Difference: 31108.66 – 30112.33 = 996.33ms (or 0.99633 seconds) C'mon. Difference: 16.11 – 16.05 = .06 Stop it now lol.
The difference is 3% faster with PF disabled, and regarding Crysis "A closer look reveals significantly higher minimum fps without a pagefile ( 9.30 to 6.54)" I really don't care if it's faster or not, that's isn't why mine is disabled. I was illustrating that there are actually benchmarks that show a performance comparison between enabled/disabled PF, contrary to IPlayNaked's comment.
Oh, here we go again...is this a repeat performance or some sort? Just keep the debate on its own thread; you guys probably freaking the OP out. @Anfield82 If you're using SSD, it's advisable to either use another HDD (if present) to put the page file in, or disable it completely. Another option is to put the page file on a RAM disk, which would mean you don't have to put the page file on your SSD but programs and Windows will still have a fast page file. If you want to know a way to do that, check my thread.
I have had the Page file disabled for about 6 months mainly because I'm only running an SSD, no standard hard drive apart from my old 640GB WD as an external USB backup. I didn't do it for performance, more to save the SSD the read / write cycles. Haven't had any issues running anything on my system, BF3 and COH included. Like someone else said, try it and see
I hate being flamed when I post regarding storage. I was a benchmarker / system tests engineer at : Netapp storage IBM storage / XIV Kaminario (SSD based storage) After 5 years of constant benchmarks, I can safely say with 100% certainty that disabling SSD's have NO noticeable effect. Modern machines with 8GB or more function just as well with as without. The reasons NOT to disable it are many. So , my suggestion, move paging file to a physical hard drive. You wont wear out your SSD (not likely anyways) and performance wont suffer either. NoamM Please , if you feel the need to flame me, do it in pm.
Out of curiosity sake I ran the Crysis bench tool with PF and without. As expected, nothing conclusive. The min frame has wide variances if bench is run 3 times even under same test conditions (with PF or without). However the averages show much more consistency. Minimums can be misleading because if it was only a micro-dip affecting one frame (which you will not likely notice in play), it would still show up as a minimum in the results. Thats why I tend to look at average FPS, esp if it is far off from the minimums. Re micro-dips, a background service, AV or whatever suddenly acting up can probably cause it, I dont know. All I know is you cannot run this bench (and most fps benches) and get identical results to the last frame each and every time. Same applies to all the other benches in that article, they are all so close that the results are statistically insignificant. I never even get the same exact boot time every time I boot, it can vary by a few secs at times. Therefore, as there has been nothing compelling to push me to no PF, I stick with 2 small PFs (1gb x 2), on different drives. With PF: (3 runs) TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s) !TimeDemo Run 0 Finished. Play Time: 47.09s, Average FPS: 42.47 Min FPS: 21.53 at frame 143, Max FPS: 52.22 at frame 977 Average Tri/Sec: -12808418, Tri/Frame: -301586 Recorded/Played Tris ratio: -3.04 TimeDemo Play Ended, (1 Runs Performed) TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s) !TimeDemo Run 0 Finished. Play Time: 47.76s, Average FPS: 41.88 Min FPS: 21.81 at frame 139, Max FPS: 52.51 at frame 1016 Average Tri/Sec: -12627139, Tri/Frame: -301530 Recorded/Played Tris ratio: -3.04 TimeDemo Play Ended, (1 Runs Performed) TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s) !TimeDemo Run 0 Finished. Play Time: 47.16s, Average FPS: 42.41 Min FPS: 24.45 at frame 138, Max FPS: 52.92 at frame 1018 Average Tri/Sec: -12793844, Tri/Frame: -301695 Recorded/Played Tris ratio: -3.04 TimeDemo Play Ended, (1 Runs Performed) No PF: TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s) !TimeDemo Run 0 Finished. Play Time: 47.56s, Average FPS: 42.05 Min FPS: 17.48 at frame 145, Max FPS: 52.11 at frame 1786 Average Tri/Sec: -12694409, Tri/Frame: -301854 Recorded/Played Tris ratio: -3.04 TimeDemo Play Ended, (1 Runs Performed) TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s) !TimeDemo Run 0 Finished. Play Time: 47.02s, Average FPS: 42.54 Min FPS: 19.43 at frame 145, Max FPS: 52.07 at frame 973 Average Tri/Sec: -12847132, Tri/Frame: -302021 Recorded/Played Tris ratio: -3.04 TimeDemo Play Ended, (1 Runs Performed) TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s) !TimeDemo Run 0 Finished. Play Time: 47.22s, Average FPS: 42.35 Min FPS: 23.49 at frame 138, Max FPS: 52.47 at frame 1017 Average Tri/Sec: -12789660, Tri/Frame: -301973 Recorded/Played Tris ratio: -3.04
It's 100% safe to leave a pagefile on an SSD so long as the SSD has sufficient space available. There isn't enough writes to the SSD as a result of the pagefile being stored there to have any real impact on it's life span. Pagefile activity is mostly reads, which have no effect on SSD lifespan. That's what I did....
I've just re enabled mine but to 4GB, wasn't sure how much I needed, experienced two random losses of power in the last week and wanted to consult the crash dump file only to find nothing as the PF was off, so now I gots to wait for another random loss of power to see if I can find something out about it. Never sure why some people fuss so much about it tbh, since it seems thats SSDs aren't affected much by it presence think I'll just leave it on in future, has seemingly no affect on system performance off or on really.
Yes its safe to turn it off, turned mine off after installing OS on a intel 520 SSD and had zero issues
Last time i tested, when pagefile is off, whole PC feels faster. But, i had problems when working with archives, maybe because i use HDD or just 4GB of RAM (usage never goes above 1.5 - 2GB, so i doubt it's that reason), anyway it as way slower when compressing/decompressing archives. I doubt it will be a problem on SSD, if it will make SSD last longer, it is safe to turn it off.
Cheers for the Link. However it seems even with the PF enabled nothing is being written out. Just had another random power loss and nowt was created, was hoping to try find out what was going on.