|
Advice Beginners BIOS Guide CPUs Links Mainboards Memory Network Storage Video/Sound Cards Contact Forum SiteMap Sponsors WebNews Home |
. | . |
Prices: Mainboards ABIT ASUS Chaintech Shuttle Soyo Tyan CPU Intel P4 2.4C-800 P4 2.6C-800 P4 2.8C-800 P4 3.0-800 P4 3.2-800 AMD AthlonXP XP 1700+ XP 2000+ XP 2400+ XP 2500+ XP 2700+ XP 3000+ XP 3200+ Athlon64 Athlon64 3200+ Athlon64 FX-51 Opteron Opteron 240 Opteron 242 Opteron 244 Opteron 246 Memory Corsair Crucial Kingston Mushkin OCZ |
LOSTCIRCUITS |
SHORTCUTS:
The Multiplier Legacy Validating the New, Supporting the Old Apples to Apples and Clocks to Clocks System Disclosure, SiSoft Sandra Memory PCMark2002, A Different Perspective Benchmarks that won't: CodeCreatures, Comanche4 SPEC ViewPERF and diminishing returns Fraggin the Quake3 Arena Scores Expendable and the Chipset Frequency, 3DMark2001SE 3DMark2001SE: Behind the Looking Glass Profiling Performance, Will Barton Challenge Hammer? AMD Processor Steals |
| 333 MHz FSB for the Athlon The Unkept Promises | |
| (Review by MS, September 2, 2002) |
Expendable
Another outdated benchmark in that hardly anyone still plays the game, nonetheless a very good indicator of what happens when a script is executed, meaning that this is not a recorded demo in the conventional sense but a sequence of logically connected events where one frame depends on the previous.
Expendable 640 x 480 x 16
There is hardly any prefetching involved in Expendable which is why we would have expected a higher relative impact of increasing the memory bus frequency per se. An almost even distribution in gain of 4.2 and 5.2 % respectively reflects among other things the simple effect of the chipset ticking faster, that is, internal latencies including turnaround times are lower. We know that those chipset latencies are critical for performance in Expendable and therefore we interpret the additional increase at 166 MHz FSB mostly as a result of the shortening of these latencies (in absolute ns measurements).
3DMark2001SE
Despite the fact that 3DMark2001 SE scores reflect no less than a total of 17 benchmarks (withough the image quality tests) with different weightings as well as the fact that several of the benchmarks are mostly limited by the graphics card as the prevalent bottleneck, the overall score shows quite a performance delta between our three settings. Interestingly, moving to the higher FSB yields a greater gain than increasing memory bus frequency alone.
As briefly touched upon, 3DMark2001 is so complex and graphics card limitations are playing such a dominant role that it is surprising to even see the amount of delta that we find. When we are analysing the results in more detail, we find that the majority of sub-scores is not even affected beyond gains that are just above noise level, however, there are several applications that show substantial differences. This is getting more interesting ....
next page: => "Car Chase" and "Lobby" =>
