After careful consideration I have decided to transfer all hardware review activities to a new domain. I purchased Hardwareasylum.com in 2012 and have been working hard to build a new and improved Ninjalane on that domain. If you are reading this you have reached one of the archived articles, news, projects and/or reviews that were left behind during the site migration.
Please update your bookmarks and be sure to visit the new and improved Ninjalane at Hardwareasylum.com
Corsair Twin2X DDR2 5400C4 ProSeries Review
Author: Dennis Garcia
Published: Thursday, October 21, 2004
Benchmarks
The system as it was tested
DFI LanParty UT 915P
Pentium 4 530 (3.0Ghz) 1MB L2 cache 800MHz FSB
OEM Heatsink
2x Corsair Twin2X 5400C4PRO DDR675 (4-4-4-12) 512megs
Chaintech PCX5900
IBM Hitachi Deskstar 80gig SerialATA 7200RPM
Antec NeoPower 480 power supply
12x/40x DVD/CDRom
Windows XP Professional Service Pack 1
Pentium 4 530 (3.0Ghz) 1MB L2 cache 800MHz FSB
OEM Heatsink
2x Corsair Twin2X 5400C4PRO DDR675 (4-4-4-12) 512megs
Chaintech PCX5900
IBM Hitachi Deskstar 80gig SerialATA 7200RPM
Antec NeoPower 480 power supply
12x/40x DVD/CDRom
Windows XP Professional Service Pack 1
SiSoft Sandra
Sandra is a software collection of synthetic benchmarks
that will give us a basic idea as to what a system is capable of. It should be noted that SiSoft numbers change depending on what version you have installed, these were
recorded using Sandra Professional Version 2004.8.9.131
Cachemem
Cachemem is a synthetic benchmark used to test bandwidth and latency.
FarCry
For this benchmark the training level demo was used and the built in benchmarking utility recorded the in-game framerate.
If you would like to try this benchmark out for yourself add the following arguments to your FarCry shortcut.
-DEVMODE "#demo_num_runs=1" "#demo_quit=1" "#r_Width=800" "#r_Height=600" "map training" "demo training"
#demo_num_runs= - how many times you want the benchmark to run, this is 0 based
#demo_quit= - 1 will quit the game after the run, 0 will return to the main menu
#r_Width and Height= - I think you can figure this one out
Map mapname - tells to load a map called "mapname"
Demo demoname - run the demo called "demoname"
-DEVMODE "#demo_num_runs=1" "#demo_quit=1" "#r_Width=800" "#r_Height=600" "map training" "demo training"
#demo_num_runs= - how many times you want the benchmark to run, this is 0 based
#demo_quit= - 1 will quit the game after the run, 0 will return to the main menu
#r_Width and Height= - I think you can figure this one out
Map mapname - tells to load a map called "mapname"
Demo demoname - run the demo called "demoname"
Unreal Tournament 2004
For this benchmark we used a multi-player demo called Primeval.
If you wish to try the demo out for yourself you will need to download the demo files from the following website or
ask me and I'll send you a copy. The timedemo is similar to the one used in Quake III in that once the demo is over you pull down the console to see the average FPS.
Benchmarking Conclusion
The numbers tell a very strange story; from a synthetic standpoint lower memory timings and higher clock speeds increase performance. This much is relatively easy to understand though in the real world benchmarks the major gains follow clock speed with only minor gains due to memory timings.