... benefit incredibly from lowering the CAS latency. Basically, going from CAS3 to CAS2 would be like going from 700MHz to 800MHz, and sometimes more.
Actually, this doesn't apply to moldyn, which usually fits in the processor L1 and hence is more of a pure fp test.
I'm on and off thinking about writing a CAS Latency article with (tons of) benchmarks from my Duron-800 testbed. I have had some (minor) problems getting my PC133 memory working at CAS2 100MHz, though.
i can say that there is most definatly a noticable diffrence. If you work with any large files or deal with any streaming media you can feel the increase in streaming data preformance. That extra clock cycle really can show a negative effect.
In all of my memory comparisons using pc 133 222 rated crucial memory i find that running your memory at pc100 cas 222 is overall under 2% slower then running pc133 at cas 333. However when using streaming media i find the pc100 case 2 to actualy outpreform pc 133 case 3.
Its no fluke that a premimum is charged for cas 2 pc133 over cas 3, its due to the fact that there IS a real world preformance increase. I cant say cas2 will garner you any super FPS increase under q3 but id be willing to bet it will add a few frams, i find it overall i worth the extr 10 or 15 bucks.
Bottom line - cas 2 is faster. Does it feel faster than cas 3? Who knows. You can barely, if at all, feel the difference between a P3 500 Katmia and a 933EB. Noticable speed differences lies in harware such as video cards and CDR/RW speed ratings.
Go cas 2 to if you can. Every little bit helps. Don't go by Sandra, it is not real world. It does prove that cas 2 is faster, which is no revelation.
I've been lurking on these boards for a long time...kinda of a long time reader, rare poster. And i have to say it's amazing that for such a generally smart population of guys and gals how easy it is for so many to be suckered into believing hype. All people on here ever do is post their XXX scores and sh!t like that. I can't believe how many need these great NON-real-world scores. I'm willing to bet anyone anything that you could not sit down at 2 rigs running the same stuff and pick the rig running cas2 vs. cas3. People dicker over the most sensless stupid scores, it's a shame.
Mother
A little harsh there Mother. Glad you stopped lurking and posted, but don't offend peoples hobbies. As I alluded to above, the diefference is probably undetectable and Sandra is not real world. BUT, I do understand the thrill of building a rig and benching it. It is a way of seeing our results, because as you said, running real world apps will generally not prove sh!t. People that buy from Dell and plug it in do not hang out here. People that tweak and are harware junkies do. Just be repsectful to other members.
I would like to know what all the other timings mean in my BIOS. And why "2-2-2" and "3-3-3" in naming the memory?? What does that refer to exactly?
Yes, PC133 cas3 or unspecified, does not run well always at CAS2. My mobo set it to cas3 automatically. I overrid it, and ME crashed like crazy. I sent the biitch back and got CAS2 PC100.
Let's say you get a-few-percentage improvement from upgrading from 600-MHz CPU to 800-MHz, another few more percentages from going from PC-100 to PC-133 memory, another few more percentages from going from CAS2 to CAS3, even another few more percentages from going from ATA-66 to ATA-100, etc., etc.
in my system - p3 @ 866 133mhz fsb, geforce2. at 640x480 there's about a 10fps difference between cas2 and cas3. (like somewhere around 135fps - 145fps)
But since real people play at 1024x768 or higher with that kind of system, you only see about a 1 or 2 fps differnce since the video card becomes the bottleneck.
edit - oh, and to answer the original question - There's no dangers in trying cas2. It's just like overclocking your cpu. You can always turn it back to cas3 if it crashes a lot while running at cas2.
Pretty much all high quality cas3 ram can run at cas2 anyway, so give it a try.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.