• Sign in to Neowin Faster!

    Create an account on Neowin to contribute and support the site.

Sign in to follow this  

Windows 7 Pagefile/Superfetch/ReadyBoot discussion

Recommended Posts

StarLion    4

Stop trying to bull**** me. Your reg file quite clearly says this (in addition to Windows Media Player AutoUpdates)

; Disable automatic updates
[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\WindowsUpdate\Auto Update]
"AUOptions"=dword:00000001

That will kill automatic Windows Updates...

Not only are you not reading my posts, you apparently can't even read your own...

What Services? Sure keep your services, but I still want you to provide steps on how to reproduce instability / performance loss problems when running without a page file. I don't feel like arguing with you, I just want proof of your claims about stability
Not a substitute for RAM,

-- The PF is not anywhere as efficient as physical memory

-- Reliance on the PF does not mean anything good

-- You need to upgrade, remove bloat, etc if your PF is being used

Legacy Feature

-- The PF is an ancient feature

-- Its purpose being an extension of system memory

-- Without the necessity of thereof, its purpose is not there

Free Up Space,

-- The Page File wastes a good amount of space which can be utilized for storage.

Reduce Writes To Storage Medium

-- Unnecessary writes are bad, especially for SSDs.

-- Wastes clock cycles, time, moving data from RAM to PF and from PF to RAM

I must apologize but arguing with me is kinda hard because I have ran without a PF XP (3 Systems), 2003, 7 x64, XP x64, and 7 x32 (Current).

You still haven't read my post. Go back and read it please, I've already explained all this. Just about every point you've made there is either completely wrong, or the conclusion you've reached is flawed in some major way.

You know what, here, I'll quote the whole thing for you. Read it carefully this time.

1. It is perfectly safe to disable your page file in many cases. Disabling the page files DOES NOT cause any weird issues. [...] The argument that things will not run without a page file is for the most part is a lie - it most likely, if at all, applies only to legacy applications that have long been discontinued.

You obviously have no idea what the page file is actually used for. Please stop spreading this misinformation, disabling the page file is a bad thing no matter how you slice it.

Some applications reserve far more address space than they'll ever use (in some cases, gigabytes more). When this space is reserved, it means nothing else can use it. Normally Windows will attempt to allocate this reserved (yet empty) space to the pagefile. Without a pagefile, Windows is forced to allocate physical RAM. By disabling the page file you are decreasing your effective total amount of physical RAM in a lot of cases.

The pagefile is also where Windows moves unused pages. If something was loaded into RAM and then forgotten about by the host program, Windows will move that misfit data to the pagefile (this is so the data is still available but not eating up resources forever). Without a pagefile, Windows is forced to keep that information resident in RAM. Once again, you are decreasing your effective total amount of physical RAM in a lot of cases.

As you yourself mentioned, if you run out of physical memory, you're going to run into problems. Running a couple of virtual machines will reserve enough RAM to push you over the limit real quick (and VMWare crashes pretty nastily if you run out of RAM and don't have a page file). Games can easily push you over your memory threshold as well. As previously mentioned, due to the two factors above, you'll now have varying degrees of RAM locked up as reserved, so your other applications will now have even less space to use before you run out of physical RAM to allocate.

Oh, and there are quite a few applications that will refuse to run without a pagefile (I'm guessing you didn't even bother to research). Photoshop is a pretty big one, it used to just crash instantly without a pagefile (CS1), but now I believe it will just warn you and close.

Speaking of research, did you even bother to look up benchmarks to see if disabling the pagefile has any real world benefit performance-wise? As far as i can tell, disabling it will only hurt performance and stability overall.

Superfetch also becomes less helpful if you never restart as Windows for a very long time now - keeps parts of any application you ran in memory thus greatly increasing application reload times

You think reloading an application from RAM takes longer that reloading from a hard disk? wow...just wow... I can't even begin to believe how completely backwards your thought process is here...

Also, that information is only kept there as long as nothing else needs the space. If you load something large, Windows kicks little-used cache out of RAM to make room.

2. Superfetch is not by any means a crucial service. It is safe to disable. All it does is preload applications that you use in RAM, without your control of thereof. If you have an SSD it will be turned of, in most cases, but if you have an SSD with slow read speeds - it will remain on. Which actually doesn't make sense as superfetch doesn't help with write speeds.

I never said it was a crucial service... The important part here is that you should leave it in whatever state Windows decides it should be in (which is "on" for most people). In the case where Windows has automatically detected Superfetch should be on, you will only hurt performance by turning it off.

Superfetch caches resources that are loaded most often by the user so that they are always available (so in a sense, you do have control over it). Superfetch only uses RAM that would otherwise be free (if a large app loads and needs space, superfetch will purge itself from RAM to make room). I see no reason why you would EVER want to disable it on a system with a hard disk.

3. This makes no sense. Disabling something that you do not use from loading should increase speeds. It is like saying that a clean boot would slow down your boot time.

Once again, you demonstrate you have no clue what you're talking about, nor how it works. The ReadyBoost service does a lot more than you seem to think, you use it every boot (as i said, even without a flash drive dedicated to ReadyBoost).

ReadyBoot (the replacement for Windows XP's boot prefetcher) has been combined with the ReadyBoost service. ReadyBoot has been shown to improve boot times over the old Windows XP boot prefetcher by about 20%.

ReadyBoot uses features of ReadyBoost to work, so disabling ReadyBoost will kill ReadyBoot (harming your boot times greatly).

Here's a highly in-depth look at how ReadyBoot works, how it improves performance, and how it ineracts with ReadyBoost.

Windows Vista uses the same boot-time prefetching as Windows XP did if the system has less than 512MB of memory, but if the system has 700MB or more of RAM, it uses an in-RAM cache to optimize the boot process. The size of the cache depends on the total RAM available, but is large enough to create a reasonable cache and yet allow the system the memory it needs to boot smoothly.

After every boot, the ReadyBoost service (the same service that implements the ReadyBoost feature just described) uses idle CPU time to calculate a boot-time caching plan for the next boot. It analyzes file trace information from the five previous boots and identifies which files were accessed and where they are located on disk. It stores the processed traces in %SystemRoot%\Prefetch\Readyboot as .fx files and saves the caching plan under HKLM\System\CurrentControlSet\Services\Ecache\Parameters in REG_BINARY values named for internal disk volumes they refer to.

The cache is implemented by the same device driver that implements ReadyBoost caching (Ecache.sys), but the cache's population is guided by the ReadyBoost service as the system boots. While the boot cache is compressed like the ReadyBoost cache, another difference between ReadyBoost and ReadyBoot cache management is that while in ReadyBoot mode, other than the ReadyBoost service's updates, the cache doesn't change to reflect data that's read or written during the boot. The ReadyBoost service deletes the cache 90 seconds after the start of the boot, or if other memory demands warrant it, and records the cache's statistics

So, what I said stands; DO NOT disable the pagefile, DO NOT disable SuperFetch (if Windows has decided it should be on), DO NOT disable ReadyBoost. Disabling these three will only hurt performance.

Please Udedenkz, I beg of you, stop spreading misinformation.

Edited by StarLion

Share this post


Link to post
Share on other sites
Udedenkz    51

Oh, I do then. I guess I will remove it from my other recommendations , it is better to disable Updates ( so to have it manual ) through the Microsoft Management Console.

I tackled your arguments in post 75. You are arguing pretty much that stuff goes to PF leaving more free RAM when PF exists. This is a non-argument as I am saying that the PF should be disabled if having enough free RAM is a non-issue. Then you are mentioning running out of memory because there is no page file, which is something that I confirmed from my first post in this thread. But I also must note that disk thrashing is not good.

Next you go into the software. To which I simply say - if it doesn't run without a PF, you might not have enough memory without a PF. Or you need to get the latest updates / patches. Or you need to stop using that program for being a piece of ****. Or use newer versions such as CS4 for example. I also point out games like Crysis and their flawless performance without a PF.

Then you mention that it hurts stability, without actually giving any method for me to reproduce this loss of stability, on which I call you out to either stop posting or show me how to reproduce a problem (aside from what I acknowledged from the begging),

Windows Registry Editor Version 5.00

;clean disabled entries from startup
[-HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Shared Tools\MSConfig\startupreg]

Share this post


Link to post
Share on other sites
StarLion    4

Seriously, give it a rest. If you can't be bothered to read my posts, so be it. You've made it abundantly clear that it's worthless explaining anything to you, as you'll simply breeze right over anything contradictory to your little reality and post back with barely decipherable broken-English.

I've backed up my claims, you simply aren't reading. I can't help that. I can, however, keep beating the facts over your head.

Lets take a look at your little list.

-- The PF is not anywhere as efficient as physical memory

Well of course it isn't, but having access to slow virtual memory is better than running out of memory entirely (which will lead to all sorts of lapses in stability).

Having a pagefile present also compliments the performance of RAM. I explain how below.

-- Reliance on the PF does not mean anything good

Actually it does. Windows will attempt to keep actual RAM (which as you noted, is more efficient) free by moving data that hasn't been accessed in a great deal of time to the pagefile. This keeps RAM free to work with applications that actively require fast access.

The pagefile also allows Windows to map empty reserved address space somewhere other than physical RAM. Let me give an example of what this means in day=to-day usage.

If an application decides to reserve 2GB of RAM but only use 1GB, that means 1GB of the reserved address space isn't being used. Other applications can't use this reserved space, even though it's empty. If you have a pagefile, Windows can map that 1GB of empty space to the pagefile instead of physical RAM. This means you suddenly have 1GB more usable RAM with a pagefile than without one (when running said application).

That said, a few throw-away system caches such as disk cache can occupy this reserved address space, but they rarely use enough memory to actually fill it, meaning you still have wasted (reserved but empty) space in physical RAM if you don't have a pagefile.

-- You need to upgrade, remove bloat, etc if your PF is being used

Wrong. As I've already explained in the above two points, the pagefile is used beneficially even if you have ample amounts of RAM.

-- The Page File wastes a good amount of space which can be utilized for storage.

Hard disks are $0.10 per gigabyte. DDR2 RAM is about $20 per gigabyte.

Hard disk space is cheap, RAM is not. In instances where the pagefile is being used to keep extra reserve space out of RAM, what you've basically said here is that we should waste gigabytes of a $20 per gigabyte resource instead of a $0.10 per gigabyte resource.

Why would I want to spend $20 per gigabyte just to waste it with reserved empty space? Or with files that haven't been accessed in long periods of time? I would much rather use that (very expensive) physical RAM for active applications that need it.

-- Unnecessary writes are bad, especially for SSDs.

True enough, but you're overlooking a legitimate performance improving tweak.

Buy a second hard disk for storage and put the pagefile on it. This way the extra reads and writes will not effect the performance of your OS/programs drive, and will not ware out your SSD as quickly. This is also significantly cheaper than buying a cartload of RAM, and also provides you with a secondary storage space for documents, ISO's, pictures, backups, or whatever else you want.

-- Wastes clock cycles, time, moving data from RAM to PF and from PF to RAM

The few clock cycles it takes to shift address space around is negligible and a non-issue.

I know I've said this one a bunch of times already, but I'm going to keep beating you over the head with it until it sinks in. First off, many times it isn't moving actual data into the pagefile, it's moving empty reserved address space so it doesn't eat up physical RAM. When the OS does decide to move actual data to the pagefile, it means that you've either run out of physical RAM (in which case you do need your pagefile) or you haven't accessed that data in so long that the OS has decided that the RAM it's taking up is put to better use for an active application or superfetch's cache.

Share this post


Link to post
Share on other sites
Udedenkz    51
*snip*

I don't think you realize this, but in post #55 I clearly said "if go over your physical memory limit the OS will kill the memory leeching app."

Your second argument is also something acknowledged in post #55. It is obvious how that is an advantage, but my argument still holds - if you never used more System Memory than you have RAM, it is safe to disable the page file. For example, if a user has 3GB RAM and a 3GB PF, but the user never sees his System Memory usage about 1.7GB, it is safe to disable the PF as memory used by that user never exceeds his RAM limit of 3GB.

To put it simply, reserved but empty space will still be whether you have a PF or not. You are making an argument where there is none as you are making an argument that does not apply to when it is advised to disable the PF: if used memory does not exceed the total amount of RAM the system has - the PF is unnecessary. With or without the PF, Windows will still manage memory and free up unused memory if it becomes necessary - it will just have less to work with.

If I was to right now enable the PF, for some insane reason, I would not be able to install TF2. Quite honestly disabling Hibernate and the PF, seriously frees up a major amount of space which then can be used for other things that are more important.

You haven't said anything at all that even dares to imply that the PF is a performance tweak - it is still the same old argument "the PF gives you more RAM to work with". Also, I use a Dell Mini 311 - so I can't use two HDs.

Every user that met the conditions for disabling the PF, should try it and see for her/him-self whether or not there are problems.

Windows Registry Editor Version 5.00

; Removes Dependencies of Audio Service on other services
; Allows Audio Service to run independently
; Good if you can't start Audio Service due to dependency problems.

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\Audiosrv]
"DisplayName"="@%SystemRoot%\\system32\\audiosrv.dll,-200"
"Group"="AudioGroup"
"ImagePath"=hex(2):25,00,53,00,79,00,73,00,74,00,65,00,6d,00,52,00,6f,00,6f,00,\
  74,00,25,00,5c,00,53,00,79,00,73,00,74,00,65,00,6d,00,33,00,32,00,5c,00,73,\
  00,76,00,63,00,68,00,6f,00,73,00,74,00,2e,00,65,00,78,00,65,00,20,00,2d,00,\
  6b,00,20,00,4c,00,6f,00,63,00,61,00,6c,00,53,00,65,00,72,00,76,00,69,00,63,\
  00,65,00,4e,00,65,00,74,00,77,00,6f,00,72,00,6b,00,52,00,65,00,73,00,74,00,\
  72,00,69,00,63,00,74,00,65,00,64,00,00,00
"Description"="@%SystemRoot%\\System32\\audiosrv.dll,-201"
"ObjectName"="NT AUTHORITY\\LocalService"
"ErrorControl"=dword:00000001
"Start"=dword:00000002
"Type"=dword:00000020
"DependOnService"=hex(7):52,00,70,00,63,00,53,00,73,00,00,00,00,00
"ServiceSidType"=dword:00000001
"RequiredPrivileges"=hex(7):53,00,65,00,43,00,68,00,61,00,6e,00,67,00,65,00,4e,\
  00,6f,00,74,00,69,00,66,00,79,00,50,00,72,00,69,00,76,00,69,00,6c,00,65,00,\
  67,00,65,00,00,00,53,00,65,00,49,00,6d,00,70,00,65,00,72,00,73,00,6f,00,6e,\
  00,61,00,74,00,65,00,50,00,72,00,69,00,76,00,69,00,6c,00,65,00,67,00,65,00,\
  00,00,53,00,65,00,49,00,6e,00,63,00,72,00,65,00,61,00,73,00,65,00,57,00,6f,\
  00,72,00,6b,00,69,00,6e,00,67,00,53,00,65,00,74,00,50,00,72,00,69,00,76,00,\
  69,00,6c,00,65,00,67,00,65,00,00,00,00,00
"FailureActions"=hex:80,51,01,00,00,00,00,00,00,00,00,00,03,00,00,00,14,00,00,\
  00,01,00,00,00,60,ea,00,00,01,00,00,00,c0,d4,01,00,00,00,00,00,00,00,00,00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\Audiosrv\Parameters]
"ServiceDll"=hex(2):25,00,53,00,79,00,73,00,74,00,65,00,6d,00,52,00,6f,00,6f,\
  00,74,00,25,00,5c,00,53,00,79,00,73,00,74,00,65,00,6d,00,33,00,32,00,5c,00,\
  41,00,75,00,64,00,69,00,6f,00,73,00,72,00,76,00,2e,00,64,00,6c,00,6c,00,00,\
  00
"ServiceDllUnloadOnStop"=dword:00000001

Share this post


Link to post
Share on other sites
StarLion    4
I don't think you realize this, but in post #55 I clearly said "if go over your physical memory limit the OS will kill the memory leeching app."

I don't thing you realize this, but that is obviously a bad thing.

Applications randomly being closed when you run out of memory could lead to all sorts of data loss. What if you're working on something important and you run out of RAM? it could just vanish right out from under you if you haven't saved recently. Why would you want to work on such a system?

If I was to right now enable the PF, for some insane reason, I would not be able to install TF2. Quite honestly disabling [...] the PF, seriously frees up a major amount of space which then can be used for other things that are more important.

More important things? Storing just one more game is more important than the stability of your system?

Also Team fortress 2 is around 3GB in size. Are you saying you're running with less than 3GB of free disk space? That alone will hurt performance on mechanical hard disks, and increase the amount of writes that ware-leveling has to do on SSD's

My argument still holds - if you never used more System Memory than you have RAM, it is safe to disable the page file. For example, if a user has 3GB RAM and a 3GB PF, but the user never sees his System Memory usage about 1.7GB, it is safe to disable the PF as memory used by that user never exceeds his RAM limit of 3GB.

To put it simply, reserved but empty space will still be whether you have a PF or not. You are making an argument where there is none as you are making an argument that does not apply to when it is advised to disable the PF: if used memory does not exceed the total amount of RAM the system has - the PF is unnecessary. With or without the PF, Windows will still manage memory and free up unused memory if it becomes necessary - it will just have less to work with.

You keep forgetting one important fact, the situation you're describing isn't even possible (unless you disable superfetch).

Superfetch will keep as much RAM as possible full with cached resources from the programs you use most often. This allows those cached programs to open faster (loading from RAM is faster than loading from disk), while also marginally increasing the battery life of laptops by preventing the hard disk from spinning up as often. This runs as low priority, meaning it will dump superfetch cache from RAM if another application needs the space. No downsides to having this service, all upsides.

Now, the presence of superfetch compleatly invalidates your point. If you have 3GB of RAM, you will be using close to 3GB at all times, not 1.7GB.

If you disable the pagefile, all the abandoned resources and empty reserved space that Windows likes to keep there in order to free up RAM fill be forced back into physical RAM. This means you will see higher memory usage without a pagefile (which makes it easier to fill your RAM, and easier to run out of it). This also means less space will be available for superfetch and disk cache. Having less data cached by superfetch WILL harm system responsiveness.

Bottom line: the pagefile is still beneficial even when you have ample amounts of RAM.

You haven't said anything at all that even dares to imply that the PF is a performance tweak - it is still the same old argument "the PF gives you more RAM to work with". Also, I use a Dell Mini 311 - so I can't use two HDs.

Yes I have, you simply aren't reading. I've re-itterated a number of reasons why having a pagefile will improve performance in this very post. Please read this one carefully.

Every user that met the conditions for disabling the PF, should try it and see for her/him-self whether or not there are problems.

No they shouldn't. Please stop spreading this mindless FUD.

I've refuted you over and over, stop willfully ignoring the points I've made that shake the foundations of your little pagefille-less world. There is no good reason to disable the pagefile, and a whole lot of reasons to keep it enabled.

Share this post


Link to post
Share on other sites
Udedenkz    51
*snip*

Obviously it is a bad thing and a valid reason to use the page file. Also, just to make it clear, this only happened in XP - I think W7 will attempt to do a lot more to make sure this doesn't happen.

Yes because I do not believe I am compromising the stability of my system. Also, thanks for pointing out that it was 3GB+ - I have 6GB so I can install it. :) Also Windows informs you when you are running out of space, so the amount of space I am going to have after I download TF2 (1.7GB) is plenty as Windows only informs you when you have like 200MB left - that is when it might be time to clean your recycle bin.

After boot, and some idle time, my RAM usage remains under 500MB without any PF and with superfetch enabled. Because memory used by superfetch could be considered empty as superfetch will free up all that RAM if there is any need.

That is the point of disabling the page file though, to force windows to use the abundance of RAM for everything.

If you disable the pagefile, all the abandoned resources and empty reserved space that Windows likes to keep there in order to free up RAM fill be forced back into physical RAM. This means you will see higher memory usage without a pagefile (which makes it easier to fill your RAM, and easier to run out of it

I must once again, tell you that you are going back to "you will have less RAM argument" - you are on version 3 or so, but in essence it is the same argument.

This also means less space will be available for superfetch and disk cache. Having less data cached by superfetch WILL harm system responsiveness.

This is the most BS statement I have heard you make. And that is aside that in the same righteous sweep of words you demonize those that want Windows to be fully RAM dependent (by promoting the Swap File) while promoting greater reliance on RAM (by supporting superfetch). I see a major contradiction- you want RAM to be used, but on the other hand you DO NOT.

A properly maintained Windows 7 install will use about 400-600MB RAM, the rest will be cached. Lets say a system has 3GB RAM, then on a clean boot it will have 2.6 - 2.4GB RAM ( less w. memory allocated for the GPU ) therefore there will be ~2.5GB for superfetch to fill with applications that you use. I am sorry but, who the hell has that much s**t in the first place? I just want it to preload about 5 applications that I use often, that doesn't require a lot of RAM at all! Once again this is just the same argument of having less memory.

Also this argument does not apply to those that either have an SSD.

Also the argument that without a PF one will have less memory, which you have wrote here a few times does not apply to the argument that the PF is the holy grail of performance.

And obviously people need to try it themselves! That is how one can confirm for his/herself whether or not you are full of air. I mean I was skeptical that turning off the PF is a bad idea too at the start, I can understand your fear, but then I tried it and never looked back.

I am going to ask you to stop posting as it is self evident that this debate is going in circles. Surely you have approached the same argument from many sides but this is getting boring - at least start posting Task Manager screenshots! Or try to reason why I haven't had any problems since summer 2008 running without a page file. Also keep in mind that you are making an argument against Windows as Ubuntu, for example, runs just fine without a PF.

Contributing to this Thread,

I recommend W7 x86_32 every user to install Unlocker as it not only "unlocks" files but you can also you delete files by Trusted Installer! For example, Windows Mail in Program Files - you do not have the permission required to delete that folder, so you can just right click - unlocker - delete. Presto! :D

Edited by Udedenkz

Share this post


Link to post
Share on other sites
svnO.o    26
And obviously people need to try it themselves! That is how one can confirm for his/herself whether or not you are full of air. I mean I was skeptical that turning off the PF is a bad idea too at the start, I can understand your fear, but then I tried it and never looked back.

I am going to ask you to stop posting as it is self evident that this debate is going in circles. Surely you have approached the same argument from many sides but this is getting boring - at least start posting Task Manager screenshots! Or try to reason why I haven't had any problems since summer 2008 running without a page file. Also keep in mind that you are making an argument against Windows as Ubuntu, for example, runs just fine without a PF.

My thoughts exactly. If your system is rock stable without it/you don't notice any difference and you got plenty of ram to begin with then why does it matter?

Share this post


Link to post
Share on other sites
StarLion    4
I do not believe I am compromising the stability of my system.

What you "believe" contradicts reality. Stability is effected if you do not have a pagefile weather you want to believe it or not. Toms Hardware did extensive testing, here's a quote from their article on the subject (this was with Windows Vista):

[...]applications immediately crash as soon as the system runs out of RAM and Windows is unable to write the files to virtual memory on the hard drive. This situation is especially critical if the memory shortage hits one of the Windows system components - that may quickly result in the Windows Aero turning itself off as a result of the graphics driver crashing. The only remedy in such a situation is rebooting the system.[...]

There you go, reduced stability. They also note that disabling the pagefile didn't increase performance beyond the margin of error inherent to synthetic benchmarks. This means that disabling it offers no advantage, only disadvantages (aside from gaining 500MB or more disk space). Windows can't even do a crashdump without at least a 400mb pagefile.

Also Windows informs you when you are running out of space, so the amount of space I am going to have after I download TF2 (1.7GB) is plenty as Windows only informs you when you have like 200MB left - that is when it might be time to clean your recycle bin.

You completely missed the point. Windows might only inform you when you have 200mb left, but you'll start experiencing negative effects on performance quite a while before that.

On a mechanical drive, having so little free space means the hard disk cannot write data in contiguous blocks in an optimal manner; data ends up spread out and fragmented easily. On an SSD, having so little free space means that ware leveling is forced to do more write operations on the drive to swap around data from heavily used blocks to lightly used blocks, eventually shortening the life of the drive.

After boot, and some idle time, my RAM usage remains under 500MB without any PF and with superfetch enabled. Because memory used by superfetch could be considered empty as superfetch will free up all that RAM if there is any need.

It can be considered available for other applications to use, which is NOT the same thing as free. While the space is in use by superfetch it is being used to improve system performance, even though it is available for other active applications to use if needed. This means the space available to superfetch has an effect on performance.

That is the point of disabling the page file though, to force windows to use the abundance of RAM for everything.

Which doesn't work how you think it does (and you keep ignoring my explanation for some reason, which is becoming extremely annoying). As I've said before, Windows will allocate reserved, yet empty, address space to the pagefile to prevent it from sucking up RAM for no purpose. There is no performance hit for allocating this empty address space to the pagefile, all it does is leave more physical RAM available for use by actively running applications.

Windows will also move resources that have not been accessed in a great deal of time to the pagefile. What's important to remember about these old resources is that, if you use them often, they'll be cached in memory by superfetch anyway (so no problem). If these resources are being moved to the pagefile, it means nothing is currently using these abandoned resources, and they weren't loaded by any application use use commonly (which means the resource would have to be re-loaded from disk ANYWAY). There's little point to keeping such resources in RAM, as that space could obviously be better used for superfetch to cache relevant files and improve performance.

These are the facts. It seems to me you're arguing without knowing the ins and outs of Windows memory management. You're spreading FUD and you need to stop.

I must once again, tell you that you are going back to "you will have less RAM argument" - you are on version 3 or so, but in essence it is the same argument.

And I'm going to keep restating it until it sinks in. Because it's correct. You need to accept the facts of the matter.

My response to the quote previous to this one goes into great detail on why you will have less physical RAM available for more useful services. I'm going to have to ask you to read it very carefully this time and fully comprehend it before responding, because you've still failed to do so thus far.

This also means less space will be available for superfetch and disk cache. Having less data cached by superfetch WILL harm system responsiveness.

This is the most BS statement I have heard you make. And that is aside that in the same righteous sweep of words you demonize those that want Windows to be fully RAM dependent (by promoting the Swap File) while promoting greater reliance on RAM (by supporting superfetch). I see a major contradiction- you want RAM to be used, but on the other hand you DO NOT.

There's no contradiction, you just lack basic reading comprehension skills.

I want RAM to be used, obviously, but I want it to be used optimally for data that make the most sense. That means dumping some data and allocating some address space to the hard disk to keep more space free for relevant data in physical RAM. The criteria Windows does these operations based upon have been explained to you ad nauseum, it makes perfect sense to keep the pagefile around as a performance and stability enhancing measure.

A properly maintained Windows 7 install will use about 400-600MB RAM, the rest will be cached. Lets say a system has 3GB RAM, then on a clean boot it will have 2.6 - 2.4GB RAM ( less w. memory allocated for the GPU ) therefore there will be ~2.5GB for superfetch to fill with applications that you use. I am sorry but, who the hell has that much s**t in the first place? I just want it to preload about 5 applications that I use often, that doesn't require a lot of RAM at all! Once again this is just the same argument of having less memory.

Why wouldn't you want to cache as much as possible in order to make as many applications as possible load more quickly? You make it sound like you don't want to use all of your RAM optimally.

I have 8GB of RAM, and superfetch is currently caching 6.8GB, so apparently I have "that much s**it." Firefox, photoshop, steam, a media player, some office applications...I could easily see using up quite a lot of RAM for cached data. I mean, even a default install of Windows is larger than that, so even on the cleanest install there's still a lot of data that could be cached in order to improve system responsiveness and reduce hits to the hard disk.

Also this argument does not apply to those that either have an SSD.
Yes it does. And unlike you, I'll back up my statement.

SSD's are still slower than RAM, some of them are actually slower than mechanical drives in certain regards. Superfetch will still improve performance even if you have an SSD, though the results may not be as drastic as with a mechanical hard disk. Superfetch will intercept reads to data that reside in cache, reading from RAM instead of hitting the SSD. This results in an increase in net bandwidth in such situations which helps both mechanical drives and solid state disks.

Also the argument that without a PF one will have less memory, which you have wrote here a few times does not apply to the argument that the PF is the holy grail of performance.
How so? With a pagefile you have more available RAM for resources that matter, which will increase performance.

I don't believe I ever called it the "holy grail", but it certainly DOES NOT HURT to have it enabled...really, there's no reason to disable it in the first place.

And obviously people need to try it themselves! That is how one can confirm for his/herself whether or not you are full of air. I mean I was skeptical that turning off the PF is a bad idea too at the start, I can understand your fear, but then I tried it and never looked back.

You have yet to list a credible reason for disabling it, so no. Don't tell people to waste their time.

There's no "fear" here, just facts, and all the facts say that there are no real negatives to having it enabled. You really need to read up on the Windows memory management subsystem before you join a debate heavily entrenched in it.

Share this post


Link to post
Share on other sites
Udedenkz    51
* TEXT *

That reduced stability argument is still an argument from not having enough RAM to actually disable the PF. Also that is Vista - I did not experience anything of sort when I hit the roof on 7 and XP just kills the application with the memory leak.

FYI: Windows doesn't need much extra drive space aside from what it has to run. I had a few instances when I got to 100-5MB space. Nothing much happened - Windows informed me and I freed up 300MB or so. There was no performance loss, there was just either Windows informing me, me noticing, or an application not being able to save something to disk because of a lack of space. Windows doesn't write a lot to disk - some writes to registry, some writes to log files, some writes to thumbnail cache. I would advise you to run with about 200MB (your argument might work for less) space left to realize this fact - no change what so ever. I don't know how this impact performance if running with the swap file as I do not use it - and if it does negatively effect the swap file, it would undermine your own statements.

Next paragraph does not address my previous post.

I am not ignoring you - I am reading all that you say. This paragraph is also an argument from not having enough RAM.

The next paragraph is also about not having enough RAM.

Stop raging, Furry - argument from not having enough RAM do not apply to me.

All your arguments can be summarize in two sentences,

- You will have less available RAM without a PF

- If you run out of RAM, you will be sorry

You haven't shown any stability problems that result from running without a PF. You also haven't show that disabling the PF will cause a performance decrease. See above, all your argument stem from having less RAM without a PF.

For the two paragraphs on superfetch,

I already told you why the longer your computer stays on the more nullified the effects of superfetch become. I am not repeating myself.

On SSDs, you won't notice the difference either way. I didn't notice ANY difference with it on or off on my previous setup (first gen slow SSD!!).

The next two lines lack consistency "will increase performance", DOES NOT HURT to have it enabled

My reasons for disabling it still remain unchanged.

I already did - one of which may give you credibility, unless - that is - you are scared that it will reduce your credibility.

Read It Up VS See For Your Self.

Share this post


Link to post
Share on other sites
Knife Party    629

@ both Udedenkz and StarLion -> can you two somehow contibute something to the thread ? Look at the thread title it says 'Tip and Tricks' not far fetched registry hacks that provide zero gain to the average user m'kay? Stop ruining this thread with senseless rambling of utter nonsense, you come to read some 'tips'...but no....you have to see you two bikering about silly hacks and dodgy 'tweaks'.....

keep it Relevant please... :hmmm:

Share this post


Link to post
Share on other sites
StarLion    4
FYI: Windows doesn't need much extra drive space aside from what it has to run. I had a few instances when I got to 100-5MB space. Nothing much happened - Windows informed me and I freed up 300MB or so. There was no performance loss, there was just either Windows informing me, me noticing, or an application not being able to save something to disk because of a lack of space. Windows doesn't write a lot to disk - some writes to registry, some writes to log files, some writes to thumbnail cache. I would advise you to run with about 200MB (your argument might work for less) space left to realize this fact - no change what so ever. I don't know how this impact performance if running with the swap file as I do not use it - and if it does negatively effect the swap file, it would undermine your own statements.

You have, once again, ignored me. I've just told you why having so little space will harm performance (and possibly harm the lifespan of your drive). The "well, I personally don't notice it, so it isn't a problem!" argument simply doesn't work. Believe me, your hard disk won't be happy with you if it only has a couple of gigabytes remaining; ANYTHING written to the disk in such circumstances will have to be swapped in and out of that last little bit of space in whatever location it might happen to be on the hard disk (which will most likely not be optimal placement under such circumstances).

I give a giant list of reasons with sound backing and reasonable explanations, and you just ignore it...

Next paragraph does not address my previous post.

I am not ignoring you - I am reading all that you say. This paragraph is also an argument from not having enough RAM.

The next paragraph is also about not having enough RAM.

Stop raging, Furry - argument from not having enough RAM do not apply to me.

Still ignoring every point that I've made. Sorry if you don't like the truth, but this level of denial is just sad.

All your arguments can be summarize in two sentences,

- You will have less available RAM without a PF

- If you run out of RAM, you will be sorry

You haven't shown any stability problems that result from running without a PF. You also haven't show that disabling the PF will cause a performance decrease. See above, all your argument stem from having less RAM without a PF.

Wrong, you haven't read. I quite clearly stated a number of circumstances where having a pagefile will improve performance even if you have massive quantities of RAM, because other performance-improving services will take advantage of the space it clears up for information that improves system responsiveness.

I don't know how you can miss so much, especially after having it beaten over your head for pages and pages. (you're starting to look like a forum troll, actually).

For the two paragraphs on superfetch,

I already told you why the longer your computer stays on the more nullified the effects of superfetch become. I am not repeating myself.

Good, because that was wrong as well. The length of time the machine has been on does not decrease the effectiveness of superfetch.

On SSDs, you won't notice the difference either way. I didn't notice ANY difference with it on or off on my previous setup (first gen slow SSD!!).

I'm going to have to chalk that up to you simply not being very observant. There is a difference, sorry you quote can't tell, but quite a few can, myself included. I've put a fairly fast SSD in my laptop, and there is a highly noticeable improvement with Superfetch on as apposed to off. There's a bandwidth bottleneck due to the machine only having SATA 150, so any extra places I can pull bandwidth from will help on large loading operations. If you knew how these services really worked you would clam up.

The next two lines lack consistency "will increase performance", DOES NOT HURT to have it enabled

My reasons for disabling it still remain unchanged.

Seems consistent to me, if something increases performance, then it obviously isn't hurting...

At this point, I have to come to one of two conclusions. You're either incredibly dense, or you're putting me on (the lack of reading comprehension and trying to make arguments where there are none looks suspiciously like trolling). Please, give it a rest. As Kaboose said, we're all getting tired of these senseless supposed "tweaks" that only serve to harm rather than help.

---------------- end of rant ----------------

Tip #1: Creativelement power tools suite:

Now, if I may contribute something genuinely useful to the thread. Creativelement power tools work with Windows 7 about as well as they work with Vista, offering some interesting tweaks and powerful management interfaces (like a powerful file type editor). It's a 45 day trial, but it doesn't nag you and the tweaks will stay applied after, so as long as you get everything set the way you want it in the first 45 days, you simply don't have to purchase it.

Tip #2: Kallout system-wide accelerators:

During the early experimental stages of Windows 7 (before build 6801) they had integrated a feature that provided system-wide accelerators to all applications on Windows. Previously, accelerators had been a feature exclusive to internet explorer, allowing you to highlight some text and instantly be provided with actions such as search, blog, write in a journal, etc. You can probibly imagine how useful that would be if it worked in every application, document editors for example.

Well, that feature got cut from Windows 7, but that's where Kallout comes in. It's a small application that runs in your system tray and provides functionality similar to accelerators for a wide range of applications. Link to Kallout home page.

Edited by StarLion

Share this post


Link to post
Share on other sites
Udedenkz    51
* AIR *

Really a few gigabytes empty? REALLY? Tell that to a netbook user with a 4GB SSD - don't make me laugh!

The only reason to do that - that I can even conceive of - is for a dynamic swap file. Once again, Windows doesn't need insane amounts of Hard Drive space to run - applications might. Photoshop has its own swap file and it uses it if it so becomes necessary - your argument would apply there if you use gigantic files (that Photoshop cannot process in RAM alone).

I not only didn't notice any performance difference but there also was NO difference in SSD usage.

I have dedicated at least 1.5 hours to every one of your silly fears, and my arguments still stand.

"number of circumstances," re-read this,

All your arguments can be summarize in two sentences,

- You will have less available RAM without a PF

- If you run out of RAM, you will be sorry

Also, if you don't know that Windows loads parts of applications that you used into RAM and - after the application was closed - still keeps them in memory then you should stop talking. This is self evident for XP, Vista, and 7 users. Open Firefox, time it, close it, open it again, and time it - it will start faster then second time. Although honestly I would use the Firefox Preloader over any dll in memory feature, ~0.3 second Firefox startup time on a an Atom - if I wasn't using two Firefox distros at once.

FYI - when I have a 4GB system, I had XP and 7 - 7 (with superfetch & readyboost) was slower than XP. Now I use a netbook (2GB DDR3) - Atom + ION, and it comes preinstalled with 7 - surely with superfetch things are not bad - but I wouldn't doubt that XP was still faster. So much for your fancy superfetch.

I guess for you ideal RAM usage is to preload your entire HD into RAM and only communicate with the HD when writing something - you sure make it sound like this would be ideal. Sure... see there is this point where this RAM thing becomes silly, you are at it. ~200MB RAM would be plenty to preload some common used executables.

And obviously people need to try it themselves! That is how one can confirm for his/herself whether or not you are full of air. I mean I was skeptical that turning off the PF is a bad idea too at the start, I can understand your fear, but then I tried it and never looked back.

I am going to ask you to stop posting as it is self evident that this debate is going in circles. Surely you have approached the same argument from many sides but this is getting boring - at least start posting Task Manager screenshots! Or try to reason why I haven't had any problems since summer 2008 running without a page file. Also keep in mind that you are making an argument against Windows as Ubuntu, for example, runs just fine without a PF.

Here is your instability,

th_Instability.jpg

Contributing To Thread,

NirSoft ShellExView is a good way to disable shell extensions other applications added that you do not want. You can also disable parts of the default shell that you do not want / use - be careful though.

"The ShellExView utility displays the details of shell extensions installed on your computer, and allows you to easily disable and enable each shell extension. "

Share this post


Link to post
Share on other sites
StarLion    4
Really a few gigabytes empty? REALLY? Tell that to a netbook user with a 4GB SSD - don't make me laugh!

I don't see how 4GB SSD's are relevant to this discussion. Windows 7 requires at least 16GB of free drive space, so SSD's that small are moot.

I have dedicated at least 1.5 hours to every one of your silly fears, and my arguments still stand.

Actually they don't, and I've explained why. You're in some hardcore denial right now.

Also...it took you 1.5 hours to respond to each of my points? That's some serious time invested when you don't even appear to be reading half of what I've written.

"number of circumstances," re-read this,
All your arguments can be summarize in two sentences,

- You will have less available RAM without a PF

- If you run out of RAM, you will be sorry

As I said, you're missing quite a few of my points. If you're not going to read, stop posting.

Also, you haven't actually refuted either of the points that you've bothered to comprehend. Do you want less RAM available for foreground applications and superfetch? because that's what it sounds like right now.

Also, if you don't know that Windows loads parts of applications that you used into RAM and - after the application was closed - still keeps them in memory then you should stop talking. This is self evident for XP, Vista, and 7 users. Open Firefox, time it, close it, open it again, and time it - it will start faster then second time.

Of course, because now that application is prefetched and possibly in disk cache. The problem is, you're missing a crucial difference between that and superfetch. When you load something large enough, it will kick that prefetch data out of RAM. When you close the large application that kicked that prefetch data out of RAM, one of two things could happen.

If it's designated to be in Superfetch's cache, it will be reloaded automatically so that it opens instantly the next time you call upon the application.

If it's not designated to Superfetch's cache, it will wait until the next time you launch the application to reload files from disk, leaving you with additional wait time.

Although honestly I would use the Firefox Preloader over any dll in memory feature, ~0.3 second Firefox startup time on a an Atom - if I wasn't using two Firefox distros at once.

Which overrides all of Windows memory management optimizations by forcing the application to appear active at all times, even when not being used. Talk about wasting RAM...

If you have ample amounts of RAM, as you say, then you shouldn't require preloaders like that. Superfetch will handle those kinds of tasks for you.

FYI - when I have a 4GB system, I had XP and 7 - 7 (with superfetch & readyboost) was slower than XP. Now I use a netbook (2GB DDR3) - Atom + ION, and it comes preinstalled with 7 - surely with superfetch things are not bad - but I wouldn't doubt that XP was still faster. So much for your fancy superfetch.

Speculation with no facts to back you up. Invalid point.

On moderately powered systems, Windows XP, Windows Vista, and Windows 7 all perform within a couple of percentage points of one another. On slower systems outfitted with a decent amount of RAM, Windows Vista and Windows 7 can actually outpace Windows XP due to their superior resource management. This is especially true if said slower system also has a slow hard disk.

I guess for you ideal RAM usage is to preload your entire HD into RAM and only communicate with the HD when writing something - you sure make it sound like this would be ideal. Sure... see there is this point where this RAM thing becomes silly, you are at it. ~200MB RAM would be plenty to preload some common used executables.

First of all, that would be the fastest solution if you had 1TB of RAM you weren't doing anything else with. It would just take forever to get all the data off the hard disk and into memory every boot (especially since Windows background-loads superfetch to prevent it from impacting performance). You would effectively be running from a very flexible RAMDisk, with all writes backed by non-volatile storage.

Also, why on earth would you want to limit it to 200mb? That's hardly enough for anything! Use all the RAM available, otherwise it's wasted. Your ideas on how memory management should work are a decade out of date. Get with the times.

Here is your instability,

th_Instability.jpg

And that proves...what, exactly? That you've managed to keep your system up for 1.5 days?

You're still using more RAM than you would be if you had a pagefile to offload old resources into, which means superfetch has less space to cache data in and your system isn't as responsive as it could be. Also, considering you only have 1.8GB of RAM to work with total, and only about 800 or so remaining, I would be highly inclined to keep a pagefile around to prevent nastiness from occuring when you run out of RAM. Firefox alone can eat up more RAM than you have left in that screenshot.

Share this post


Link to post
Share on other sites
Udedenkz    51
*snip*

There is ONE valid reason to have at least like 1GB of HD space available that is system related - mostly Vista and 7. This reason I accept, interestingly you could have made a much better argument with it present.

The question you asked about whether foreground application or superfetch should be giver priority is also a question from having less memory available without the PF. It is also a question with an obvious answer - OBVIOUSLY it is foreground applications!

That is also a variant of the having less system memory argument. Also stuff will get offloaded when you reach the max with a PF, same behavior.

Bullsh*t! The preloader does a much better job that keeping dlls in memory and superfetch - because it keeps FF actually running in the background. That is kinda the point, keep the thing idle and loaded in memory (seen in processes - firefox.exe) - to maximize the start speed. I disabled the preloader and has been using Firefox for a very long time w. superfetch enabled - it is not anywhere near the startup speed with the preloader...

I am using all RAM available by using superfetch and disabling the PF. Surely with the PF I have more system memory and more space for superfetch, too bad, I don't care. It is still an argument from having less memory without the PF! Having less memory for superfetch is an argument from having less memory without the PF. Having less memory for applications is an argument from having less memory without the PF. Having problems with RAM heavy applications is STILL an argument from having less memory without the PF. Again this is self evident, and whenever a Power User turns off the PF he/she realizes this. I have provided a screenshot from a stable system, no problem with memory management, no swap. Running multiple apps, games, etc without crashing / instability. Frankly, to make you politely stop posting. Once again, I present to you a simple challenge,

... try to reason why I haven't had any problems since summer 2008 running without a page file.

Firefox eats up about 40MB - 100MB RAM under normal usage (1 - 8 or so tabs), not sure about other browsers. It is using 58MB with 4 Neowin.net tabs open. Eclipse on the other hand, is a more RAM hungry creature - 100+ MB.

But, I must still point out that summarized and simplified there arguments are,

- You will have less available RAM without a PF

- If you run out of RAM, you will be sorry

EDIT:

For those interested in the preloader,

http://sourceforge.net/projects/ffpreloader/ :)

Edited by Udedenkz

Share this post


Link to post
Share on other sites
StarLion    4
The question you asked about whether foreground application or superfetch should be giver priority is also a question from having less memory available without the PF. It is also a question with an obvious answer - OBVIOUSLY it is foreground applications!

Stop right there, you misread again. I said "and", not "or".

The question was "Do you want less RAM available for foreground applications and superfetch?" Which is what would happen without a pagefile, both items would have fewer resources to work with.

As I've said many times, foreground applications have priority over superfetch, I'm not sure why you would even think I would ask such a thing. I know it's probably futile to keep asking, but please read more carefully. You're now unintentionally pulling arguments out of thin air because you're missing key words.

Bullsh*t! The preloader does a much better job that keeping dlls in memory and superfetch - because it keeps FF actually running in the background. That is kinda the point, keep the thing idle and loaded in memory (seen in processes - firefox.exe) - to maximize the start speed. I disabled the preloader and has been using Firefox for a very long time w. superfetch enabled - it is not anywhere near the startup speed with the preloader...

You're most likely seeing a tiny speed increase with the preloader because it's actually keeping firefox resident, which forgoes processing the resources when you call it forward because it does it when the preloader starts. Superfetch just keeps the resources in RAM, but still requires the application to perform whatever startup tasks it usually does. At that point, the preloader being faster has nothing to do with caching (the same resources are in RAM with both the preloader and superfetch), so it can't really be compared with superfetch apples-to-apples. They're two different things.

I can tell you, however, that the preloader is a constant and inflexible drain on resources, while superfetch is not. That preloader will always be eating up RAM, even if firefox is closed and another application could be using the space to do actual work. Unless the preloader is set up to do background loading, it's also harming your boot time by forcing Windows to load and process Firefox on every boot.

Remember when Microsoft used to use preloaders themselves for large suites like Microsoft Office, before all these advancements in resource management? Ever stop to think why they stopped using them? They increased boot times (which really just moved your wait time from one place to another) and were a constant, often unnecessary, drain on resources. You were calling the pagefile archaic, and now here you are bringing up preloaders! I can't believe how uninformed you are.

I am using all RAM available by using superfetch and disabling the PF. Surely with the PF I have more system memory and more space for superfetch, too bad, I don't care.

So you acknowledge that, with the pagefile enabled, you'll have more space available for foreground applications and cache...but you don't care?

Why wouldn't you care about having a more responsive system? Why wouldn't you care about wasting RAM on dead-wood that could otherwise be trimmed off to the pagefile to make resources available for things that could actually help foreground applications?

You're just ignoring the problem, again.

It is still an argument from having less memory without the PF! Having less memory for superfetch is an argument from having less memory without the PF. Having less memory for applications is an argument from having less memory without the PF. Having problems with RAM heavy applications is STILL an argument from having less memory without the PF.

And these are all relevant and valid concerns, I don't know how you can so casually disregard every single one...

Again this is self evident, and whenever a Power User turns off the PF he/she realizes this.

A power user wouldn't be dumb enough to disable it, as the risks outweigh the gai- wait, there aren't any gains! You still havn't come up with a good reason to run without one.

I have provided a screenshot from a stable system, no problem with memory management, no swap. Running multiple apps, games, etc without crashing / instability. Frankly, to make you politely stop posting.

All it proves to me is that you spend more time watch-dogging your memory usage than actually using your computer. Oh, it also shows that you've got some serious resource management issues. You're using almost 1GB of RAM with only 37 processes running, which is a clear symptom of Windows being unable to move overzealous preallocated address space and forgotten resources out of RAM and into an alternate storage space.

Your task manager shows that your system is, for lack of a better word, constipated.

Firefox eats up about 40MB - 100MB RAM under normal usage (1 - 8 or so tabs), not sure about other browsers. It is using 58MB with 4 Neowin.net tabs open. Eclipse on the other hand, is a more RAM hungry creature - 100+ MB.

But, I must still point out that summarized and simplified there arguments are

That's extremely light for Firefox. I'm currently seeing Firefox using 200MB of RAM with 8 tabs open, which is double your estimate.

Edited by StarLion

Share this post


Link to post
Share on other sites
Udedenkz    51
*snip*

You are still going with the memory argument. The more RAM a person has, the less relevant this argument becomes. The less RAM the person needs, the less relevant your argument becomes. Disabling the Page File is an feature of Windows. Windows will manage itself, it will just have less to work with.

That is kinda the point, keep FF loaded, while no showing it. Genius, and simple.

Again you are going into the argument of having less memory to work with.

A user with 8GB RAM or so might even get insulted by your preaching. As you are pretty much saying that that user does not have enough RAM to turn off legacy features. FYI: With a Page File, I would have 4GB System Memory - there no way I am better off w. 4GB System Memory than an 8GB user with 8GB System Memory.

I don't recall that I have show you my running processes. Also that number goes down if pressured. As I said Windows can manage itself just fine. My computer is still on, and my memory usage is now 707. If I let it idle it will go down further. Stop flaming Windows - it can manage itself just fine as running without a PF is an inherent windows feature. As I said before Windows keeps stuff in memory of previous opened applications, this is a plus. My challange still stands, try to reason why I do not experience any problems running without a PF.

Also I don't monitor my RAM usage.

Share this post


Link to post
Share on other sites
StarLion    4
You are still going with the memory argument. The more RAM a person has, the less relevant this argument becomes. The less RAM the person needs, the less relevant your argument becomes. Disabling the Page File is an feature of Windows. Windows will manage itself, it will just have less to work with.

Everything you've stated above is incorrect. I've already explained how it continues to help; even with vast quantities of RAM, the pagefile is still a relevant part of how Windows handles memory management.

You need to go back and re-read. Better yet, don't post again until you've done some research into how windows actually manages memory. You're clueless, and it shows.

That is kinda the point, keep FF loaded, while no showing it. Genius, and simple.

I made it quite clear why this was a bad idea. Go back and re-read.

Again you are going into the argument of having less memory to work with.

Yes, because it's a relevant argument. Why do you keep ignoring it?

A user with 8GB RAM or so might even get insulted by your preaching. As you are pretty much saying that that user does not have enough RAM to turn off legacy features. FYI: With a Page File, I would have 4GB System Memory - there no way I am better off w. 4GB System Memory than an 8GB user with 8GB System Memory.

First off, I myself am running a machine with 8GB of RAM (I'll give you a moment to pull your foot out of your mouth...)

It's not an insult to have a pagefile with this amount of RAM, it's common sense. The pagefile isn't a "legacy feature", it has evolved along with the rest of windows to help meet modern needs (as I have now explained to you countless times). Even if I doubled my system RAM to 16GB, I would still want a pagefile around.

I don't recall that I have show you my running processes.

The total number of running processes is listed at the bottom of the task manager screenshot you posted.

Stop flaming Windows - it can manage itself just fine as running without a PF is an inherent windows feature. As I said before Windows keeps stuff in memory of previous opened applications, this is a plus. My challange still stands, try to reason why I do not experience any problems running without a PF.

What are you going on about? I'm not flaming Windows, I'm simply stating how memory management works and why your views regarding disabling vital system resources are completely ignorant.

And yes, Windows will do its best to run without a pagefile if you forcibly disable it (it's enabled by default for a reason). This is not an optimal configuration, and creates all sorts of other issues. If you run out of physical memory, applications will instantly crash. The extra data that would have been kept in the pagefile now has to be kept in physical memory, leaving less space for superfetch cache, disk cache, and actively running applications.

There is no upside to disabling the pagefile. Stop suggesting it.

Even on systems with low amounts of hard disk space, it's worth the 500MB+ to have some form of protection. Really, the minimum free space requirement of 16GB accounts for the pagefile, so it shouldn't be a problem.

Also I don't monitor my RAM usage.

That's quite surprising. Considering how close you are to running out, I would expect you to end up crashing things left and right.

Oh, by the way, you STILL haven't listed any relevant reasons for disabling the pagefile. You've spent all this time baselessly attacking me without ever backing yourself up. At this point, all you are is a troll with no idea what you're talking about.

Share this post


Link to post
Share on other sites
Udedenkz    51

*Sigh*

I could make a list of a Thousand and One uses for 8GB of HD Space, but I retract, you don't have reasons to keep the PF aside for reasons having of more memory. Therefore, there is nothing, at all, that you have said or will say here that can, in any way, change my opinion on this matter. There is noting to discuss here. I disabled the PF as per OCZ forum, I had no problems w. multiple Operating Systems and PCs, therefore I will keep it disabled and suggest anyone without a large drive and adequate RAM to do the same. If you want to pursue this discussion further, pursue it with someone who gives a damn about your arguments / find them relevant. I DO NOT!

To further nail the point home,

th_Instab.jpg

Share this post


Link to post
Share on other sites
Subject Delta    108

I have 8GB of memory, and I never disable the pagefile, simply because I have never perceived a performance benefit from doing so. Windows enables it for a reason, and I see no reason to do disable it.

Share this post


Link to post
Share on other sites
StarLion    4
I could make a list of a Thousand and One uses for 8GB of HD Space

That's nice, but we're talking about 8GB of RAM. Why are you bringing up hard disk space?

You don't have reasons to keep the PF aside for reasons having of more memory. Therefore, there is nothing, at all, that you have said or will say here that can, in any way, change my opinion on this matter.

Actually, yes I do; I've listed a number of reasons over the course of this thread, you've chosen to ignore them while also vastly oversimplifying the issue. Lets take a look at all the points you've missed, shall we?

1. Applications will instantly crash if you run out of RAM without a pagefile.

2. Crashdumps do not work without a pagefile.

3. Reserved, yet empty, address space will be kept in RAM instead of put in the pagefile (this could waste multiple gigabytes of RAM holding nothing but zero's).

4. Unused and abandoned pages will be left in RAM rather than moved to the pagefile.

5. Points 3 + 4 result in less memory being available for applications, superfetch, disk cache, etc.

6. Preloaders like the one you use for Firefox undermine modern resource management and are not helpful in the slightest when other factors are taken into account.

7. lacking drive space is not an excuse to disable the pagefile, the minimum requirement of 16GB of hard disk space for Windows 7 includes space for a pagefile.

I've reiterated the above points countless times, but you've ignored them over and over. I don't believe I've ever encountered anyone with quite so much trouble with reading and comprehension...

There is noting to discuss here. I disabled the PF as per OCZ forum, I had no problems w. multiple Operating Systems and PCs, therefore I will keep it disabled and suggest anyone without a large drive and adequate RAM to do the same. If you want to pursue this discussion further, pursue it with someone who gives a damn about your arguments / find them relevant. I DO NOT!

There's plenty to discuss, unfortunately you appear to be incapable of carrying on a conversation. I'm sorry to hear you've chosen to run with reduced performance and stability. I must ask you, however, not to suggest to anybody else to disable the pagefile.

There's no up side to disabling the pagefile, only downsides. I've asked you a couple of times to state your reasoning behind disabling it, but so far you've come up with nothing. You're trying to argue with no backing, which simply doesn't work.

To further nail the point home,

th_Instab.jpg

What are you trying to prove with these taskmanager shots? That you can cherry-pick a bunch of lightweight applications to run for a screenshot? I see you have Photoshop CS4 available, you should try doing something that's actually resource intensive with it, it can chew through 4GB of RAM all by itself quite easily.

Edited by StarLion

Share this post


Link to post
Share on other sites
Udedenkz    51

Any mod can come in and tell use to ST*U any time.

There should be a topic "Does Running Without the Page File Work Out For You?" with a poll IMO - long overdue. We could go for pages and pages there

*snip*

1. Argument from Having less Memory

2. Windows warns you about this.

3. Argument from Having less Memory

4. Argument from Having less Memory

5. Argument from Having less Memory

6. Irrelevant.

7. On a 16GB SSD, disabling the PF and Hibenation will free up a lot of memory for application and data.

Here is what I said,

1. It is perfectly safe to disable your page file in many cases. Disabling the page files DOES NOT cause any weird issues. There only obvious issues with disabling the Page File is just that - if go over your physical memory limit the OS will kill the memory leeching app. I had this happen to me once when I played 4-5 Hrs straight Fallout 3 with 2GB RAM. I have been running without a page file for a year or so on multiple setups (1GB being the minimum) without finding a single application that refuses to run without a page file. The argument that things will not run without a page file is for the most part is a lie - it most likely, if at all, applies only to legacy applications that have long been discontinued.

Now I will restate the above in more detail,

If RAM consumption is a non-issue (aka there is enough of it) and dump files are not necessary, I will recommend to everyone to disable the PF. You arguments ONLY apply if the user does not have enough system memory for his applications ( or whatever he/she does ) without the PF. If the user does not have enough RAM for continual and stable day to day operation, then he should not disable the PF. It is also self evident that there will be less system memory without the PF and Windows warns you about the dumps. Therefore any arguments that you throw about having less memory do not apply as they are either self evident OR do not apply to when I recommended to disable the PF.

Now I will restate the above with examples,

"You will have less memory for x" argument - self evident.

"You will have less memory for x as x blah blah blah" argument - self evident, with an unnecessarily long explanation.

"X and Y will crash because you will not have enough memory" argument - does not apply, user should only disable the PF is there is plenty of RAM.

"Stability will be diminished because you will not have enough memory" argument - does not apply, user should only disable the PF is there is plenty of RAM.

"Program X will not work" - applies. User should find alternative to program X, patch/fix for program X, or not disable the PF.

"Dump will not work" - does not apply. User acknowledges this when disabling the PF.

"Less Space With a PF is not a bad thing" - per user argument.

That is pretty much all applications that I do run. Maybe that is where you once again mistake me for someone who uses professional level applications or something like that. If these applications are all lightweight, then how am I going to get any problems you are talking about? I am not. And that is why it is safe for me disable the PF and use the PF room for additional space.

Share this post


Link to post
Share on other sites
StarLion    4
If RAM consumption is a non-issue (aka there is enough of it) and dump files are not necessary, I will recommend to everyone to disable the PF.

Bad advice, please don't tell others to disable the pagefile when you have no idea how it's used. You still haven't listed any real reasons to disable it, instead choosing to continue mindlessly arguing without any facts to back you up.

Put up or shut up. List the credible reasons you have for disabling the pagefile. If you have none, then there is no reason to disable it.

What you need to learn is, there's never enough RAM. If you're wasting gigabytes of RAM on empty reserve space and old pages, that's less data that can be cached. Less data in the cache means you're going to have to go to the hard disk more often to load your commonly used applications. Going to the hard disk is, obviously, slower than loading from RAM.

How can you not understand something so simple?

You arguments ONLY apply if the user does not have enough system memory for his applications ( or whatever he/she does ) without the PF.

Wrong, the facts I've presented still apply even if you have excess RAM. In the case that you miraculously never go over your physical RAM limit, then it comes down what Windows does with whatever free space is left over (which always matters). to Lets try an example this time, maybe that will sink in.

Lets say you have 4GB of RAM, and that with a pagefile, your commonly used apps leave 2GB free. Lets say the same situation, without a pagefile, only leaves 1.2GB free (because of all the junk Windows can't put in the pagefile).

By disabling the pagefile, you've just lost 800mb of space that you could be caching resources. Even if you never use that space for a foreground application, Windows still uses it to improve system responsiveness. By disabling the pagefile, you're reducing the number of applications that can be sped up by superfetch.

If these applications are all lightweight, then how am I going to get any problems you are talking about? I am not. And that is why it is safe for me disable the PF and use the PF room for additional space.

You're ignoring the problems that apply to you (can you really be THAT blind?)... In this case, as you've somehow assured that you will never go over the physical memory limitations of your computer (which i find to be a highly arrogant position), there's one big issue you still can't escape from. You will have less space left over for superfetch (because you've forced unnecessary data and reserved address space into RAM). Just to make it abundantly clear, the less free RAM is available, the less superfetch can use, the less applications get cached, the less responsive your system becomes overall.

Enjoy your sub-optimal configuration, but please, don't tell others to disable the pagefile. As you've proven many times over, you really aren't qualified to give advice on the subject.

Share this post


Link to post
Share on other sites
Udedenkz    51

Lets see, in the above post, you are still using the not enough memory argument. I am not sure how think in the head you are, as I already, said I reject your memory argument - no matter what. Feel free to keep restating it until infinity. restating != saying something new.

You <Snipped-Max™> to think that the PF will be the miracle cure for apps using too much memory, as if one cannot run out of memory for applications to use with a page file. Tsk. Tsk. Tsk. Extension of memory != Magical Unlimited Memory Pool.

Refer To This,

"You will have less memory for x" argument - self evident.

"You will have less memory for x as x blah blah blah" argument - self evident, with an unnecessarily long explanation.

"X and Y will crash because you will not have enough memory" argument - does not apply, user should only disable the PF is there is plenty of RAM.

"Stability will be diminished because you will not have enough memory" argument - does not apply, user should only disable the PF is there is plenty of RAM.

"Program X will not work" - applies. User should find alternative to program X, patch/fix for program X, or not disable the PF.

"Dump will not work" - does not apply. User acknowledges this when disabling the PF.

"Less Space With a PF is not a bad thing" - per user argument.

Edited by Max
Removing insult

Share this post


Link to post
Share on other sites
Max    50

Lets not get personal now. If the personal bickering continues, the posts will be removed and warns dished out. Keep this on topic as well.

Share this post


Link to post
Share on other sites
StarLion    4
Lets see, in the above post, you are still using the not enough memory argument. I am not sure how think in the head you are, as I already, said I reject your memory argument - no matter what. Feel free to keep restating it until infinity. restating != saying something new.

No I'm not, you're still not reading. I made a special concession for you, since you're so convinced that you have nothing to fear from running out of RAM. Without the pagefile, even if you aren't using all available RAM for applications and you have extra left over, you're negatively impacting the responsiveness of your computer by having less RAM available for superfetch than you would otherwise with a reasonably sized pagefile.

As for your rigidness, your loss. I'm not asking you to change, I'm simply asking you not to spread this misinformation around. There are newbies here that might not know to disregard your advice.

You <Snipped-Max?>i> to think that the PF will be the miracle cure for apps using too much memory, as if one cannot run out of memory for applications to use with a page file. Tsk. Tsk. Tsk. Extension of memory != Magical Unlimited Memory Pool.

I never said anything like that. Stop putting words in my mouth, it isn't appreciated.

*sigh* with nothing left to defend yourself with, you've devolved all the way to spouting nonsense while slinging random baseless insults. Go away troll.

---------------- actual contribution to the thread ----------------

Microsoft has added a great deal of functionality to wifi adapters in Windows 7. Unfortunately, a lot of these goodies are hidden, and only accessible via the command line. One such feature is the ability to turn your PC into an access point/hot spot without interrupting any already active connections. This allows an easy way of setting up a small impromptu centralized network (without using ad-hoc mode, which not all devices support).

Now, while it is possible to expose this functionality through the command line, there's an easier way. A little bit of software called Connectify adds a frontend, making these new capabilities easy to configure and manage.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

  • Recently Browsing   0 members

    No registered users viewing this page.