DieHardware
Golden Member
- Jan 1, 2001
- 1,706
- 0
- 76
Originally posted by: valky
Benchmarks using SuperPi - 1M
XP2100+ default - 1733MHz(13x133) = 62s
1754MHz (10.5x166) = 58s
------------------------
1837MHz (11x166) - 56s
1704MHz (8.5x200) = 56s - 400FSB v. 333 FSB = +133MHz
------------------------
1921MHz (11.5x166) - 55s
------------------------
2005MHz (12x166) - 53s
1837Mhz (216x8.5) - 53s - 432FSB v. 333FSB = +168MHz
------------------------
2088MHz (12.5x166) - 51s
2105MHz (10.5x200) - 49s
2170MHz (11x197) - 48s
------------------------
2205MHz (11x200) - 47s
2172MHz (10x217) - 47s - 434FSB v. 400FSB = +33MHz
------------------------
2269MHz (10.5x216) - 46s
This is a limited set of tests based on a single benchmark, but it looks like going from a 333MHz FSB to a 400MHz FSB gives the same increase in performance as a 133MHZ increase in CPU core speed.
Going from a 400MHz FSB to a 432/434MHz FSB yields a rather dissapointing increase in speed equivalent to a 33MHz increase in CPU core speed.
It looks like FSB's higher than 400MHz may not be all that important after all. This would negate the NForce2's principal advantage over the KT600. I cannot run higher FSB's than those used in this test, but I would assume that higher FSB's would produce diminishing returns.
Any thoughts gentlemen?........
:beer:
Yes, stop posting.