Asus P4S800D-E Deluxe & Gigabyte 8S655TX Ultra: SiS 655TX Challenges Intel's Best
by Wesley Fink on December 9, 2003 1:52 PM EST- Posted in
- Motherboards
Gigabyte 8S655TX Ultra: Stress Testing
We performed stress tests on the 8S655TX Ultra in these areas and configurations:1. Chipset and motherboard stress testing, conducted by running the FSB at 242MHz with synchronous (1:1) memory speed (DDR484) with OCZ 4200 EL memory.
2. Memory stress testing, conducted by running RAM at 400MHz with 2 Dual-Channel (1 bank) DIMM slots filled and at 400MHz with all 4 Dual-Channel (2 bank) DIMM slots filled at the lowest memory timings possible with Mushkin PC3500 Level II memory.
Front Side Bus Stress Test Results:
We ran a full range of stress tests and benchmarks to ensure the 8S655TX was absolutely stable at the overclocked FSB speed. This included Prime95 torture tests, and the addition of other tasks — data compression, various DX8 and DX9 games, and apps like Word and Excel — while Prime95 was running in the background. At default voltage, 242MHz was the highest overclock that we were able to achieve with a synchronous memory setting.Memory Stress Test Results:
This memory stress test is very basic, as it simply tests the ability of the 8S655TX Ultra to operate at its officially supported memory frequency (400MHz DDR), at the lowest supported memory timings that our Mushkin High Performance ECC Registered Modules will support:| Stable DDR400 Timings — 2 DIMMs 2/4 DIMMs populated — 1 DC bank |
|
| Clock Speed: | 200MHz |
| Timing Mode: | 128-bit DC |
| CAS Latency: | 2.0 |
| Bank Interleave: | N/A |
| RAS to CAS Delay: | 2T |
| RAS Precharge: | 6T |
| Precharge Delay: | 2T |
| Command Rate: | Auto |
We had no problem running 2 DS 512MB DIMMs of our standard Mushkin PC3500 Level2 or 2 DS 512MB OCZ 3500 Platinum Ltd in the 8S655TX Ultra at the most aggressive timings. We did achieve complete stability at 2-2-2-6 timings at the default voltage setting. However, our past experience with Gigabyte SiS designs would suggest that the 8S655TX may actually be running closer to 2.6V at “normal” settings. We are pleased that default voltage was fine for 2-2-2-6 timings, but you should not make too much of this finding.
Filling all available memory banks is more strenuous on the memory subsystem than testing 2 DIMMs, but we had no problem at all running with all 4 DIMMs filled providing 2 Dual-Channel Banks.
| Stable DDR400 Timings — 4 DIMMs 4/4 DIMMs populated — 2 DC bank |
|
| Clock Speed: | 200MHz |
| Timing Mode: | 128-bit DC |
| CAS Latency: | 2.0 |
| Bank Interleave: | N/A |
| RAS to CAS Delay: | 2T |
| RAS Precharge: | 6T |
| Precharge Delay: | 2T |
| Command Rate: | Auto |
The required memory timings with 4 DIMMs were the same as 2 DIMMs, the fastest 2-2-2-6 timings our memory could achieve. The 4-DIMM setup worked fine at “normal” memory voltage, the same as our 2 DIMM setup. The 8S655TX Ultra was completely stable at these timings and voltage.
Facebook
Twitter
RSS
24 Comments
View All Comments
FFS - Wednesday, December 10, 2003 - link
Q? to Wesley does these mobos have control of CPU fan speed (again like MSI Neo2), so it run more slow with low CPU load (= more silent)?Another one is how it's compatible with Radeon 9700 Pro? F.e. My SIS648 even with latest SIS-AGP drivers have have a conflict with Hercules R9700Pro when Fast Write is turned on...
Competition is G R E A T E thing,... but you know we also want stability :))) So far, as I know Intel chipsets before were far more stable and compatible with other hardware...
The bottom line "as fast" does not always mean "as good"... :(( And since I also love competition let's hope that SIS655TX as good as 875P.
ripdude - Wednesday, December 10, 2003 - link
A good article on a serie of good boards, I'm very happy you took the P4C800-E dlx in the test.I'm more then happy with the increased competition these boards bring, the P4C800-e dlx is still about 199 in euro land.
SmoiL - Wednesday, December 10, 2003 - link
When is the Asus P4S800D-E coming out? Nada on pricewatchWesley Fink - Wednesday, December 10, 2003 - link
Several of the cells in the Feature Table on page 3 were reversed and have now been corrected.Dual Channel (2x64 bit) is called 128 bit mode on the SiS. The BIOS has the option to run 128-bit (Dual-Channel) mode or to force Dual 64-bit.
FFS - Tuesday, December 9, 2003 - link
Also, as I remember VIA VT6307 support 3 (not 2) 1394a FireWire ports (as on MSI Neo boards), so why Gigabyte would cut them to 2 ???And besides Wesley always trying to stress out color-coded panel connectors.
For me it's more important that you can change RAM without taking off videocard and ASUS had such problems with some of their mobos (now it looks better... but talking about overclocking ... hmmmm.
Even it's damn stable: heatsink on ASUS SiS 655TX is passive (already -) and too close to CPU (plus wrong orientation - same mistake as with original
P4C800 Deluxe. So - NO oversized cooler for CPU on this board.
FFS - Tuesday, December 9, 2003 - link
Realtek ALC658 codec AC '97 2.36-Channel with UAJ Shoud go to Gigabyte and vice versa...
Oh well.., I had my SIS648 for a while and .. so now I'd rather prefer to pay eXtra for board with Intel chipset :((
Plus Gigabit LAN will work faster(=better) with CSA...
FFS - Tuesday, December 9, 2003 - link
Hey guys... In the features table on-board audio codecs are mixed up :)Shalmanese - Tuesday, December 9, 2003 - link
Er, I don't think Sis supports 33Mhz Memory speeds... (pg 5) ;).What happened to the tech support turnaround time bit? That was always the favourite part of your reviews.
Tokelil - Tuesday, December 9, 2003 - link
On page 3 it says its the Gigabyte MB that uses the Marvel Gigabit controller, while on page 4 it says its the asus board that uses the Marvel controller... My guess is it is Asus that uses it and its wrong on page 3.FishTankX - Tuesday, December 9, 2003 - link
What on earth is the difference?There is no commercially avaliable 128 bit DDRSDRAM. All DDRSDRAM is 64 bit.