Massman

Tried pretty much everything ... PSC just doesn't play nice. With the UD3H S3 bios it's the same issue: DDR3-2600 will just not boot up ('51'). Only the UD5H S4 seems to be playing nicely if you want high PSC/BBSE clocks. To be honest, though ... I'm not entirely convinced that you need PSC or BBSE to get a competitive score in 32M. At least not on air. With a good Samsung kit, you'll get close to the performance of a good PSC/BBSE kit. Just need about 1300-1350MHz on the memory and finetune the subs.

Not entirely sure who 'we' is, but just wanted to say they're pretty much fucked if all they have is me

Thanks for the tip, bro! Now I can't boot with DDR3-2600 divider! :celebration: Quickly tested with the BBSE and 8-11-8 seemed fine. But I'd prefer pushing the MHz a little higher over tightning up the subs. FYI - the flare kit did 1333 on the Gene V a while back:

Playing with other stuff first, Dino. Maybe I will beat you the night before I fly out for Computex and do a massive "in your face"-move when we meet up there! :celebration: Fyi, started playing with the UD5H again ... can't believe I would punish myself like that. Anyway, for some reason the board started playing nice with this Flare kit. It's failing to boot with the 2666 divider, though. Always stuck at '51', no matter what timings (auto or C9+) or voltage.

Tried using 2 cores again? With the Giga boards I'm always booting up at 39x when using 1 core (after crash). Just using software to go up in OS works fine.

LOL - thanks for the tip, man! That was the solution for my instability all along! (ps: i'm being sarcastic - no thanks were actually given at all (maybe only fake ones); I'm not a retard, of course I run Waza at slower frequency)

Ever got it back to normal, Splave?

Time to buy this chip and flip the IHS.

Yes. Tried Waza and PI crashed after 1 retarded loop.

Looking beastly. Nice!

With the MSI I tried reflashing all BIOSes as well as the Intel ME firmware and it never recovered from 16x. Thanks for the input Carpo5. So we got 5 x16 boards already ... seems like a systematic error to me!

Improved a little.

Alright, seeing more and more people report having boards stuck at 16x CPU ratio. I, myself, have had two boards with this issue: - Z77A-GD65: still stuck at x16 - Z77X-UD3H: after night of drying out back to functioning normally I've heard reports of people having this issue with the Gene V too, so it doesn't seem to be a specific brand issue. What happens? Well, essentially, your board gets stuck in 16x CPU ratio and will not allow any other multipliers to be set. With the GD65 I'm stuck at 16x whenever I'm increasing the CPU Vcore, but the board works perfectly normal without overclocking. With the UD3H I got totally stuck at 16x, but after a night of drying out, the board went back to normal life and is doing all multipliers again. In both my cases, the lock appeared following a hard crash at 6.7G+ PI-32M. With the UD3H, I could actually notice the CPU going to 16x during the PI loops as up until loop3 all times were normal, but loop4 was incredibly much slower. So, why does it happen? No idea. Is there a fix? Yes, there is a temporary fix for this problem Just confirmed the solution on my board here at home. Using: - 3770K - Z77A-GD65 - Bios V10.5b3 - ThrottleStop 5.00 b1 (http://www.softpedia.com/get/System/Benchmarks/ThrottleStop.shtml) Board boots in at 16x, enabling ThrottleStop makes it jump up to 45x (as set in BIOS).

If I enable two cores, I lose about 100MHz. I'll settle for a loss of 0.9s efficiency then, thank you! It's actually the board. I've an MSI one that does exactly the same (because of the same CPU and the same type of crash). Bios says 1600MHz CPU. I'll do the BIOS switch thing later today, but I'm not too confident it'll work. As I replied to ME4ME, I've an MSI board that does exactly the same. I re-flashes all BIOSes on that board as well as reflash the ME firmware. Nothing fixes the issue.

Waza would give me about 1.5-2 seconds; subtimings are not maxed out yet either. S3 is just for the 3D01 boost and >1.9V vcore.

Well done, nice efficiency! I need to try harder!

Small update - been playing with the memory subtimings a bit. I'm now at 6.22.7xx at 5gig, so still need to find 2 addition boost seconds. The board is stuck at 16x CPU Ratio now ... daaaaaaaamn!

+1 for turning chaos into order

In response to some questions regarding the upcoming 680 lightning and the APAC qualifier. Regardless of the release date of the 680 Lightning, whether it's before or after APAC qualifier kick-off, it will not be allowed to use it during the qualifier. This to make sure the qualifier event is not just function of Lightning availability and/or sampling .

Can't even believe people are thinking about doing that . Is that really how you guys want MOA to go down?

I think with IHS is actually the best way because: - the mounting mechanism presses on the IHS to make sure the CPU is perfectly mounted - without IHS, you will need a perfect mount to avoid cracking and optimal contact Imho, the way der8auer did it is the best approach.

Using Kelvin makes the entire thing more scientific.

Anyone of you overclockers up for a trip to the beach on the 9th or 10th? Kingpin is a self-proclaimed master of the "route to the beach" and knows a "good spot", so I'm pretty confident it'll be very good . Just drop a post if you're in. So far: - Dinos22 - Kingpin - Massman - Redmax - Saaya - Smoke - Youngpro

Is this a serious question?

I use three pointers: 1) 5G+ on air 32M (really bad air cooler) 2) 5.5G wprime at -30°C 3) 6G wprime at -90°C If the chip meets these criteria, it's a good one (imho). So far, I only found 2 chips that could do that.