ARM Processor
CPU Core | MMU/MPU | ISA |
---|---|---|
StrongArm | MMU | v4 |
ARM7TDMI | none | v4T |
ARM7EJ-S | none | v5TEJ |
ARM720T | MMU | v4T |
ARM920T | MMU | v4T |
ARM922T | MMU | v4T |
ARM926EJ-S | MMU | v5TEJ |
ARM940T | MPU | v4T |
XScale | MPU | v5TE |
ARM946E-S | MPU | v5TE |
ARM966E-S | none | v5TE |
ARM1020E | MMU | v5TE |
ARM1022E | MMU | v5TE |
ARM1026EJ-S | MMU+MPU | v5TE |
ARM1136J-S | MMU | v6 |
ARM1136JF-S | MMU | v6 |
CPU Core | Pipeline Depth | Typical MHz |
---|---|---|
ARM7 | 3 stage | 80 |
StrongArm | 5 stage | 133 |
ARM9 | 5 stage | 150 |
ARM10 | 6 stage | 260 |
XScale | 8 stage | 400 |
ARM11 | 8 stage | 335 |
NOTE: increased pipeline length reduces the amount of work done at each stage in the pipeline, therefor enabling higher operating frequencies and performance. however, as the pipeline length increases, system latency also increases due to increased number of clock cycles needed to fill the pipeline before an instruction can be executed. an example would be an ARM920T running at 400MHz might have comparable performance to an Xscale running at 600MHz