|
Untitled Document
Benchmarks continued..
3dmark 2001SE
|
1024x768 |
1280x1024 |
Canterwood P4 @3.0ghz |
10941 |
8033 |
Canterwood P4 @2.4ghz |
10470 |
7920 |
Nforce2 XP @1743mhz |
8695 |
6981 |
The numbers speak for themselves, the Canterwood scores 2000 or so points above
the Nforce2, regardless of the CPU clock rate.
Memory Performance (Sciencemark)
| Canterwood |
4035MB/s (Bios Default timings) |
| Nforce2 |
2295MB/s (Bios Default "Aggressive" Timings) |
Here we see just what a difference the massive lead the Canterwood's FSB makes.
Though it doesn't reach its theoretical peak of 6.4GB/s, it is still nearly
twice as effective as an Nforce2 running 333FSB and Dual Channel DDR400!
SPECviewperf 7.1 (OpenGL)
|
Canterwood @ 3ghz |
Canterwood @ 2.4ghz |
3dsmax-01 |
8.655 |
8.454 |
drv-08 W |
32.77 |
32.93 |
dx-07 |
52.37 |
46.10 |
light-05 |
13.88 |
12.04 |
proe-01 |
11.74 |
11.47 |
ugs-01 |
6.488 |
5.972 |
The Nforce 2 results are omitted as they were totaly inconsistent
with typical Nforce2 results, and I expect to get to the bottom of this during
writing the upcoming Nforce2 Shootout article and include the correct results
for this board.
Conclusion
The Canterwood chipset is simply going to be the best available for all out
performance in the near future, though as usual with Intel you pay for the privilege.
Will somebody who already owns a fairly current system wish to upgrade to Canterwood?
I doubt it. But if you insist on the last word in performance no matter the
cost, you are going to be very pleased.
Many thanks to Intel and Asus for supplying their respective components.
Matt Skinner (Random Nonsense)
<< Back | 1 | | 2 | | 3 | | 4 | | 5 | Next >>
|
