DDR3 1600MHz on Z77M-D3H/Sandy Bridge?

[Rivergoat]

Wondering the group's thoughts....

Other forums say Sandy Bridge can handle 1600MHz no problem; the manual for the Z77M-D3H states that in order to use 1600MHz you must have an Ivy Bridge CPU.

I had been running Corsair 1600MHz with unstable results. Without XMP enabled it was running at 1333MHz 1.5v relatively OK, but sometimes a random reboot/freeze occurred. Yep, I ran MemTest86 on the chips with no problems found (oh, 2x4gb I should say). Anytime I clocked the system up to take advantage of 1600MHz (even trying some timing tweaks) the system would most certainly freeze up. Maybe in 5 minutes, maybe 5 hours, but it would be a certainty.

I really couldn't play the RMA game with Corsair and cripple the machine; I had been just running it at 1333 until today.

I picked up some "Vengeance" series DDR3 2x4gb. Those are advertised by Corsair as having been designed for Sandy Bridge, and only list 1.5v as their rated voltage. OK, so those are in now. Rather early to tell, but everything is running fine.

So....a crap shoot to try 1600MHz in this board on a Sandy Bridge (i5 2500), or just unstable chips?

PS.................I had asked in another thread about Gigabyte listing the current BIOS for this board as F18 (a jump from the beta F14d last listed). OK, tried to use it...got an error about compatibility or some such. This mobo is rev 1.0, not 1.1. Not going to bother, especially testing the Vengeance chips instead. Running F13 OK for the moment.

[Rivergoat]

Addendum....

I somehow missed the miniscule notation in the user's manual. Seems that it's the H77M-D3H, no the Z77 that requires an Ivy Bridge to support 1600MHz.

Just the same, everything else I posted about instability with Corsair XMS3 stands.

[Del]

hard to understand why Gigabyte cannot keep up with RAM standards

[Rivergoat]

Maybe that's true. Still one issue for me may *I repeat MAY* be that the 1600MHz chips I had been running are unstable. The board certainly will not run them clocked at 1600, that much I pointed out.

I'm not sure how long running MemTest86 is required. Probably an overnight repeated test would prove/disprove something. But even if it passed, can the chips be suspect? Other chat boards seem to think that the MemTest would catch a problem, it is the best test to use. Others may disagree. When I was doing desktop support in my last job I could run MemTest, and 99% of the time it would show massive errors right away on test #1. On a few instances it would show errors once it got to test #6.

I know that when I did run MemTest on the 1600 chips it completed all tests with no errors. I don't know everything about RAM as far as if an unstable product would pass tests, but fail intermittently in use....?

I'd also like to see better quality control all around. I started this new box with ASUS mobos, but ran into bad board after bad board (memory slots bad, on-board audio failed, that sort of thing). I sold the 6th returned RMA ASUS board for $20 to get it out of here, having bought the Gigabyte (some friends had recommended them). I just had to swap this board due to a flaky PCIe x16 slot (I have another thread on this subject). The RMA board arrived last Monday, and so far has been stable. The Corsair Vengeance 1333 DDR3 has been running nonstop now for approx 8 hours without a hiccup. Definitive test? I hardly think so; I'd say if it runs flawlessly for 4-6 weeks I'd be more confident.

I'd rather get onto other things than constantly ripping open the box and re-working it (oh, and OT...the rear fan on the new case, an Aero Cool box, failed within 2 months....began making a ton of noise. Not a huge deal, I replaced it with a higher quality and quieter fan, but still...I may be naive, but I like to think companies can actually sell a quality product).