2 sets of F3-14900CL8D-8GBXM degraded within a month on Asus P8Z77-v pro

[Brando]

Specs

Intel i5 3570k at stock speed
Asus P8Z77-v pro with 1206 bios
Gskill F3-14900CL8D-8GBXM
Corsair AX850 Power supply
Asus gtx 670 with dcuII cooler
Plextor M3 pro 256gb ssd

Hello all. I've been over this with one of your techs on the phone but thought I should get some more opinions before I spend more money on shipping and you spend more money on replacing my ram. The first set I bought from newegg was never stable at it's rated speed of 8-9-9-24 1866mhz @1.5v per xmp profile as far as I could tell. I got a replacement from the egg and ran memtest86 the next day all day while I was at work for a total of 13 hours with no errors. I also ran prime 95 in blend mode for a couple hours for good measure, also without errors. Not long after I got crashes in some games and weird behavior in win 7. I warrantied or replaced my psu, motherboard, and boot ssd. I thought I had figured it out as being a sound card driver issue when switching back to onboard audio let me game for about 5 hours without crashing. Unfortunately this was a fluke and I had a freeze or 2 soon after which led me to double check my ram after a round of banging my head on the wall. I got errors in prime 95 blend almost immediately at the same settings that had passed testing with flying colors immediately after I bought the ram. However, it seems to be stable at default safe settings of 11-11-11-28 (I think). Am I really that unlucky or is there some kind of incompatibility I'm unaware of? I looked up this model of memory on the gskill site and it says it works with my motherboard. Is this just going to keep happening if I keep sending the set back for warranty? Please help. For the record yes I use a grounding strap and touch my case too.

[Tradesman]

Have you tried BIOS 1406 (just released on the 1st I believe), it specifically addresses the newer 35/3770K CPUs, overall stability and a number of sets of DRAM

edit, yep the 1st

http://www.asus.com/Motherboards/Int..._PRO/#download

[Brando]

Have you tried BIOS 1406 (just released on the 1st I believe), it specifically addresses the newer 35/3770K CPUs, overall stability and a number of sets of DRAM

edit, yep the 1st

http://www.asus.com/Motherboards/Int..._PRO/#download

I haven't been able to upgrade to it because the file was bad and kept saying "not an efi bios". Looks like the one in your link is a different size so maybe they fixed it just today. I'll try it out now. Thanks. I didn't see anything in the description about memory compatibility but here goes nothing.

[Tradesman]

Basically every BIO update is going to contain numerous DRAM updates, kust part of doing BIOS...new sticks are always coming out and since the advertising for the mobos is in part based on chipset capabilities then they play catchup, when the P67s were released they were advertised as supporting 32GB of DRAm and I don't think there were even any 32GB sets available, and extremely few 8GB sticks

[Brando]

Thank you for your quick reply and the link. I didn't realized asus had fixed the file. It worked on the first attempt to flash unlike the one they had on the site a few hours ago. Unfortunately one of my prime95 threads failed just as I was posting this. I'll try xmp profile #2 and see what happens. None of the settings looks any different between profile #1 and #2 but maybe it changes some important sub settings or something. Here goes nothing.

[Tradesman]

Let us know, if any problem try raising DRAM voltage .05

[Brando]

Let us know, if any problem try raising DRAM voltage .05

XMP profile #2 is almost 10 minutes into prime95 blend so far so good. What is the purpose of having more than one XMP profile by the way? Also, isn't going over 1.5v bad for the cpu?

[Brando]

Well, it's been 20 minutes of prime 95 with no error which is 19 minutes longer than yesterday with xmp #1. XMP #2 actually tightened some sub timings which should have made it LESS stable if anything. I don't get it at all. I also have no idea why XMP #1 worked for 13 hours with no errors a few weeks ago but not now. This is just plain weird. BUT if it works now then who cares why. I'll start memtest when I go to bed to make sure. Thanks!

[Tradesman]

Not all mobos manage voltage all that well, so to help stabilize, one often uses slight voltages adjustments, especially on Intel with sticks 1866 and up. These are actually safe to 1.60 and even slightly higher. If you take a look at even faster sets you'll see the voltages often rise accordingly and may require a raise of the MC (memory controller) voltage also. As an example my Z77 runs at 4.6 and I have a 32GB set of Ripjaws X, spec is 2133/CL9 1.6, have been running them at 2400/CL10 1.655 for some time now. The specs laid out for the CPUs are at stocks specs i.e. 1333/1600 sticks so as you OC the CPU and run faster DRAM the voltages will/may go up accordingly

[Brando]

Intel makes it sound like going above 1.5v is like juggling with dynamite. Good to know.

[Tradesman]

An example of Intel being (shall we say) conservative, when the Q6600 came out, they strongly suggested against OCing it and to leave it at 2.4GHz, that was 2007, it ran on a 1066 FSB, I had one of the originals, and as mobos progressed, it went to a 1333, FSB, then 1600 and even higher, that 2.4 CPU ran 24/7 at 2.6 to 4 GHz up till I just sold it about 3 months ago. I've got other Q6600 out there I build for clients running on average about 3.8, and a few running over 4 on water. Big thing is to monitor your temps, heat can be a killer.

[Brando]

An example of Intel being (shall we say) conservative, when the Q6600 came out, they strongly suggested against OCing it and to leave it at 2.4GHz, that was 2007, it ran on a 1066 FSB, I had one of the originals, and as mobos progressed, it went to a 1333, FSB, then 1600 and even higher, that 2.4 CPU ran 24/7 at 2.6 to 4 GHz up till I just sold it about 3 months ago. I've got other Q6600 out there I build for clients running on average about 3.8, and a few running over 4 on water. Big thing is to monitor your temps, heat can be a killer.

I've been on the paranoid side since this is my first build made of all sub 30nm process parts. Nobody really knows how touchy they are yet. Especially the 22nm ivy bridge stuff. Idk if the tri gate tech helps or hinders oc reliability.

[Tradesman]

On the tri-gate only time will tell, think they started talking about it something like ten years ago, so they've been working with it for a long time, hopefully it's pretty well stable and solid. Haven't really seen or read about any real meltdowns w/ IB and their are plenty of folks puching them much further than I recommend here on the forums, I push beyond what I recommend, but then if I fry anything it's on me, don't want to put it on others.

[Brando]

I'd be happy with a 3570k at 4500mhz with less than 1.3v but it wasn't in the cards. Then again what I thought was an unstable oc could have been my ram settings so maybe it's not out of the question. My approximate max oc was 440mhz with about 1.27-1.28v. I got a lame one.

[Tradesman]

DId you try using an offset voltage for the CPU? Like + 0.030 to 0.50, at 4.6 I'm using 0.085 right now (still playing), C states off