Performance and resource usage

A benchmark is used to measure the speed of the software on available targets. This benchmark is in no way complete, but it gives an approximation of the length of time required for common operations on various targets.

Configuration and performance table

All values are in KBytes/sec.

Description of the performance tests

The performance tests are executed as described and in the order below.
Performance test procedure:

1. Format the drive.
2. Create and open a file for writing.
   W: Start measuring of write performance.
3. Write a multiple of 8 Kbytes.
   W: Stop measuring of write performance.
4. Close the file
5. Reopen the file.
   R: Start measuring of read performance.
6. Read a multiple of 8 Kbytes.
   R: Stop measuring of read performance.
7. Close the file
8. Show the performance results.

The Performance test can be reproduced. Include Performance.c (located in the folder .SampleAPI) into your project. Compile and run the project on your target hardware.

Memory footprint

The file system is designed to fit many kinds of embedded design requirements. Several features can be excluded from build to get a minimal system that can efficiently access any FAT media.

In the following section memory usage for different typical 16 and 32- bit systems is listed.
Please note that the values are valid for the given approximate values. Features can affect the size of others. For example if FAT32 is deactivated, the format function gets smaller because the 32 bit specific part of format is not added into the compilation.

This minimal configuration supports FAT12/16, but has no media format features, no directory support, no free disk space function (DiskInfo) and only one file can be opened at a time:

The configuration file (FS_Conf.h) looks as follows:

16-bit CPUs

System

Memory requirements

32-bit CPUs

System

Memory requirements