Memory faster

[Puma/Sire]

I played "GT Legends" With my memory at at 570mhz with like 3337 latency and it ran the game slower than when I had the memory at 400 mhz at 2225 latency. Would 2 Gigs of PC 3200 run games faster than 2 gigs of pc 4000 because of higher latency sets, because I can always keep my cpu at 2.6 gigs and get the geforce 7900gtx?

[red930]

Having your memory at 570MHz 3337 timings would actually be faster then 400MHz 2225. I don't think that is the issue unless your memory is unstable? Have you ran Memtest86 on that frequency for 8-10hrs or even tried running Test#5 for 25-30 passes? Do you have your video card overclocked? You have no overheating issues? I think it would be better to run at higher frequencies with loosin timings. Did you really get 570MHz or 470MHz with that particular memory? Have you tried different games on the faster memory other then GT Legends?

[Puma/Sire]

Yes it passed prime 95 at 570 mhz for 15 hours with tighter timings than 3337. BUt fails test#7 at 3337 timing. But is passed prime 95 at 9 hours 4 times in a row. But usually would fail at prime 95 the second or third test when I had the timings tighter and passing mem test. It also failed mem test#7 at 2225, 400 mhz but passed prime 95 4 times in a row. But I lowered the row cycle from 3120 something like that to the next lower time wich is like 2000 somthing, and it passsed Test #7 I was in aw. So I'm going to jump back up to 570 agian with 3337 and see what happens . Yes Call Of Duty 2 ran faster with lower mhz and tighter timings, for example, IN intense moments in the game the games would lag much more at 570 mhz than when it was at 400mhz. I haved the Video Card overclocked to 395 / 1075 and runs all games fine with no fragments I had to eventually get to those stable clocks. The PC runs very cool , the cpu idls at 32degrees the vide card max load is like 78 degrees and the ambient temperture idles AT 40 degrees. I never ran test number 5 by itself

[RGone]

The question of tighter timed but slower running memory versus looser timings and faster running memory will never be answered fully. There are too too many exceptions to the "rule" that everyone wishes to have in place or try to make.

 

In general the conditions that allow you to run the same cpu total mhz with the memory running @200 FSB> 2, 2, 2, 6 will bench faster in 3dmarks than memory running at 250mhz @ 2.5, 4, 4, 10.

 

But and the but is huge; when you move to 2 gig sticks of memory the entire situation changes since there are NO 2,2,2,6 sticks of memory available actually. So the 'rule' is of no value. The later memory hogging games, all like to run with 2 gigs and the "system'' performs better overall with 2 gig sticks because of the greater amount of memory available and the game not having to go to swap file or virtual memory and this makes the whole timings arguement more or less of no value because the looser timed 2x2gig sticks appear quite more effective generally in gaming use even if the timings are loose.

 

So the 'rule' is busted again and everyone goes back to clamoring my stuff is faster and better than your stuff. Woohoo woot woot. It is a balancing act that each really needs to prove to himself.

 

RGone...

[Puma/Sire]

Yeah that crossed my mind that if I had 2 gigs at 500 mhz I'd be in a bettter situation I just wanted to be sure since I never had 2 gigs to try out. Thats what sucks about pc hardware , once you buy it your stuck with it.