Archived

This topic is now archived and is closed to further replies.

RickSinGA

New Build, right ram?

Recommended Posts

Will be putting together a new build this week. ASUS P8P67 Deluxe R3 MB with an I7-2600 BX80623I72600K. I will be using the Noctua NH-D14 cooler and wasn't able to reuse my existing 6GB Corsair Dominator DDR3-1600 triple channel memory because of the high cooling fins on the dimms. The salesman convinced me to buy 8GB Corsair XMS3 CMX8GX3M2A2000C9 which is 9-10-9-27. I plan to try get it up to around 4.5G which seems promising looking at this forum.Will I be okay? Really don't want to drive the 2 hour trip back to Atlanta to exchange them.Yeah, gave up trying to up my 920, think it was the H-50 cooler.Thanks for any helpRick S.

Share this post


Link to post
Share on other sites
Help AVSIM continue to serve you!
Please donate today!

I have the same memory. It won't OC to 2133 on my current mobo, so you will have to run it at 1866 instead. Sandy Bridge doesn't support 2000MHz. I would try any of these sets instead:2133MHz CL9http://www.newegg.co...N82E16820231435http://www.newegg.co...N82E168202262001600MHz CL7http://www.newegg.co...N82E16820226203http://www.newegg.co...N82E16820231431Cheap and will definitely fit under the NH-D14 --> http://www.newegg.co...N82E16820226201I'm fairly certain all the mushkin sets will fit, but I'm not really sure how high the gskills are. I've attached an image of the a Mushkin ridgeback set so you can see the height.

Share this post


Link to post
Share on other sites

I removed the fins of my Corsair Dominators before getting my current RAM sticks so they fitted under the NH-D14, and run them for months like that in an OCed I5 750 (4GHz) system, zero problemsAbout the suitability of those 2000 CL9 sticks, you'll find different opinions. Personally I think you'll be ok running them at 1866 CL8 or if you're more lucky than Corey at 2133 CL9 or CL10 (My 2000 CL7 will do 2133 CL8, but not CL7)

Share this post


Link to post
Share on other sites
I have the same memory. It won't OC to 2133 on my current mobo, so you will have to run it at 1866 instead. Sandy Bridge doesn't support 2000MHz. I would try any of these sets instead:2133MHz CL9http://www.newegg.co...N82E16820231435http://www.newegg.co...N82E168202262001600MHz CL7http://www.newegg.co...N82E16820226203http://www.newegg.co...N82E16820231431Cheap and will definitely fit under the NH-D14 --> http://www.newegg.co...N82E16820226201I'm fairly certain all the mushkin sets will fit, but I'm not really sure how high the gskills are. I've attached an image of the a Mushkin ridgeback set so you can see the height.
Dang, I thought it took DDR3-2000 memory, where did I get that?So, It won't work for OC'ing then? It's a long lonely drive and I don't believe Microcenter carries Mushkin! Not sure they will give back my money as many times as I have swapped out this week.ThanksRick S.

Share this post


Link to post
Share on other sites

They should work for OCing though... I have the same kit except the 1866mhz variant... Running @ 4GHz, but I'm at 9 10 9 27 1T. I think you'd be able to do better than that.

Share this post


Link to post
Share on other sites

I haven't tried dropping timings to CL8 yet, so I'll definitely report back if that's successful. I wouldn't worry about taking them back, Rick. Memory has such a small effect on FSX and so far these sticks do seem reliable at the very least. I know they will do 1866 at less than 1.55v! I will also try running them at 1600 CL7 and report back. I'm sure we can find some sort of good compromise here!

Share this post


Link to post
Share on other sites
I haven't tried dropping timings to CL8 yet, so I'll definitely report back if that's successful. I wouldn't worry about taking them back, Rick. Memory has such a small effect on FSX and so far these sticks do seem reliable at the very least. I know they will do 1866 at less than 1.55v! I will also try running them at 1600 CL7 and report back. I'm sure we can find some sort of good compromise here!
Thanks everyone.Cory, I was getting ready to head back this morning so your post was timely. Please let me know what you find. You should also know, if you haven't already figured it out, that I am not to knowledgeable about OC'ing. I have managed to do it but I have never messed with the memory setting's just CUP clock speeds and vCore. Any help would be greatly appreciated.Rick S.

Share this post


Link to post
Share on other sites

Rick,I tried several different settings and came up with two that seem stable so far. So far meaning about 20 minutes of prime95 blend.2133 @ 9-11-10-28-1T1600 @ 7-8-7-24-1TI had several other settings successfully boot into windows (including 1866 @ 8-9-8-24), however they tended to cause errors while running prime95 blend. Try either of the two settings above and see if you don't have any stability issues yourself. Out of curiosity, I tried decreasing DRAM voltage to 1.60v for both settings above. That made things unstable, so I suggest leaving them at 1.65v. I've attached some screen shots of the relevant bios settings.I think I'm going to run with the 2133MHz setting for a while only because it yields better results with the MaxxMEM benchmark. I've also attached a summary of those results based on a 3-test average.

Share this post


Link to post
Share on other sites
Rick,I tried several different settings and came up with two that seem stable so far. So far meaning about 20 minutes of prime95 blend.2133 @ 9-11-10-28-1T1600 @ 7-8-7-24-1TI had several other settings successfully boot into windows (including 1866 @ 8-9-8-24), however they tended to cause errors while running prime95 blend. Try either of the two settings above and see if you don't have any stability issues yourself. Out of curiosity, I tried decreasing DRAM voltage to 1.60v for both settings above. That made things unstable, so I suggest leaving them at 1.65v. I've attached some screen shots of the relevant bios settings.I think I'm going to run with the 2133MHz setting for a while only because it yields better results with the MaxxMEM benchmark. I've also attached a summary of those results based on a 3-test average.
Thanks Corey, I will try that when I get done. Probably be tomorrow though. I am only about half way through my build, UPS didn't show up with the cooler until after 7PM tonight. But they did show up:)BTW, you show 1866 memory and mine is 2000; is that the same?Rick S.

Share this post


Link to post
Share on other sites

It's the same 2000 memory as yours, but I list it as 1866 because that's what I've been running it at.

Share this post


Link to post
Share on other sites
It's the same 2000 memory as yours, but I list it as 1866 because that's what I've been running it at.
ahh

Share this post


Link to post
Share on other sites

Excuse me for butting in here but why would you buy a more expensive, higher than usable frequency ram and then downclock it?Just looking for an answer to a serious question from a guy trying to learn about this ram thing which seems to be discussed almost as much as CPUs.Regards,

Share this post


Link to post
Share on other sites
Excuse me for butting in here but why would you buy a more expensive, higher than usable frequency ram and then downclock it?Just looking for an answer to a serious question from a guy trying to learn about this ram thing which seems to be discussed almost as much as CPUs.Regards,
Can't speak for Corey but mine was recommended by the sales guy, probably more my fault than his because I thought I had read somewhere that it took 2000 memory.Rick S.

Share this post


Link to post
Share on other sites
Can't speak for Corey but mine was recommended by the sales guy, probably more my fault than his because I thought I had read somewhere that it took 2000 memory.Rick S.
Hi Rick. I wasn't really questioning your purchase but rather the fact that i have noticed that some people seem to do this intentionally and there must be a logical reason for doing this. So i being uneducated to this point on this subject, would like to find out what the logic is.Regards,

Share this post


Link to post
Share on other sites

No real technical reason, Rick H. It's usually dictated by price or what's available where you're willing to shop. Alot has happened in the last six months. New sets of memory have become available and prices have dropped like a rock. I picked this set up on sale in January for $160, whereas normal price was $199. Now it sells for $129 on newegg!!! Other 2x4GB sets of 2133 memory were $200 to $220 at the time, so I figured I would get this and try to clock it up to 2133.The reason for clocking it down would be because clocking it up was unsuccessful or perhaps the need/want for tighter timings.

Share this post


Link to post
Share on other sites
No real technical reason, Rick H. It's usually dictated by price or what's available where you're willing to shop. Alot has happened in the last six months. New sets of memory have become available and prices have dropped like a rock. I picked this set up on sale in January for $160, whereas normal price was $199. Now it sells for $129 on newegg!!! Other 2x4GB sets of 2133 memory were $200 to $220 at the time, so I figured I would get this and try to clock it up to 2133.The reason for clocking it down would be because clocking it up was unsuccessful or perhaps the need/want for tighter timings.
Corey, like Rick H, I don't understand most of this either. Would FSX get much benifit from tighter timing etc or is mostly just a let's see what I can get?.ThanksRick S.

Share this post


Link to post
Share on other sites

I don't believe anybody knows for sure which is better. Dazz did a bunch of testing that showed higher bandwidth (2133 CL9 for example) provides slightly better FPS. However we all know that stutters are what really matter. Stephen has said that tight timings reduced stutters for him. It's definitely a try both and see which you prefer. I sure would love to nail this, but it would require some way of being able to record/benchmark for stutters. As it is, even FPS benchmarking using FRAPS has been brought under question by Nick, so I'm not sure of how best to benchmark FSX other than try it for yourself and see what feels best.

Share this post


Link to post
Share on other sites