MSI Z68A-GD80(G3) and Corsair's New RAM

HU16E

New member
GENERAL OF THE ARMY
Joined
Jul 9, 2008
Messages
11,111
CorsairPlatinum-2.jpg


Yeah, ok, it all looks great, but how do they work together outside of a test laboratory in a real world home built pc?

That question was answered very quickly, without any special fiddling about required beyond normal setting changes.

When the Corsair kit was installed, the usual 'Memory Has Changed' screen popped up. No big deal, a simple check in the UEFI/BIOS looked like everything defaulted to ensure a successful post and bootup. Yup, that's exactly what took place. Corsair took the right approach IMO to set the profile to default to DDR3 1333. At the next restart, it was a simple matter to set the manufacturer's recommended rated speed of 2133, CR 1, 9 11 10 27, tRFC 171 on Auto DRAM V. Posted and booted right up.

Yeah, ok, how about if 8GB isn't enough RAM for me?

Figured that might be a question asked. Here is the finding. To make it more real, often pc builders buy components at different times. This can be problematic, but that is best discussed on a case by case basis. Back to the answer. Purchased on purpose a kit from a different batch, even the serial numbers were way apart. After they were installed, the usual 'Mem. Change' screen again, but at F10 - Enter, it posted and booted right up! No increase of DRAM V. was required. In fact, with it set to Auto, the 'Read Only' actually shows the voltage to be 1.472v's (by meter, it was just below 1.50v's).

Note that CPU-Z X.M.P. for this RAM shows wrong. In the UEFI/BIOS, Memory-Z, it shows correctly.

MemoryTab2133-1.jpg
SPDTab2133-1.jpg
Mem213316GB.jpg


CorsairTest-1.jpg


Conclusion? I would say MSI mainboards and this RAM are going to work out just fine together. :biggthumbsup:







 
Froggy Gremlin said:
If it works, it should be fine. I'm still in a bit of shock at your results!

I am shocked myself, best kit of RAM i've owned! So I ran IBT on maximum, 20 passes, at tRFC 85!

EDIT: Was able to post at 82, but gave an instant BSOD. 83 failed IBT after test 2. 84 failed after test 15.

I'll go back and do more IBT at 85 and prime 95 blend while I sleep to ensure its stable at 85.

Do you guys know of any good RAM benchmarking programs I can try out tomorrow to test out this RAM? :-P)
 
I am testing a kit like that myself at the moment. They are just awesome.
You can use the 7zip benchmark. shows the actual compression and decompression performance for a typical zip file.
MaxxMem2 is another useful benchmark. http://www.softpedia.com/get/System/Benchmarks/MaxxMEM2.shtml
 
I am shocked myself, best kit of RAM i've owned! So I ran IBT on maximum, 20 passes, at tRFC 85!

:shocked_anim: Simply speechless! Can't wait to do some more experimenting & seeing some performance numbers. Just to recap, 2133, 8GB, CR 1, 9 10 10 21, tRFC 85 @ 1.50v's. I knew MSI's mainboards all liked this stuff, but this is something else! :-))
 
Froggy Gremlin said:
:shocked_anim: Simply speechless! Can't wait to do some more experimenting & seeing some performance numbers. Just to recap, 2133, 8GB, CR 1, 9 10 10 21, tRFC 85 @ 1.50v's. I knew MSI's mainboards all liked this stuff, but this is something else! :-))

Yup, I have the 8GB model. So @2133 1T-9-10-10-21-85 1.492v, IBT went through 40 passes at maximum, and prime 95 went 9 hours on blend, no problems. Seems like these are my numbers! :biggthumbsup: This is on my GD80 system below, btw.

And heres a benchmark result (since I couldnt figure out how to add a picture, heres a link. Lol):
MaxxMEM2_a3054.jpg

 
I too ran these setting with Corsiar vengence 2133 memory. 10 passes of IBT no problems  :think:

I never would have tried to tighten them quite this much, thanks!

Corsair really make some great memory.
 
xmad said:
I too ran these setting with Corsiar vengence 2133 memory. 10 passes of IBT no problems  :think:

I never would have tried to tighten them quite this much, thanks!

Corsair really make some great memory.

Might wanna try a little more than 10 passes with IBT, as I had a failure at pass 15. Maybe like 20-30 passes. But I agree, they do make great memory! :biggthumbsup:
 
To add images to your posts:

https://forum-en.msi.com/faq/article/how-to-post-images-in-the-forum

Photobucket is a good image hosting service, I've been using it for quite a few years.
 
XFM said:
To add images to your posts:

https://forum-en.msi.com/faq/article/how-to-post-images-in-the-forum

Photobucket is a good image hosting service, I've been using it for quite a few years.

I tried that, but it gives me the broken link picture. It may be dropbox (I use it because I got 50GB free when i bought my Samsung GS3).

Ill give photobucket a try right now.

EDIT: It was dropbox, heres the benchmark. Thanks man.
MaxxMEM2_a3054.jpg
 
Further tests are now complete, and must say they are very good. Thanks to 00 oh 7's testing, I doubt I would have tried tighteneing timings as far. With DRAM V. unchanged for the tests, 1.507v's by DMM, results were stable for 16GB @ 2200, CR 1, 9 10 10 21, tRFC 85 with an average 7-Zip of 27,634. However, this proved to be worse than loosening timings just a bit to 9 10 10 22, tRFC 88 for an average 7-Zip of 28,115! This proves to me there is a 'sweet spot' to be found with each individual system configuration. Depends on how much fiddling and tweaking one wants to do to find the 'optimum'. 

This is very good. 7-Zip 27,634;

450022001910102185.jpg


This was even better with the slightly looser settings. 7-Zip 28,115;

450022001910102288.jpg
 
Froggy Gremlin said:
Further tests are now complete, and must say they are very good. Thanks to 00 oh 7's testing, I doubt I would have tried tighteneing timings as far. With DRAM V. unchanged for the tests, 1.507v's by DMM, results were stable for 16GB @ 2200, CR 1, 9 10 10 21, tRFC 85 with an average 7-Zip of 27,634. However, this proved to be worse than loosening timings just a bit to 9 10 10 22, tRFC 88 for an average 7-Zip of 28,115! This proves to me there is a 'sweet spot' to be found with each individual system configuration. Depends on how much fiddling and tweaking one wants to do to find the 'optimum'.

Very interesting. Guess I will have to try to find that "sweet spot" as well! :emot-tip-wink:
 
badboy2k said:
im going to buy some of them for my next build as they look so nice to the eye ;)

GIMME THEM LOL

Always apprehensive, reluctant, and cautious to recommend something, but you should be successful with them as well. Corsair makes a lot of Platinum kits now, so can't vouch for all of them, however, this model that was tested is a winner. 

To make the test results more comprehensive, tested some Corsair CMGTX3 2400's for comparison. Surprisingly, the Platinums beat them in every test. Another excellent example of RAM speed doesn't always equal better. The test modules also require considerable more voltage than the Platinums, stable at 1.648v's (tRFC was 132). Out of 9 runs, the best 7-Zip was 27,950 falling short of the 28,115 average of the Platinums at 2200. Attempts at adjusting the timings other than the manufacturer's recommended settings did result in failure to POST.

CPU-ZCorsair2400.jpg
   

7-Zip45X2400.jpg


 
Yip, must be the best RAM around currently.
Corsair started this leap in RAM with the Vengeance series and now they have perfected it with the Platinum.
The others will have to follow. None other with such timings and SPD at 1,5V.  :biggthumbsup:
 
Just a question out of curiosity, but what makes this kit perform so well over others? The soul factor can't be that they are "hand picked," right? Is it that they are also high density?
 
corsair never cheap out on the chips (to save money) they use unlike some others! they get high grade chips mainly from samsung now that are well tested cherrypicked chips from there batches that have next to no defects meaning they peform much better!
 
bb2k has been researching. :lol_anim: The 2133 Platinums are indeed Samsung IC 256Mx8's. The 2400 CMGTX3's are PSC IC 256Mx8's. 
 
Just some update on the Vengeance kit I got myself. This was after I was given a Platinum kit for testing. The 16GB Vengeance cost was less than the 8GB Platinum. Corsair CML16GX3M4A2133C11B Low profile.
Firstly 1T is a no-no. It boots alright but the performance then is equal to a 1333 kit.
TRFC tightning only really has a marginal influence on the 7ZIP benchmark. 7ZIP is of course highly dependent on the amount of CPU threads, and I will do another run when I have my 3770K.
MaxxMem does show the Latency difference though between 1T and 2T.
The stock TRFC is 171. I was able to get these down to 127 and my 7ZIP benchmark came out better than with the Platinum, taking into consideration that I did not change the TRFC on the Platinum.
I tried TRFC107 but the scores dropped off drastically to around 1333 values.
Overall a bargain kit that does a mighty fine job, at 10.11.11.24 when the design XMP SPD is 11.12.12.28  I'll play around a bit more with timings to see how much further I can push them.

memorytest.png


I ran IBT for 20 runs to test stability.

memorytest1.png
 
Mr. B, would you report the IC's on the label? ver4.13 are Samsung. I'm not sure I will ever learn all the secrets of RAM. What we think should be better doesn't always turn out that way with the opposite being true. Loosen timings and get better performance. Go figure!  :shocked:  For the tinkerer, it is very interesting (sehr interessant) and kind of fun.
 
Froggy Gremlin said:
bb2k has been researching. :lol_anim: The 2133 Platinums are indeed Samsung IC 256Mx8's. The 2400 CMGTX3's are PSC IC 256Mx8's.

yes i have! i want some really badly and have been looking at how there put together (i always do some resarch before i part with my own hard earned cash lol)!

samsung make some of the better made (but expensive) and more standardised RAM chips now and im glad to see corsair are using them,  they make a point of testing the chips fully before there sent to ram manufacturers to maake sure they meet there specs needed to prevent problems!  :agrees:

i like corsair recently as they never seem to let an accountant neer the manufacturing side of there things they make! guess thats why corsair stuff tends to be more expensive then other premium brands (quality costs more money so its worth it)!
 
Back
Top