Quick personal test, 680i vs X48

[skypine27]

I recently became sick of my 680i Striker Extreme mobo, it seemed to have too many problems with my overall setup. So, I made the switch to a Rampage Formula X48. Even though I'm a hard core gamer, I'm totally against even the concept of SLI and Crossfire, so SLI support was not an issue for me. (However I'm not against "self contained/transparent SLI" like a GX2).

Anyway, I ram some benchmarks using all my same system components on the 680i, and then after the swap to the X48.

680i 3D Mark 06:

http://img203.imagevenue.com/i...am_mark06_122_17lo.jpg

X48 3D mark 06:

http://img174.imagevenue.com/i...am_1066_122_1040lo.jpg

14449 (680i) vs 14526 (X48). Basically identical, well within the standard 3% margin of error, as far as 3D peformance goes. However, check out the CPU side of things using PCMark 05.

680i Pc Mark 05:

http://img179.imagevenue.com/i...pcmark05_122_550lo.jpg

X48 Pc Mark 05:

http://img139.imagevenue.com/i...ram_1066_122_956lo.jpg

9925 (680i) vs 11638 (X48). That's a 17% increase for the X48. Definitely an improvement and not a fluke.

Now the sad news, Crysis:

I use the included benchmark; file in the Bin64 directory, Benchmark_GPU.bat.

Unfortunately, using the exact same settings (including CPU overclock, in game graphical settings, and BIOS memory speed and latency) my new X48 just got its butt handed to it by my older 680i.

The Benchmark_GPU.bat used to give me an average FPS of about 29 when running through the test on my 680i Striker Extreme. It now runs a flat 19 FPS average when running it on the X48, again with all the same settings. :-( A huge loss for the X48.

I first noticed this when loading an old save game from Crysis (after I reinstalled it on my new system) and was like OUCH! Something is wrong. It was playing noticeably more sluggish than before. So I ran the benchmark and confirmed it.

In any case, though the PC Mark CPU scores are much higher on the X48, the 680i definitely games better on Crysis. (Again, using all the old system components and same settings...) Not sure if this is due to the newest nvidia drivers working best w/ nvidia chipsets or what. I am on vacation and have not had time to test any other games after making the switch to the X48.

I ran each test several times, with various small background programs running and then shut off (ie, free download manager, and/or Everest Ultimate) to make sure nothing was messing with the results.

Just food for thought.

 

[Sylvanas]

Hmm well I wouldn't use Futuremark tests to determine whether something is 'better' than another, try a Winrar/SuperPI bench. But needless to say changing a mobo will not result in a 10fps difference, obviously something else is at play here most likely drivers/individual settings. Did you do a fresh install when you installed the X48? Are the memory latencies/CPU speed the same?

 

[skypine27]

Yes, I did a fresh install when I went from the 680i to the X48. My boot drive is a Raid 0 setup, and the X48 uses different a different SATA controller than the 680i, so the new mobo wouldn't even recognize the old drives until a new Raid had been created/formatted.

I run all the same settings. Memory Speed, latency, CPU overclock, etc, installed Vista programs, nvidia video drivers, etc. All identical to the previous setup.

I'm not sure whats causing the 10FPS loss in Crysis with the X48 because, like I said, I kept everything else the same. Everything.

I went on vacation the day after I setup the new mobo, so the only game I had time to test it with was Crysis. The only other game I was recently playing on the 680i was Stalker, and will test that with the X48 to see if there is a FPS loss with it as well.