tighter timings or higher clock

[martini161]

i have been messing around with my memory somemore, and i was just wondering, what would be faster: 5-4-4-10 at 900 mhz, or 6-6-6-20 at 1066?

[kenc51]

i have been messing around with my memory somemore, and i was just wondering, what would be faster: 5-4-4-10 at 900 mhz, or 6-6-6-20 at 1066?

900

[Bio-Hazard]

Bench the rig and see which one does better...........

It works both ways sometimes depending on if you can get high enough to offset the looser timings.........

[ibleet]

Interesting topic...anyone here do any actual benches on this?

There seem to be arguments going both ways on this subject. I would recommend running some ram speed tests (passmark?) each way and see which is best. Anyone here know of a good memory benchmark utility?

Or if you just want to check gaming, set up FRAPS and run the ram each way and see which delivers more FPS in the same part of the game. I highly doubt that the numbers will be that much different though.

It will be different for everyone depending on what types of apps you run, your hardware, and your system configuration, ie. overclocking.

[Zertz]

Sandra and Everest are good for benchmarking memory. The difference in real-world applications will be next to nothing.

I can do a few tests on my AMD computer

[martini161]

yea, i thought i could do 1066, but my memory could handle it without something like 2.5 volts, which im not comfortable with

[Methious]

I ran some benches on my new OCZ Reaper and posted them down in the AMD Motherboard section in the M3A Swamp Thing Thread. Just got done reloading fresh XP and SP3RC and going to rerun some stuff anyway I'll try to get some more posted.

Btw I did see a pretty decent performance increase after installing SP3. Saw the on going discussion about it and ran 3dmark06, Lightsmark, and AquaMark before and after. 3dmark went from 10500 no OC to 11100+ no OC, Lightsmark went from 350 to 367FPS, and AquaMark went from 122.XX to 133.XX. That was all under the fresh install. If you do the math that's an almost consistent 9.5% performance increase across the board with all variables kept the same except the addition of SP3RC. (Downloaded from windows update)

Any body know if you can slipstream SP3?