Shouldnt this apply?

[daos]

does the internal clock multiplier affect overclocking the front side bus? could you overclock the speed of the bus higher, lets say, on a Pentium4 2.53ghz than a Pentium4 2.8Ghz? in theory, you should not be able to, since the stepping is the same as well as the core and main architecture. you should not be able to since the processors rating is at 2.8 and should be able to reach as high as the 2.53.
lets say we overclock the bus of the 2.53 from 133 x 19 to 166 x 19. your effective clock speed would be 3154. if you did the exact same thing to the 2.8 which would be 166 x 21 you get 3486. both have a speed of 667 for the bus but each have a different speed for the processor because of the multiplier. if you could reach 3154 on the 2.53 shouldnt you be able to reach 3486 as well for the 2.8?

 

[Duvie]

No the logic is flawed...

It is not the front side bus (ie your 667 example) that is similar per stepping it is the ovrall mhz....For example many may packing date early b0 stepping chips did 2.53 to 2.7ghz with anywhere from default to 1.7v...Later b0 stepping chips from september on have been doing 2.7-2.88ghz more regurlary from default to 1.7v....

To correct you example....


19x166 = 3154 with a 2.53b and the 2.8b should get 21x150.....

The reason ppl around here go for the lower chips like the 1.6a, 1.8a, 2.26b, and 2.4b is more for the low multiplier as the cost. The lower the multipler the higher the fsb can likley go thus increasing overall system bus and potential of jacking up mem speeds to 400mhz and beyond. In a given stepping and packing date the overall mhz is likely very similar across the board of manufactured chips at the time.

When Intel is producing chips they don't produce individual speed chips. They are intended to be the same speed chip. Some just get binned down to meet market demand or don't cut the mustard at the given speed. However Intel was getting good yields so most chips were severly downclocked to fill those speed demands inthe market.

I got a 1.6a back when Intel was pumping out the 2.53ghz chips...I got lucky and got a chip that does 2.53ghz at default vcore all day stabily and prime tested numerous times for 24hours. Now the chance at that time was the 2.53ghz may not do mch better then 2.8ghz and thus would never get much better then 2.85ghz for a total system bus of 600 (150x4)...but my little chip could do 2.74ghz or 684fsb (171x4)....You could not get a 171fsb out of those 2.53b chips without some serious water cooling and dangerous vcore voltage.

 

[daos]

i see your point. anyone else feel up to responding?

 

[DAPUNISHER]

Originally posted by: daos
i see your point. anyone else feel up to responding?

I think Duvie's answer was rather definitive and requires no further elaboration.