40% more instructions per clock is something i need numbers for.. it makes no sense how it can do 40% more work in a single clock.
3.0Ghz single core no ht can do 3 thousand million operations a second. which is its clock speed. But the base measurement is still 1hz and i don't see how it can do 1.4 operations in a 1hz cycle without turnicating data which would leave that .4 useless any way and would get discarded then it all gets loaded up on the 2nd cycle, hyper threading is one thing, atleast then you are getting 2 opperations per 1hz (if possible) but 40% makes no sense at all.
(obviously we arent just talking 1-2 hz here but 3 thousand million of them per clock. but you are still going to be left with 60% of the workload needed to be processed the next clock cycle to complete the work load of the 1st cycle, so it wouldnt even be added to the 1st cycle at all and would just be done on the 2nd cycle instead)
maybe im missing something?
but the only way it makes any sense at all is if Clock speeds are 40% faster. so its 4.2ghz instead of 3.0ghz (or whatever +40% of 3.0 is) but then that's not 40% more instructions. that's 40% faster clock speed.
