Navigate: 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 Search Prices:

LOSTCIRCUITS	SHORTCUTS: A new bus speed Heat and Hertz Test Configurations Cachemem, Sandra Everest and Speed Derating Sysnark2004 Internet ContentCreation Office Productivity WorldBench5 -1 WorldBench5 -2 3D Rendering 3ds max Cinebench2003 Futuremarks Gaming A-Z DOOM3 FarCry Conclusions and Paradoxes Give Us Some Feedback to Help Us Improve our Reviews
Intel Pentium4 3.46 Extreme Edition / 925XE chipset
(Review by MS November 16)

WorldBench5

The discontinuation of Ziff Davis WinStone / Winbench suite earlier this year after the entire enterprise was taken over by PC-Magazine has left a void for system benchmarks to at least offer an alternative to Sysmark. This void has finally been filled by PC-World's benchmark suite dubbed somewhat expectedly "WorldBench". In contrast to the Winstones and Sysmarks, WorldBench5 examines individual applications that appear relevant for home, office and professional users. Some of the benchmarks are very system oriented, though, for example Roxio's Nero Express depends more on I/O transfers than on CPU power.

That is, simply by changing the IDE setting from Standard SATA to an Advanced Host Controller Interface (AHCI) in the BIOS, the run time for the P4 3.46XE dropped by 36% from 758 to 560 seconds. In other words, unless the exact chipset and controller is used, this particular benchmark is not suited to look at CPU performance. We have a similar gripe with the 3ds max benchmarks that are specifically oriented towards professional graphics cards. In other words, even with a hopelessly outdated ATI FireGLX1 as AGP card, it is possible to completely trash any high-end consumer card and since we did not have a matching pair of e.g. PCIe and AGP FireGL or nVidia Quadros, there is no point in showing these results since they are non-representative of CPU or even system performance.

This does not mean that the above mentioned benchmarks have no merit, on the contrary, they are actually great tools for showing what they were meant to show -- albeit not CPU performance.

Adobe Photoshop 7.0.1

Runtime in seconds, lower is better. The only difference, if any, is in the clock speed but that was to be expected.

Intel P4 Northwood 2.4 (hard to find)

Adobe Premiere 6.5

Runtime in seconds, lower is better. Same as above. What is interesting here is that the performance almost exactlyn mirrors the results we showed earlier for Everest's memory benchmark with the Prescott having a substantial lead over the ExtremeEdition despite a slower host bus interface.

Microsoft Office XP

Runtime in seconds, lower is better. Office applications with a lot of multitasking show at least what can be justified as difference. Interestingly, the 875 chipset seems to be able to regain ground based on latencies. One different explanation for the low performance of the original Alderwood chipset here would entail the penalties for synchronizing the memory transfers between the memory bus and the host bus - neither the 875 nor the 925XE are burdened with that task.

next page:     => WorldBench5 Continued =>