecture me on how the main goal is to provide correct info (I have a patent on that lecture, by the way, and you own me royalties), why not follow your own advice and stick to posting facts, giving links, or offering up some suggestions on tests we can all try out? Why stoop to the level of making personal attacks.

I was enjoying the reading of both points, but now we just reverted to the usual bullshit.

I'm voting that a mod lock this thread. There's no real point in debating this, as it serves no purpose or benefit. There's no reason to disable the page file, even if it causes no harm to the system. Let's leave a dead topic in the grave where it belongs.
 
I'm voting that a mod lock this thread.
I vote not.

I will be running my main system with 6GB installed without the PF enabled, and noting any problems that come up over the next week for this experiment. :) Well, if it is locked I'll just make another thread with the results next week.
 
If we stick to testing and work out some other tests that several people can try, than great. If it's just going to continue to degrade into more personal attacks, then there'd be no point.
 
you know if it all matters get some small SSD of about 4 to 8GB move the page file to it and bam done.

/thread
 
I can care less one way or the other, it's just that I HAVE seen Windows re-enable a page file after disabling it (took it a couple days before it did that, but it eventually did).

The problem with this theory is that it's non-falsifiable.
 
The problem with this theory is that it's non-falsifiable.

Not a theory, did it twice in a row, actually...

But like I said, I've royally jacked some systems up not having a page file, so I really don't want to test it out anymore and have to repair my system afterwards.
 
Not a theory, did it twice in a row, actually...
And apparently you're not a scientist (or are scientist that doesn't know the definition of theory).

If you can provide instructions for how to reproduce your results, or provide evidence that it actually happened, then you might have a basis for your argument. Otherwise, the only reason we have to believe you is faith in your expertise and honesty, despite evidence which suggests you are not correct.

Maybe you are right, but you certainly aren't making a compelling case for it.
 
Really?
I would have thought most of you would have gotten past the point of these silly little arguments.

I'm going to leave this open if you guys want to actually discuss this, but the first signs of personal insults, even vague ones will earn you an infraction, not a warning.
 
The null hypothesis "Windows will never re-enable the page file spontaneously" would be disproved by a single case in which Windows was observed to re-enable the page file spontaneously, I'd say. However, you'd have to provide evidence supporting that in order to meet the necessary burden of proof that it had indeed re-enabled the page file. A model for when Windows might re-enable it would be useful, but not required to disprove that it never happened.

This having been said, I'm not sure if it does re-enable the page file or not. I occasionally see mention of temppf.sys, but the Microsoft KB articles only mention it for NT/2000.
 
See, I knew my faith in some of the people around here wasn't unfounded. Just a bit misplaced at large... but the info is out there, it always has been, and it's relevant across every NT-based OS that Microsoft has ever created. No matter how much RAM you have in the machine, if and when Windows "needs" a pagefile, it will create one on-the-fly - and since all NT-based OSes have a virtual memory subsystem of which the pagefile is a component, that means... <drumroll please> you can't disable the pagefile.

Sure, you can set it to zero - which does not disable it but only gives it a zero byte size, and you may never encounter a situation where Windows requires one past that point (how fortunate for you) which is usually the case where you have a machine with a lot of RAM and you simply never put enough of it to use to matter or breach that point where Windows has to create a temporary one...

But that doesn't change the fact that you (as a User or even an Administrator) never really have the absolute control over the pagefile that a simple "No paging file" radio button seems to grant. Windows will do what it needs to do to keep functional.

It's smarter than people seem to think... or at least the Microsoft folks did a better job than people give 'em credit for.
 
Joe, you need to reference some real sources now, instead of just saying stuff.
 
Back
Top