• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Which ram to get?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
yeah, fairly basic question here, but trying to get some 'P67' ram so I can push 2133 a little easier than with my 1.65v Hyper's.

Should I get these Ripjaw X at CL7 and 1.65v?

Or these Dominator GT's at 1.5v and CL9?

Or any other..

I currently have 3x2GB Super Talent Chrome's that I will be getting rid of and wanted somethign with the lower voltage b/c I so far believe the 1.5v hype or the more stable those are made for P67.

Thanks men! :thup:
 
The ripjaws for sure, one thing i've seen with SB is that the IMC is not that forgiving when you try to force lower timings even with higher voltages, also sticks buck when you turn up the juice, especially the 1.5v ones, I have two sets and one is 1.5 CL6 and 1.65 CL7 and the low voltage pair don't like voltages over 1.6 at all
 
I'm torn on what to suggest. The Ripjaws kit you have listed is very solid, Brolloks has a set and they do great. The other one I'd consider is this one, which is also 2133, but with 8-9-8-24 timings. I'm not sure if CL + tRP or tRCD + tRAS will help more. G.Skill seems to think the 8-9-8-24 kit is harder to bin; they're $20 more than the 7-10-7-28 kit.

Wish I had both to test. Not sure what to tell you, sorry!
 
I ran some SPI32 M tests and 2133 7-10-7-25 outperforms 1900 6-7-7-18 by a good measure, I did these with Ripjaws vs GT's respectively
 
Yes, but the MHz scaling on SB is strong too. I have a kit that'll do 6-9-8-24 installed right now. I'll try to run a couple 32M at the two different settings. ED, if you can be patient, I'll try to get that done by tonight.
 
I guess my concern stems from another thread where that IntelEnthusiast member stated to have 1.5v... I PM'd him to explain why he/intel suggests that and he said he was researching it (though I wish there was knowledge behind the suggestion BEFORE it was posted).

So 1.5v ram doesnt scale well with voltage...teh 1.65v ram may/may not be good for stability for the IMC...

Im still a bit confused. Should I just keep my darn ST Chromes? They do 2133 @ 9-11-9-28 w/e HOkie's were set to for the review (never tried tighten them down), but...
 
Well then, crank 'em down and see what you get!

1.5V is JEDEC spec for the SB chips. I haven't (yet) had IMC stability issues and have run 1.65V for the life of this chip. FWIW.
 
2133 is about the highest your gonna be able to run anything on SB i belive anyways. Can't get any ram to post at much higher than 2300mhz on my Maximus IV Extreme. I am using dominator GT 2400's. At 9-11-9-27 @1.65v.
 
Well then, crank 'em down and see what you get!

1.5V is JEDEC spec for the SB chips. I haven't (yet) had IMC stability issues and have run 1.65V for the life of this chip. FWIW.
They are 77721 at 1.65v 2k speeds...on Bloomy chips so I assume that should be good here at 2k.. lets see what happens...
 
I guess my concern stems from another thread where that IntelEnthusiast member stated to have 1.5v... I PM'd him to explain why he/intel suggests that and he said he was researching it (though I wish there was knowledge behind the suggestion BEFORE it was posted).

So 1.5v ram doesnt scale well with voltage...teh 1.65v ram may/may not be good for stability for the IMC...

Im still a bit confused. Should I just keep my darn ST Chromes? They do 2133 @ 9-11-9-28 w/e HOkie's were set to for the review (never tried tighten them down), but...

My point exactly, the 1.5 kit is not going to do CL 7 at 2133 even at 1.65v, the 1.65 kit is already there
 
7-10-7-28 vs. 8-9-8-24

Well, it appears we have a winner by a nose.

CPU @ 4.5GHz, RAM @ DDR3-2133. Only the timings differ.

First, 7-10-7-28:

4500_2133_7-10-7-28.jpg

Then, 8-9-8-24:

4500_2133_8-9-8-24.jpg

Thus, 8-9-8-24 wins by a nose, beating the other by a mere 0.748s.

This is NOT scientific, as tests were only run once. Will give it two more runs each to be sure, but this is what I have on the first run.
 
Finished up testing at lunch. I ran each test three times and averaged the results.

Test System
  • CPU: i7 2600K @ 4.5 GHz
  • MB: ASUS P8P67WS Revolution
  • RAM: 2x2G G.Skill Pi (F3-19200CL9D-4GBPIS) run at DDR3-2133
  • OS: Windows 7 x64
  • Benchmark: SuperPi 32M, run at high priority with the 32M runs immediately following a 16K run (my SOP for running any SuperPi test).

Contender #1, based on this G.Skill kit rated at DDR3-2133 / 7-10-7-28 that retails $139.99.
(Whoops...that's a 7-10-7-27 kit, completely invalidating this test, which was run at 7-10-7-28. I'll re-run them hopefully tonight. Sheesh. As it is the least important of the four timings, I'd wager it makes little difference.)

7:31.215
7:31.356
7:30.794
Average: 7:31.122

Contender #2, based on this G.Skill kit rated at DDR3-2133 / 8-9-8-24 that retails at $159.99.

7:30.467
7:29.920
7:30.139
Average: 7:30.175

So, on average, at standard settings of 4.5GHz on the CPU and DDR3-2133 on the RAM, the 8-9-8-24 kit will beat the (non-existent) 7-10-7-28 kit by 0.947 seconds.
 
I'm keeping my comments on this for the lounge. Methinks this thread, or at least recent developments, should be there too.
 
If you want to get super specific you would need to check the subtimings too. All I know is the 7-10-7 kit is far less frustrating than my hyperx kit :).
 
Oh yea, hypers don't get along with SB for some reason. The kit I'm testing with, and the ones linked are all PSC I believe.

Unfortunately we don't know sub-timings of the proposed kits. Those I'm using are all "Auto" on the WS Revolution.
 
I must be the luckiest hyper owner on SB...these problems no doubt are real, but Im trying to figure out why I seem to be ok up to 2133 vs some that cant even boot and other stability issues. Like I said, Im a bit noided out at all these issues, hence the move to something supposedly more appropriate for the platform.

Thanks for the info.. now I just need 2 of my exact same kits sold here before I can sell mine!!!!
 
Oh yea, hypers don't get along with SB for some reason. The kit I'm testing with, and the ones linked are all PSC I believe.

Unfortunately we don't know sub-timings of the proposed kits. Those I'm using are all "Auto" on the WS Revolution.

If those are on a WS revolution it makes me think my MIVE should run them ok, it spat out Bnova's Hypers.

Would be a vast improvement over my 9-11-9 27 ddr3 2400's
 
I want to try the hypers again in your next board, I don't think your bios profile was as cleared as it was said it was.
 
Back