Massman

Who from TW ?

Updated list. Looking good!

Like: S / N XXXXXXXXX142X But it's just for this batch, not a general rule ...

Two words: "Ivy Bridge".

I don't see any statistical indication that cache would be the limiting factor. Or, at least, that lower cache would yield higher OC results in some way. That 3930K 'seems' better is just down to sample test size. And that's just down to pricing: who cares about a $1000 X if you can buy a $550 K that has similar performance and similar overclocking capabilities.

- MSI Big Bang XPower II --- V1.2b1: download This is supposed to be the XPower bios with cold fix. Please report back if you tested it (I will test during lunch time for sure!).

SOF: http://hwbot.org/forum/showthread.php?t=37294

- Asrock X79 Extreme4 --- V1.70: download - MSI Big Bang XPower II --- V1.1b1: download - Intel DX79SI --- 0424: download - Intel DX79TO --- 0424: download - Asrock X79 Extreme9 --- V1.98: download

If it's not a joke, why am I smiling then?

Just tested this beta BIOS and I can now confirm that the X12 lock when subzero cooling is now fixed. Even after shutdown, the board will no longer lock in the x12 ratio when below a certain temperature. So, now I can unsticky this thread! The BIOS for XPower should follow next week! Thumbs up, MSI!!

- MSI X79A-GD65 8D --- V1.5b1: download This bios should fix the cold issues! I'll test this tonight - feel free to test and feedback yourself! //EDIT: confirmed!

I'm pretty happy about the BIOS too. Just plugged in CPU and memory, set 1.25 ratio, set memory freq and all boots just fine. Nice!

By the way, if someone wants some binning advise: At a certain moment, we were hitting a rough patch in our session as we were only hitting 52-54x for 10-15 chips straight. Then I picked two from the bunch based on the serial number and both of them were 57x chips. The first 57x of that batch had 152x as end of the serial and from the pack I selected 143x and 141x (those were 57x). The batch of SB-E doesn't mean that much, but if you know the serial number of a good clocking one, pick one that is close to that. It's not a 100% certainty, but it feels like there's at least a higher chance on hitting a good one. I saw the same with 2600K, by the way, so it's not just these 3930Ks.

Anyone already tested the bios?

Seems like a bug. Added to the bug tracker: http://bugs.hwbot.org/browse/HWBOT-706

S-T-R-O-N-G, strong!!

Yeah. The most exciting about this kind of testing is that we can really see the variance within a specific stepping. Doesn't happen too often, so I'm really happy that I get the chance to do this. It's also why it took 13 minutes to bin them. We didn't just look for the 55x+ ones, but noted down the max mp for each chip. The 50x chips too sooo long to test. 55's and 56's => contact Tones. The 57's, not sure what to do with them yet. If they're all 6G chips, we'll just keep them LOL!

Whoah!

L150B490. We tested 5x L149B708 too, but only 2x 55x //edit: just realised we tested 45x

Emanuel is the person to talk to in Tones, so email manu [at] tones [.] be and check if they can ship to your country.

No decision about the 57x's yet. First, need to see how good they are exactly!

They said they would ship internationally if possible. I'm not sure about the shipping policies, so you'll have to ask them

Chances for a Core i7 3930K capable of 55x or more is 13/44 ~ 30%. Almost 1/3, so that's nice!

Hey all, Yesterday (Friday), Leeghoofd and myself were invited in the Tones.BE office to test a bunch of 3930K chips to check for maximum mulitplier. Using the known binning method for SBE, we tried about 45 different 3930K C2 samples from batches L149B708 (5x) and L150B490 (40x). It took about 2 and a half hours to go through them all, which is about 13 minutes per chip. Pretty fast considering we wanted to find max ratio for all chips and not just the 55x+ ones. Binning method Results As you know, SB-E doesn't support multipliers about 57x. So, the five chips that do 57x+ we'll need to test with better cooling ... soon . By the way, you might remember we did something similar with the 2600K a while back. The average chip did around 53x, with max= 56x and minimum 48x. With 3930K, it seems that the results are much more consistent and slightly higher too. It was a lot easier to find higher clocking chips too. Further info Now, read this message posted by Tones on their Facebook page: https://www.facebook.com/pages/Tonesbe-The-Most-Powerful-Computer-Shop/95869329998 (you should like this page!) Translated: In other words, if you're looking for a chip you can check with Tones and ask. No overclocking result is guaranteed! You can make your decision based on the binning method and the results we got. The only additional cost might be shipping if it's international, though, so no extra $$ for a specific ratio . One per person, though, and please use them for overclocking ... If you're checking with Tones for a chip, please be friendly and polite people! It's pretty cool they're so into overclocking, I'd like to keep it that way .

Intel.