Will using FAT32 provide better pagefile performance than NTFS?

Question

I was discussing with my others personalities, and came up with a conflict. In http://technet.microsoft.com/en-us/library/cc938440.aspx , says that FAT32 is faster when using smaller volumes.

Ok separate disk, will give more performance than same disk. But did anyone test this?

• Scenario 1 : Separate hard disk FAT32 (small volume)
• Scenario 2 : Separate hard disk NTFS

which one will win? minimum gain?

Answer 1

It makes some sense that FAT32 might be a bit faster than NTFS, since it doesn't have the overhead of journaling and other advanced features that NTFS offers. See here, here, and here for some references that are equally likely to be verified or hearsay. The first one suggests that the volume and cluster sizes are going to determine which is faster for you.

Putting the pagefile on a separate disk is going to buy you way more performance than the particulars of the filesystem. Most likely, it's not enough to lose any sleep over.

Answer 2

Journaling does have a small impact on performance. But this only occurs when the metadata of the file changes, such during file creation, deletion, resizing, etc. In the case of the pagefile this would normally only occur at bootup. Journaling, and most of the other features of NTFS, do not come into play at all during normal pagefile activities.

In any event pagefile performance only becomes important when it is a limiting factor, which in most cases it isn't. Most attempts at optimizing the pagefile are wasted effort and have virtually no impact on performance.

Answer 3

Actualy i have a ramdisk partition, used for page file,which is around 1.24 gb i used ntfs at first and found that it used up around 40 mb as waste vs fat21 which uses little over 5 mb as waste. And secondly when using virtual machines using vmware there is a performance increase when using fat32 ramdisk as pagefile only partition.