| ??? 06/01/07 20:37 Read: times |
#140184 - Normalizing for the subtle differences Responding to: ???'s previous message |
Erik said:
however if you are going for 'fast' it would be ridiculous to start out with a `12 clocker Well, yeah. Of course. What I'm trying to get at here is a good way to evaluate entries in a "fastest subroutine" contest. From what you guys are saying, it sounds like counting processor cycles is the way to go, and further that it is necessary to specify a particular derivative so that the occasional variation among processors doesn't confuse the issue. For the purposes of a contest, then, it seems that one 8051 variant is as good as the next, and we might as well use the one whose instruction times are given in Table 7 of "the bible" -- Russ |
| Topic | Author | Date |
| How fast is fast? | 01/01/70 00:00 | |
| Cycles | 01/01/70 00:00 | |
| One cycle == one cycle??? | 01/01/70 00:00 | |
| there are subtle differences | 01/01/70 00:00 | |
| Normalizing for the subtle differences | 01/01/70 00:00 | |
| Want something fast? | 01/01/70 00:00 | |
can make all at once | 01/01/70 00:00 |