Benchmarks: 3D performance and storage subsystems
We're left with the gaming and storage benchmarks now.
The 100MHz boost to the clock speed of the RS485M-M's integrated graphics compared to the RS482-M's does yield a performance improvement, but it's of little use seeing as the game will still be unplayable. The X1300 gets slightly better framerates in the Socket 939 platform.
It's exactly the same story in Quake 4, except with even worse framerates from the integrated graphics solutions.
In Splinter Cell: Chaos Theory at 1024x768, the X1300 performs the same in both boards, while the Xpress 1150 graphics keep their (still unplayable) edge.
The percentage drop in performance for all graphics configurations is pretty much the same when the game resolution is upped to 1280x1024.
The Radeon Xpress 1150 graphics isn't going to win any gaming performance awards, then, but who would have expected it to? It might have been nice to have integrated graphics on an AM2 board with a bit more grunt than that, maybe somebody will surprise us in due course? Integrated graphics, it seems, are still only fit for rendering at 800x600 with minimal quality settings. That, as we remember, was the case a few years' ago, too.
However, an X1300 combined with this board and our RAM/CPU configurations should give a playable level of graphics performance at acceptable (but not breathtaking) resolution and quality settings.
Storage
Here are the SATA performance results from HD-Tach:
Must we really comment? If we were to be really nit picky, we'd say the RS482-M performed fractionally better, but it seems unfair to say such a thing... they both perform the same.
SB460 outperforms SB400 in USB throughput, but in reality the performance is still dire. Seeing as USB performance won't be critical to every users, some will be happy with the performance of the southbridge, while for others it won't be satisfactory.
Look at performance in some areas and ECS's AM2 board trumps its S939 offering, but in others the tables turn. Either way, the difference is minimal. We have to wonder what would happen if the RAM had been running at DDR2-533? Well, we know what would have happened... the RS485M-M would have had an advantage in benchmarks that depended largely on system memory. The integrated graphics might have benefited too (seeing as HyperMemory is used, thus the system RAM is used by the Xpress 1150 chipset).
