![]() |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
:
News
: Archives
|
|
BFG Tech 590 SLI Motherboard Review |
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
BFGRAN590S… or Something |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
February 26, 2007 |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
By Josh Walrath |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Sciencemark 2.0 This handy little number takes a good look at CPU and memory performance in both synthetic and “real world scientific” style tests. Since this does not stress other components on the motherboard, we can expect similar results from each board.
Since both boards run the X2 5000+ at 2611 MHz, the L1 and L2 values are nearly identical. With this benchmark we see confirmation that the memory on the BFG board is running slightly slower than that of the Asus. The Primordia and MolDyn benchmarks also show parity between the two boards (less than 1% difference).
The southbridge cooler is a nice copper affair with the fancy colored cap. The space between the graphics card and the cooler is minimal, but should not impact the seating of any standard PCI-E card. 3D Mark 2006 3D graphics are probably one of the most stressful things to be done on a computer. From the graphics card, to the PCI-E bus, to the CPU, and finally the memory a 3D graphics load can push multiple components to 100% utilization. With 3D Mark 2006 we get a good feeling of where a motherboard sits in connecting all the parts so they can work together in harmony. Default 3D Mark settings were used.
As we can see both boards are essentially identical when it comes to performance. Note as well the CPU scores are pretty much the same, so we can assume it is not just because of the video card that the scores are so close. Commanche 4 I resurrected this benchmark because it is so nicely CPU bound. I set the resolution at 1600 x 1200 and watched the bullets and missiles fly.
The two 590 SLI boards are again very much in the same league when it comes to performance. No appreciable differences between them. Realstorm 2004 Benchmark This realtime raytracing benchmark is 100% CPU and memory limited. As such it makes for a wonderful benchmark in these situations. I set resolution to 640 x 480 with the rest of the settings at default.
We see a little bit of separation here between the boards, with the Asus coming out consistently on top. These are not huge differences by any stretch of the imagination, but they do show a pattern. WinRAR 3.62 For the final test I compressed a 1 GB folder into a .rar file and timed how long it took.
Here we finally see some separation in the benchmark results. For some reason the BFG board is slower than the Asus board. 34 seconds is pretty significant. I am not sure where the issue lies, but the results are reproducible here.
Though this would rarely happen in the real world... both cards fit nicely onto the board. The 4X PCI-E and bottom PCI slot are still exposed and usable. Overclocking Honestly, I was not expecting much from overclocking this setup. Earlier reviews have the board being able to reach between 270 MHz HTT to 295 MHz HTT. Very few people were able to get the board above 300 MHz HTT, if any at all. With a small amount of work I was able to take the BFGRAN590S to 314 MHz HTT. Apparently the rev. 2.0 board has allowed a little extra overclocking room. While this is not as good as my recently reviewed T-Force 550 which runs at 340 MHz HTT, it is a hefty jump up from previous reviews I have read. Stability was rock solid at 314 MHz, running for several days straight without issue. Once I increased HTT to 316, the motherboard failed to boot altogether. There were no strange performance issues while overclocked, and all benchmarks tested showed performance gains and no drops. I set HTT to 314 MHz and put the X2’s multiplier at 9, giving a clock speed of around 2820 (give or take a few MHz). Memory was right around 403 MHz (806 MHz effective) using the same 4:4:3:8 timings. Here are a couple of benchmarks showing the mild overclock.
The 200 MHz overclock did produce higher results, but the most interesting numbers are the memory. An extra 750 MB/sec is pretty impressive. Remember, at stock speeds the memory is running around 743 MHz or so. The extra CPU speed and extra MHz give bandwidth a nice little bump.
If you have found this article interesting or a great help, please donate to this site.
Copyright 1999-2007 PenStar Systems, LLC. |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||