The difference between 1600MHz and 2000MHz is very slim, I don't think it's worth the tweaking. If you have the time and willpower though, why not?
Help with optimal balance G. SKILL Ripjaws DDR3 2000 on my i5 750/GIGABYTE UD2 rig...
The i5s were launched right as I was about to do a quick semi-portable build to take with me for a few months of international travel, so I slapped together a little SFF box that I've been tweaking when I've had time over the past month and a half.
First let me say, the parts are not optimal, but I build for friends and as of right now, just about all of the parts in this rig are spoken for when I get back and am ready to start my high-end 1366 build.
That said, here's where I'm at. I'm doing VERY mild overclocking on the stock cooler (mostly under 3.2GHz) until I put in an aftermarket cooler. My RAM is the G.SKILL Ripjaws DDR3 2000mhz (when I bought it at $100 the price difference with slower RAM was almost non-existent.)
Now the main problem here is obviously that the memory multiplier is locked at 10x max on the i5 750, so I have to bring my base clock up to 200 to run my RAM at its intended speed. Because I'm passing these parts on to friends I don't want to push them insanely hard (though they're fully aware of what I do) so I try to keep my cores at a max of 70c and CPU at no more than 60c. Luckily its been pretty chilly here in Paris so my ambients are low, but I still don't have far to go before reaching my max's.
I haven't had much time to do detailed benchmarks but I've been using the full Passmark tests as a loose comparison between configurations. This is a rough summary of what I can run 24/7 (I think?) with my RAM at only 1500mhz.
150 BCLK x 21 = 3.15Ghz - 9h45m stability test 4 core average of 70c, CPU temp at 60c, 1.1vcore
RAM Speed : 775 MHz (2:10) @ 9-7-7-21
I've also been able to get 155x21 mostly stable at 1.09v for 3.25GHz but have had some trouble tweaking my settings properly to get back to that at safe temps.
Now obviously the timings are significantly tighter running at 1500mhz, but what my question really comes down to is: would I still be better off running with a base clock of 200? I'm about to do a proper stability test again (I can't find my stats I originally saved) but I believe to achieve stability/safe temps in this scenario I had to run at 2.8 or possibly 3GHz. Even at 3GHz, am I better off than I would have been in the previous scenario? I've tried 166 and 180 but have had trouble finding stability at decent speeds.
Though I've built plenty of systems and keep up to date with the more theoretical tech aspects, RAM is something I've always had trouble understanding in terms of its real world performance, so this may be a lot simpler than I'm making it. Thanks for your help.
[I'm going to edit to add a few more BIOS details after I reboot]
First let me say, the parts are not optimal, but I build for friends and as of right now, just about all of the parts in this rig are spoken for when I get back and am ready to start my high-end 1366 build.
That said, here's where I'm at. I'm doing VERY mild overclocking on the stock cooler (mostly under 3.2GHz) until I put in an aftermarket cooler. My RAM is the G.SKILL Ripjaws DDR3 2000mhz (when I bought it at $100 the price difference with slower RAM was almost non-existent.)
Now the main problem here is obviously that the memory multiplier is locked at 10x max on the i5 750, so I have to bring my base clock up to 200 to run my RAM at its intended speed. Because I'm passing these parts on to friends I don't want to push them insanely hard (though they're fully aware of what I do) so I try to keep my cores at a max of 70c and CPU at no more than 60c. Luckily its been pretty chilly here in Paris so my ambients are low, but I still don't have far to go before reaching my max's.
I haven't had much time to do detailed benchmarks but I've been using the full Passmark tests as a loose comparison between configurations. This is a rough summary of what I can run 24/7 (I think?) with my RAM at only 1500mhz.
150 BCLK x 21 = 3.15Ghz - 9h45m stability test 4 core average of 70c, CPU temp at 60c, 1.1vcore
RAM Speed : 775 MHz (2:10) @ 9-7-7-21
I've also been able to get 155x21 mostly stable at 1.09v for 3.25GHz but have had some trouble tweaking my settings properly to get back to that at safe temps.
Now obviously the timings are significantly tighter running at 1500mhz, but what my question really comes down to is: would I still be better off running with a base clock of 200? I'm about to do a proper stability test again (I can't find my stats I originally saved) but I believe to achieve stability/safe temps in this scenario I had to run at 2.8 or possibly 3GHz. Even at 3GHz, am I better off than I would have been in the previous scenario? I've tried 166 and 180 but have had trouble finding stability at decent speeds.
Though I've built plenty of systems and keep up to date with the more theoretical tech aspects, RAM is something I've always had trouble understanding in terms of its real world performance, so this may be a lot simpler than I'm making it. Thanks for your help.
[I'm going to edit to add a few more BIOS details after I reboot]
1 - 6 of 6 Posts
Ram is not worth tweaking unless you're just running benchmarks. I would just tighten up timing to 7-7-7-20 at 1500 and be done with it
Case closed, much thanks to both of you. One more question I'd never make thread for, and am a bit embarassed to ask about: I understand why the four standard measurements with RAM are the most discussed, but would it be adviseable (and I do have the time) to mess with anything else, such as the tRFC?
Quote:
It doesn't make difference in performace. If you're running into stability issues, then you can increase that, but otherwise just leave it on auto.
Originally Posted by infiniteengine Case closed, much thanks to both of you. One more question I'd never make thread for, and am a bit embarassed to ask about: I understand why the four standard measurements with RAM are the most discussed, but would it be adviseable (and I do have the time) to mess with anything else, such as the tRFC? |
The other timings are for fine tuning. Some increase performance by a marginal value and some are pretty much useless performance and stability-wise.
1 - 6 of 6 Posts
- This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Overclock.net
A forum community dedicated to overclocking enthusiasts and testing the limits of computing. Come join the discussion about computing, builds, collections, displays, models, styles, scales, specifications, reviews, accessories, classifieds, and more!
Full Forum Listing
Explore Our Forums
Top Contributors this Month
View All
HappySheep
367 Replies
GRABibus
206 Replies
Ichirou
201 Replies
Recommended Communities
