System setup and notes
Here's a quick rundown of the test system should you wish to compare
benchmark results with your own.
AMD Athlon 64 FX-53 S939 CPU - 2.4GHz - 1.5v
AMD Athlon 64 Model 3800+ S939 CPU - 2.4GHz - 1.5v
AMD Athlon 64 FX-53 S940 CPU - 2.4GHz -
1.5v
AMD Athlon 64 Model 3400+ S754 CPU - 2.2GHz - 1.5v
Intel Pentium 4 3.4GHz Prescott S478 CPU - 1.4v
Intel Pentium 4 3.4GHz Northwood S478 CPU -
1.6v
Intel Pentium 4 3.4GHz Northwood Extreme
Edition S478 CPU - 1.6v
Other components
ASUS P4C800-E i875P (Rev 1.15 BIOS)
EPoX 8HDA3+ VIA K8T880 S754 motherboard (02/12/03 BIOS)
MSI K8T Gem (MSI 6702E) VIA K8T800 Pro S939 motherboard
ASUS SK8N nForce3 Pro 150 S940 motherboard (Rev 1005 Beta 006)
2 x 256MB Corsair XMS3500C2, run at 2-2-2-6 for all S478 CPUs and S754
2 x 512MB Legacy Electronics DDR400 Registered / ECC memory used for FX-53 (S940), run at 2.5-3-3-5
2 x 512MB Corsair XMS3200LLPT DDR400 for S939 CPUs, run at 2-3-2-6
ATI Radeon 9800 XT 256MB (412/730)
Western Digital 160GB PATA hard drive
Toshiba 8x DVD
Samcheer 420w PSU
Dell P991 19" monitor
Zalman 7000Cu S478 / 754 / S939 cooler
Software
Windows XP Professional SP1 with latest updates
DirectX9.0b
VIA Hyperion 4.51 drivers
Intel 5.02.1002 chipset drivers
ATI CATALYST 4.1 drivers and control panel
Pifast v41 to 10m places
Lame v3.92 MP3 encoding with Razor-Lame 1.15 front-end using U2's Pop album
(607MB)
Gordian Knot - XviD encoding test using first vob of Sleepy Hollow. 1433kbit/s
bitrate
Kribi Bench v1.1
ScienceMark 2.0
Realstorm Raytracing benchmark 320x180x32
3DMark 2001SE v330
UT2003 Retail (Build 2225)
X²: The Threat - Rolling Demo
Comanche 4 benchmark
Quake 3 v1.30 HQ
Notes
AMD was kind enough to supply us with a 1GB TwinX pack of Corsair's excellent low-latency memory and MSI's upcoming S939 VIA K8T800 Pro motherboard. As ever, results obtained from a 'new' chipset and new processors aren't totally indicative of future performance. We expect performance BIOSes to be released and benchmarks to rise accordingly.
There were no problems to report during installation or testing. Our main concern is with the S939 processors' increased heat output. They push out a Prescott-like 103w TDP.
Not what we're used to seeing from AMD's Athlon 64 line of CPUs. It's difficult to ascertain why S939 processors, using the same packaging as S940, run so much warmer.
Overclocking
Our overclocking tests met with disappointment. The processors' high heat output and basic air-cooling isn't best conducive to overclocking. Using the base 12x multiplier the FX-53 became unstable at 212MHz driven clock (2544MHz) with 1.6v. Model 3800+ fared little better, with a top stable speed of 214MHz or 2568MHz. Either our samples are below par or AMD's got its work cut out if FX-55's to remain at 1.5v.
Pre-benchmark notes
It's important to note that MSI's K8T Gem shipped with a BIOS that had the 2T command option enabled. Disabling this, that is, forced the board into running a 1T memory command. Benchmarks were given a healthy rise, with 3DMark2001SE rising by more than 300 marks and Pifast over a second faster. Yours truly ran all the benchmarks using 2T then realised the error of his ways. So the following set of results were obtained in the performance 1T mode.
We've got 3 different 3.4GHz Pentium 4s and 3 different 2.4Ghz AMD Athlon 64s, as well as a 2.2Ghz S754 Model 3400+ thrown in for good measure. Each processor can be used in a cutting-edge system, so which is the one to go for. Let's find out as we move on to the eagerly-awaited benchmarks.
