- Jan 20, 2010
- 2 (0.00/day)
- Hamburg, Germany
|Cooling||Eheim HPPS, Alphacool NexXxoS XP rev.2, EK Water Blocks EK-FC280GTX SLI, Thermochill PA120.3 [old]|
|Memory||OCZ Reeper / PC2 6400 4x1gig|
|Video Card(s)||2x Nvidia GTX280|
|Storage||2x WD Raptor ADFD 74gig Raid-0, 1x Seagate Barracuda 7200.10/320gig|
|Display(s)||LG FLATRON W2242T|
|Case||Antec Performance One Mini P180 - ATX`mod|
|Audio Device(s)||X-Fi Titanium|
|Power Supply||Corsair HX850|
Just removed SLIPatch and installed Halmod .9a, SLI still enabled even through reboot. GPU2 now appears to be working and GPU1 is showing the same 30-40% usage as it did when the system appeared to be on only 1 GPU. It's like the GPU's are working but aren't working together. I'm thinking an SLI bridge is going to unlock this puzzle.
the tricky thing is: not all games are using multi-GPU setups optimized. That means: u never will benefit from power of your SLI-setup in your results as expected -> if u play "the wrong" [not multi-gpu supported] game. Doom3 for example: Both cores are running here with only 30-40%! The timedemo in Doom3 with SLI-setup is about 25% SLOWER, as with only a single card Battelfield 2 on the other hand: the FPS has nearly doubled by using SLI and the Vantage bench shows about the double performance, with SLI-setup
So for all who want to go SLI: keep in mind that it is very helpfull to know what your system with only one card can go in benchmarks and games..(write it down) to know! later exactly what is going on and you are talking about, when you are using SLI.
The SLI-bridge in use: it`s about ~ +900 points here in Vantage bench.
sli patch 0.4 + NVidia 196.21
sli patch 0.55 + Nvidia 196.34 //replaced with 197.13
Best regards TeleX