FAT versus NTFS
Home Up

 

WHAT FORMAT SHOULD I USE: FAT OR NTFS?

Exec Software sent us this short update which I thought was useful. I have abbreviated it a little bit for space reasons. We are often asked "Which format is better, FAT or NTFS?" It's really quite simple, once you understand the basic differences between the two. NOTE: In this article, we refer only to FAT-16, not FAT-32, since NTFS does not support the FAT-32 format.

FAT stands for File Allocation Table. This table is located at the start of the partition. It contains information about how and where the files are stored on the partition. In FAT, critical file system data is not duplicated; the loss of a single critical sector can result in the loss of everything on the partition. The File Allocation Tables and the root directory must be at fixed locations so that the files needed to start the system can be located.

The FAT format can be used by several operating systems, including MS-DOS, Windows 3.1x, Windows 95 and Windows NT. It can handle a maximum of 65,535 clusters, and a maximum of 65,518 files.

FAT is best suited for smaller partitions (roughly, to 500MB). On partitions larger than 500MB, the cluster size must be 8KB or more, since FAT will only support 65,535 clusters. Plus, the available cluster sizes are limited; they must be a power of 2. That is, you cannot have a 10KB cluster size; you would have to use 16KB or more. This increases the wasted space due to unused space in the last cluster of a file. Of course, if you only have very large files on the partition, this wasted space is not significant.

NTFS stands for New Technology File System. Central to it is the Master File Table, or MFT, essentially an index to all of the files on an NTFS volume. The critical data is duplicated, so that the loss of a single sector cannot cause you to lose the entire partition.

NTFS includes advanced security features, such as data access control and ownership privileges. Files and folders can have permissions specifying whether the files are shared or not, and permissions can also be assigned to individual files. However, these permissions and attributes will be lost if the file is transferred to a FAT partition.

NTFS can only be used by Windows NT. It is much more fault-tolerant than FAT; NTFS includes recovery capabilities which ensure that the file system can be recovered in the event of a catastrophic system failure. It can also recover data from a bad sector and automatically allocates the bad sector to a special file, so the bad sector will not be used again. The maximum partition size is 2 terabyte.

NTFS does not limit the number of files you can have on a partition; the volume size, and thus the number of file entries you can fit into the MFT, sets the limit. NTFS is more efficient than FAT for larger volumes (roughly, more than 500MB)

Microsoft recommends using a small (250MB to 500MB) FAT partition for the system partition (which contains the hardware-specific files used in loading and initializing the operating system) and for the boot partition, and using NTFS for all of the rest of your partitions. In fact, on RISC-based partitions, the system partition must be a FAT partition. On an x86 system, one of the advantages of this setup is that if you have a startup failure you can boot from a bootable DOS floppy to repair the system.

So here are the guidelines:

  • If you will have more than 65,518 files, you must use NTFS.

  • If your partition is to be accessed by any operating system other than Windows NT, you must use FAT.

  • If you have a RISC-based system, the system partition must be FAT.

  • If data recoverability is important, you should use NTFS.

  • If data security is important, you should use NTFS.

  • If the partition is less than 500MB, FAT is more efficient.

  • If the partition is greater than 500MB, NTFS is more efficient.