Not sure what this means. Windows will not let you actually "defrag" a SSD. That's why the defrag program in Windows is now called "Optimize Drives". When the system detects an attached drive, it determines if it is a hard drive or SSD and performs the applicable tools for that type drive. For example, TRIM for SSDs and defrag for hard drives. You say "not swap file". What does that mean? As I said above, you should let Windows manage your Page (swap) files. That includes having the page file on your SSD. SSDs are ideally suited for Page Files. See Support and Q&A for Solid-State Drives and scroll down to, "Frequently Asked Questions, Should the pagefile be placed on SSDs?" While the article is getting old, it applies even more so today since wear problems of early generation SSDs are no longer a problem and each new generation of SSD just keeps getting better and better. And there's no harm in using hibernation with SSDs either - with notebooks. With PCs, "Hybrid" mode should be used. Again, it was only first generation SSDs from many years ago that suffered from limited write issues. Newer generation SSDs still have limits, but the numbers are so high, you will never see it. TRIM and wear leveling featured see to that.
Many article I read on the Web when I got a new pc, with a SSD, always said this: not swap file, not defragmentation, not hibernation..... Anyway, at least I used HD, I never saw advantages in performance using the swap file.
Got link to any article that says that? A swap file is the same thing as a page file. And there's no reason to disable it.
Swap is not about performance. It is about increasing available system memory. Example: If you have 2GB RAM and your OS + programs need 2,5 GB at the moment, least used 0,5GB memory can be dropped on HDD/SSD. It is going to be slower than having 4GB RAM. It may let you run your programs which will not be possible without swap file having 2GB RAM. It is going to be slower because HDDs, SSDs are slower than RAM. It is best for performance to have more RAM and disable swap file. The only disadvantage is that without swap you can't have hibernation (suspend to HDD/SSD), but you still can sleep (suspend to RAM). Given that dropping memory to SSD (i.e. while hibernating) means writing a few GB amount of data to SSD, it also should increase your SSD life expectancy, but in depends how often you hibernate.
If I were to use an SSD I would use it on Windows 7 (or possibly Linux at some point in the future). Windows 7 has 'defrag' ...
What? Sorry but that makes no sense at all. What does more available system memory do? Improve performance. Totally wrong! It amazes me when people believe they have more expertise than all the PhDs, computer scientists, and programmers at Microsoft, Linux, Apple and every other OS maker. Microsoft has exabytes of empirical and scientific data gathered from 100s of millions (billions even) of computers over decades of real world use. Microsoft has no interest in degrading your systems performance. Why? Because if they do, all the Microsoft haters will bash them relentlessly. Windows employs a "dynamic" Page File. That means it will expand and contract as needed to optimize performance. Don't you think if you had gobs of RAM they would set the PF to 0 if that improved performance? Your comment about SSDs and PF and hibernation is just wrong too. SSDs today are rated to transfer 20GB or more of data every day for 5 years or more.TRIM and wear leveling ensure that usage will be evenly distributed. This ensures you are not going to wear it out prematurely. Hibernation was was developed for notebooks! More and more notebooks are coming with SSDs only. If reasonablePrivacy's comment made any sense, these notebook SSDs would be failing right and left. But that is not happening. There is no reason or advantage to remove the Page File. But don't believe me! Read the following: See also The How-To Geek: Understanding the Windows Pagefile and Why You Shouldn't Disable It. The Out-of-Memory Syndrome, or: Why Do I Still Need a Pagefile? As I said above, Windows (including Windows 7) knows if the drive connected is a SSD or not. While the program in W7 may be called Disk Defragmenter, it will NOT defrag the SSD. To eliminate that confusion, they renamed that tool to Optimize Drives in newer versions of Windows.
Or even 600 GB of data every day for 5 years. We shouldn't be concerned about writing too much data to our SSDs.
Users have diverse needs and use cases so Microsoft can't predict what setting is best for everybody. Something that is optimal for somebody is not optimal for somebody else. I remember times when I had computer with 128 MB RAM. Back then swap was necessary. My OS used heavily swap. Now it is not. I have disabled swap on my all systems, because I have enough RAM and guess what? They work like a charm. So if somebody has 16 GB of RAM and let's say on average uses about 70% of RAM and hibernates 4 times a day everyday it means 44,8 GB written daily to the SSD just for the process of hibernation. Add GB per day from rest of system usage.
I did the same and been working for 4 years like this. Confirmed, they do work like a charm! Especially at boot time. Pretty fast! Yes I know, some applications/scenarios might still need swapping. Not mine. Edit: Tbh and accurate, I actually left a 350 MB swap file just for Windows to be happy.
They are not predicting. Windows analyzes in real time and adjusts as necessary. So Samsung, ASUS, MSI, ACER and all the other notebook makers selling notebooks with SSDs only are all involved in a giant global conspiracy to rip off consumers by using storage devices that will fail prematurely to force users into costly repairs and buying new SSDs? Right? Because that is what you are suggesting. Yeah right. Using the justification that disabling the PF didn't break anything, therefore it must be better to disable it is twisted logic. What is broken by sticking with the defaults? Nothing. What is improved? Read the links I provided. Use Google. Learn. Don't assume you are a master at virtual memory management. I provided several links to reputable sources showing why we should leave them enabled. You have provided nothing. Of course that is not true either. While Windows does not require a page file to run, Windows itself needs one to run optimally. Just because you have not noticed any degradation just suggests your hardware is fast enough - you just choose to not let it run to full potentials. Again, since Microsoft designed the PF to adjust dynamically, why wouldn't they set it to 0MB if that was better when large amounts of RAM was installed? It would be simple for them to do. And it would prevent bad criticisms if that was better. Why don't the other operating systems recommend disabling the PF if you have lots of RAM installed? Simple there too. Because those operating systems also use the PF to optimize virtual memory.
Added: https://www.redhat.com/en/blog/do-we-really-need-swap-modern-systems End of added. I didn't said that. I am not interested in discussion where somebody throws false accusations against me. I am out.
Tbh and accurate, I actually left a 350 MB swap file just for Windows to be happy. Wonder what those full potentials are...
You implied that by suggesting hibernation will increase aging (writes) so significantly on SSDs that it should be disabled. That is wrong on today's SSDs. Yes, I was being sarcastic - but to illustrate my point. And that is that your claim is just wrong. If it were true, notebooks and many new PCs would not come with SSDs as their boot drives. But many do - especially notebooks which use hibernation. And it is interesting how you left out the significant part of that Red Hat article. Did you actually think I would not read it? So clearly, the do NOT recommend disabling it! In fact, no where do they recommend running without one. Yes, they acknowledge there are times where running without one is supported - but as you quoted, "the customer should know the behaviour in the above situations". Are you seriously claiming to be an expert in virtual memory management? And not just for the OS but your running programs? And if you are making such a claim, since virtual memory requirements vary depending the tasks being performed and programs running, are you constantly adjusting the PF as required? Because disabling the PF, or manually setting the PF size is NOT a "set and forget" setting - unless the tasks perform never change (like on a POS computer). And of course, Red Hat is a version of Linux - which does NOT use a dynamic PF. Again, read the links I provided. The 3rd one in my post #31 is very informative. Now I realize this is your thread - but it is about SSDs and Data recovery. Not Page Files.
Me too ! @Bill_Bright: If someone work fine disabling swap file what is the problem ? I always worked so, no problem and fully performing system.
Even if you haven't had any issues after disabling the page file, there is no need to do it. SSDs last so long, that the extra writes from paging, simply do not matter. There'a good article on it at How To Geek. https://www.howtogeek.com/126430/ht...-windows-page-file-and-should-you-disable-it/ The article also explains that there is no performance benefit from disabling the page file.
Because most people who disable it are not experts on virtual memory - yet they pretend to know better than the true experts who develop modern operating systems, or they blindly follow the suggestions of people who don't know what they are doing. Most who disable it don't know what they are doing, or why - and the excuses given are typically loaded with fallacies, misconceptions and total falsehoods. W7, W8 and especially W10 are NOT XP yet many feel they need to treat these modern operating systems like they are. Not to pick on or single out blackknight but just as an illustration, the Page File is NOT about increasing "system memory". Yet reasonablePrivacy said that and you believed it. The page file is about increasing "virtual memory", not system memory. BIG DIFFERENCE! If you don't fully understand the difference between virtual memory and system memory, leave the PF settings alone! Let Windows manage it! It does know the difference. If you don't fully understand how virtual memory affects performance and stability, and if you don't understand OS and program "private commit rates", "pool usage", and "mapping memory", leave the page file settings at their defaults! Let Windows manage it! Other fallacies and misconceptions mentioned and used as an excuse to disable the PF is the claim that "it is best for performance to have more RAM and disable swap file." That's a total falsehood. Operating systems can run out of "virtual memory" and still have gobs of "system memory" left unused! BEWARE! When people say, "My programs don't need a page file" or, "I have lots of RAM, I don't need a PF" or, "Disabling the PF forces the OS to use faster system memory for better performance" or, "I disabled it and didn't see any difference" or, "I disable it and my computer still works fine" or, "I have a SSD so I don't need a PF" or, "I don't want to wear out my SSD" or, "I set my PF to 1.5 times my RAM" (the old obsolete XP rule of thumb).Watch out! Those are clear red-flag warnings they don't know what they are doing and/or are blindly following the advice of someone who does not know what they are doing! Even the world renown, preeminent expert on virtual memory management, Mark Russinovich, who wrote the fantastic white paper, Pushing the Limits of Windows, NEVER recommends disabling the PF. He goes into great detail on how to manually determine the correct page file size for those who want to, but disabling it completely is never suggested as an option. And note for anyone who wants to study Mark's paper and set your own fixed page file size, just remember it is NOT a "set and forget" setting. Any time you make a major change to the OS, apply major updates to major programs (like real-time security apps), make significant hardware changes, or you change your computing tasks/habits, you need to recalculate your requirements. Or better yet, just let Windows do it for you.
The same might apply to those who claims quite the opposite, right? We are so ignorant so, to whom we shall believe? Btw you are not upset, are you? I mean you're writing style...
This: and this: I must admit it. I have faced the first statement twice thrice last year when running a vm and several Chrome profiles opened. Annoying issue it is. Second one, well I haven't faced it and don't want ever in the future. Thus I ended reverting every registry and any other tweak about disabling page file. So I for one I'm done with this off-topic about page file.
Sorry, I believe that reasonablePrivacy mean that Page File, increasing virtual memory, increases the system availability. Anyway what I said I thought for many years before reading this thread. Why should I ? Actually: - I have lots of RAM and I don't need a PF that would slow my system. - I disable it and my computer still works fine - I have a SSD and using the PF would uselessly increase writing operations. - I didn't experimented advantages to use PF. If my system works fine and faster without the Page File, why I can't do as I want, and I should let do to Windows ? Windows already control too many things, and MS trend, especially with 10, is to limit the user as much as possible.
First, no. I am not upset. Sorry if I come off that way. But I am passionate about ensuring readers understand what they are doing, and any potential consequences for doing it BEFORE they dink with settings. That means doing their homework and seek out the experts. Don't assume some forum poster (including me) knows what they are talking about. Don't believe me just because I said it - I posted several links to real experts. You mean Microsoft, Ed Bott, The How-to Geek, that Red Hat author, and all the other recognized experts in the world who recommend we just let Windows manage it? Ummm, no. It is not the same. How about the real experts - like those you quoted in your last post above? The same article reasonablePrivacy cited? Ummm, no. First, that does not even make sense. What do you mean by "system availability"? If you mean "system RAM", that is wrong as system RAM is the physical RAM installed on the motherboard. If you meant "total memory" available to the system, that is "virtual memory" since virtual memory equals system memory (RAM) plus the PF memory. And if you did mean the PF increases the amount of "total memory" available to the system, then that is a good thing but then would totally contradict everything else reasonablePrivacy claimed. So either way, he is incorrect. Thus proving my point. A PF does not slow a system down when you have lots of RAM. It may cause a very small delay when maximizing a minimized window, but after that, performance is full speed - and with a SSD, that delay would not be noticeable. The SSD is ideally suited for the PF and write operations are no longer an issue. Ever heard of Superfetch and how it improves performance? Without a PF, Superfetch cannot do its job. And you admit you didn't do any testing. Ah, now the truth comes out. I wanted to state earlier that many choose to disable the PF just because they are biased and don't want Microsoft telling them what to do. That is sad and foolish. I am the first to admit that some of Microsoft's "marketing" and executive decisions have been outlandish, intrusive, and often offensive. But that is on the marketing weenies and executives. The developers, on the other hand, are top notch - clearly some of the best in the world and totally committed to ensuring Windows, and especially Windows 10 is the best Windows operating system ever. Best in terms of security and performance. And they have done that and continue to improve it. So to change technical settings that affect system optimization just because of some bias against Microsoft, well that's just silly and naive. My advice to you blacknight is to switch to Linux. Just note they recommend you use a page/swap file too. And since you are the OP, I respect that decision and I will step out of this mess too.
Why ? Because I don't want to use Windows as Microsoft want ? May be that in the future I'll chose Linux, it's customizable, and - not last - it has very nice GUI that can be changed, modified, customized... while Microsoft 's OS is, by a graphical view, always the same since 98. But now I'm quite good fine with 7 without Page File. Naturally, we are here only for debate.
It is unusual for program to request 6 GB in on system call. Not to mention that after failed request programmer can handle that failed request, because i.e. C language function is going to return value indicating error. I would say usually this indicates error in memory management of that program. In most cases program is going to crash soon anyway. I admit that additional memory via PF can in some cases rescue badly behaving program from crash, but only in some rare circumstances.
