I was just about do the same on socket 939 though and I did a performance test. With 1T I got 113 fps on the source stress test (1280x1024 aa x4, af x8, all settings maxed.) Then I ram it at 2T just to see what it would be like with 4 dimms and I got 112 fps. As for latency in everest home edition, it went from 40 to 45, memory read basically stayed at 2700 mb/s but write dropped from 2600 to 2100. But as I said before, the source engine (cs:s source+hl2 +more to come) is actually pretty cpu dependant (ram+cpu) so any changes to the cpu/ram should result in a significant change to fps yet it didn't do anything. I ram the test a couple more times at 1T and 2T to make sure there were no bad renders or anything. Results stayed basically exactly the same with 2T actually getting a higher fps a few times because I was able to increase my memory speed due to the loosened 2T timing.