dreamsr Posted October 15, 2006 Posted October 15, 2006 What would you regard as the best CAS latency for dual channel memory sticks. Here is an example: Memory CAS Latency: 3-4-4-8-1 Is this too high? Share this post Link to post Share on other sites More sharing options...
Praz Posted October 15, 2006 Posted October 15, 2006 If you had a detailed listing of your system in your sig per the rules we would be able to answer your question. Share this post Link to post Share on other sites More sharing options...
cuoficr Posted October 15, 2006 Posted October 15, 2006 Memory timing's have a pretty mute affect on real life performance in S939 applcation's. The only time you'ld really notice the difference is in benchmarking. At DDR500+ with 2x1gb kit's 2.5 is very hard to achieve, especially for UCCC, with 3 cas being the normal, I run my OCZ at DDR500 3-3-7-3 bit I notice no real difference from that to 3-4-8-4. Hope this answer's the question you asked, since you are asking in the recomendation's section it's safe to think that you are asking in general and not related to your signature. Share this post Link to post Share on other sites More sharing options...
Kingfisher Posted October 15, 2006 Posted October 15, 2006 Depending on the size , some folks can run their CAS @ 1.5. Any 2 GB kit will more or less run 2 or 3 depending on the model and frequency. Share this post Link to post Share on other sites More sharing options...
dreamsr Posted October 15, 2006 Posted October 15, 2006 Thanks guys, I was infact asking generally. I am thinking of getting 2GB kit dual channel memory and I'm just wondering what factors to consider other than pricing of coure! I'm going to be running sli with a pair of eVGA 7900 GT KO video cards. Share this post Link to post Share on other sites More sharing options...
Kingfisher Posted October 15, 2006 Posted October 15, 2006 If you are overclocking, look for DDR500( pc4000 )2 GB kits. Crucial Ballistix, G.SKILL , OCZ and Mushkin offer some really nice choices. If you are not going to overclock, DDR400(pc3200) is in order. Share this post Link to post Share on other sites More sharing options...
Angry_Games Posted October 15, 2006 Posted October 15, 2006 forget about timings they are moot stick with something that has some headroom to overclock if anyone tells you timings make a big difference, tell them they are full of crap and then link them to this: http://www.dfi-street.com/forum/showthread.php?t=40178 Share this post Link to post Share on other sites More sharing options...
cuoficr Posted October 15, 2006 Posted October 15, 2006 forget about timings they are moot stick with something that has some headroom to overclock if anyone tells you timings make a big difference, tell them they are full of crap and then link them to this: http://www.dfi-street.com/forum/showthread.php?t=40178 So do I get credit for already saying it Share this post Link to post Share on other sites More sharing options...
old_geekster Posted October 15, 2006 Posted October 15, 2006 So do I get credit for already saying it Duely noted! Share this post Link to post Share on other sites More sharing options...
Angry_Games Posted October 15, 2006 Posted October 15, 2006 So do I get credit for already saying it you would have...except you didn't link to the thread where it was proven that timings have no real effect (other than minor effects that only truly anal arguing monkies will complain about) Share this post Link to post Share on other sites More sharing options...
Thraxz Posted October 15, 2006 Posted October 15, 2006 *It doesn't have a substantial effect on gaming and graphics. It does, however, have an effect on non-GPU reliant things like video transcoding and compiling. Share this post Link to post Share on other sites More sharing options...
Angry_Games Posted October 16, 2006 Posted October 16, 2006 *It doesn't have a substantial effect on gaming and graphics. It does, however, have an effect on non-GPU reliant things like video transcoding and compiling. no, not really I've found nil difference in overall performance on a variety of programs so maybe 2-2-5-2 will encode a dvd 3 minutes faster than 3-4-8-4 when testing same cpu. wow (not meaning to sound snarky...but 3 minutes out of 90+ is not really something anyone is going to jump up and down about). I just tested encoding a 1.2GB raw DV file into a 120MB WMV on 4 different machines, from timings ranging 2.5-3-8-3 all the way up to 5-6-6-15 and the only real difference is the strength of the cpu...and on the same cpu, there's no difference when it comes to Intel to Intel or AMD to AMD that I have seen.... again, this is myth stuff that you guys keep believing because of the old days when memory timings and 1T was everything (AthlonXP that really relied on tight timings). A64 and Core2Duo...the cpu is the strength, not the memory timings, bandwidth, nor command rate. Share this post Link to post Share on other sites More sharing options...
Recommended Posts