Freja currently supports analysis for a number PowerPC and ARM processors. For these processors Freja has an analysis model complete with all the essential information such as cache hierarchy layout, cache sizes and cache line sizes, the availability of prefetch instructions and their effect at different cache levels and whether hardware prefetching is available different at cache levels.

The supported PowerPC processors are:

The supported ARM processors are:

The ARM Cortex processors can be configured with many different combinations of L1 and L2 cache sizes, and with or without L2 cache. The Freja default models for these processors include a 16 kilobyte L1 data cache and a 128 kilobyte L2 cache.

If this does not match the configuration of the processor you are using, select the closest processor model and the desired cache level, and then specify a different cache size to override the default.

If your processor does not have an L2 cache cache, simply do not perform an L2 analysis. The L1 cache analysis done by Freja is still valid.