Quote:
|
Originally Posted by ChaosBlizzard
If it's using 100% of its processing power the cache is full. You tell me otherwise and you need to go back to the books.
Runway thread and pre-emptive don't belong in the same sentence with each other. When was the last time you seen a "run-away" thread in a pre-emptive system? Besides, today’s systems use 2, 3 or sometimes even more ways to schedule a thread and run a process.
You do realize Windows 2000 is both a co-operative system, pre-emptive system, and a priority scheduling system?
|
I should hit the books? You're the one making outrageous claims.
Are you saying that if I write a simple two-instruction program, a NOOP and a jump to that noop (Which would take the CPU to a hundred percent), that it would use up the entire L1 and L2 cache, both instruction and data? That makes absolutely no sense.
I guess you're going to also tell me where everywhere it's been mentioned that having more cache on a processor helps it keep it's pipeline full, despite the fact that it's at 100% usage either way, that those people were either lying or incorrect? Perhaps you should go tell Anand to "hit the books".