| ??? 06/19/10 12:19 Read: times |
#176766 - There are some operations ... Responding to: ???'s previous message |
Andy Neil said:
Richard Erlacher said:
I'm not certain whether 3 32-bit ARM's could outperform 3 805x's. The answer hinges on how quickly the 32-bitter can isolate a specific bit or byte and act on its state. No, that's not true! What you should say is: IF the answer - in the specific application at hand - hinges on how quickly the processor can isolate a specific bit or byte and act on its state THEN it's not certain whether ARM's could outperform 8051/2s. Well, that might be the case if you're using a compiler, but if you're in ASM, which you probably would be if you're in a time-critical situation, it's really not application dependent. Unless you're supporting some sort of vectored interrupt, you have to isolate the byte, and then do something on the order of JMP @A,DPTR, which is pretty quick. I do believe it's not uncommon for 32-bitters to require an extra step or two in order to get down to a small table with offset. AFAIK, that's the only performance-impact of the architectural difference. How easy is it to get 3 ARM's to play together? How easy is it to get 3 805x types to work together? That is, indeed, a very good question! There really are things an 8-bitter can to more quickly or more easily than a 32-bitter. You don't trip over them too often, but they're out there. RE |
