Printable View
FWIW the 2500k I bought ~3 weeks ago is C-batch, L041C119. Takes 1.54V Vcore to hit 5.2GHz (not stable).
Someone earlier posted 4.5GHz with 1.2V, this one seems nearer 4.1GHz at that Vcore. :(
At 1.2V seems okay for Linx @4.2GHz
http://i1181.photobucket.com/albums/...th_42_Linp.png
But Prime with same settings above using less power and lower temp fails @4.1GHz.
http://i1181.photobucket.com/albums/...1_Prime264.png
Bist du Ace Kia aus dem Luxx???:up:Quote:
Originally Posted by Ace.
yo :up:
Yep more proof that batch nr's say nothing ATM.Quote:
Originally Posted by Dumo
So what does? Wouldnt it be good to know....
:)
Yes I was aware that v26.x was more stressful than the old one. I hadn't gotten around to downloading it yet on that particular setup. I'm testing 2600K but I'll put the 2500K back in and test 48x with 26.5 as a possible (lower voltage) daily o.c.Quote:
Originally Posted by Ace.
not trueQuote:
Originally Posted by PolRoger
it just finds small tweak errors here and there
very good for 2nd set subtimings
I thought they said last 4 of the Serial mean more than batch?
Just the combination of Batch and Serial is saying a bit...
What's the key with Serial then? I see low serial give low max multi/bclk and higher serial giving low bclk/multi too within any given batch, so either way still seems like a crapshoot?
And then trying to find best batch and serial, no way without knowing what's what :cussing: And that's if there really is any what's what to it all :shrug: :hitself:
I wish Dr.Who would at least be kind enough to give us a few hints, couldn't hurt him or Intel any I wouldn't think. Anyone chat with him via PM, ask him if he'll stop by and give us some pointers :D
Many of the good chips i have seen so far have a s/n with last 4 digits in ~ 1000-1200 range. Not saying this means anything, i just noticed it.
LOL, I knew this would get tricky, but this is getting really tricky now. Anybody would volunteer to make a "Sandy Bridge s/n list" ? :rofl:
Just picked up a 2600K batch "L044B441". First time I've seen a L044. Hope it's not a dud.
Some German stores replied but none will post overseas, do you guys know any that will? Im trying to find a C batch since we dont seem to have them here, or not yet anyway.Quote:
Originally Posted by aussie-revhead
:up:
I agree, any two chips can be very different in peak and manners within the same batch.
I've made note of serial numbers when looking at different batch(s) and even for chips within the same batch it still seems to me that the o.c. results are quite random.
There has always been variation within each batch#s, but I don't think s/n plays a big role, or if it does, I don't think we can crack the code without having a big amount of data.Quote:
Originally Posted by PolRoger
Absolutely, it is really not a great indicator as a batch produced from any area within the wafer can display drastically different ability. I have had same batch chips that acted very different at high speeds not to mention major load temp differences, hot and cold chips and non consistent voltage requirements at a set speed. The search for the golden chip continues.
4.8 @ 1.34 volts only made 14 passes tho then blue screened more vcore for sure.
http://www.overclock.net/attachments...ub-4.89ghz.jpg
Flesh,
I wonder if "hot and cold" chips have to do with the quality/solder of the IHS? (But no one has been brave enough to try removing one ever since the failures with core i7's...)
Whats up Quad. More vcore no doubt. How do you like the P8P67 and what ram are you running?
What batch Quad?
:)
[QUOTE=Falkentyne;4739058]Flesh,
I wonder if "hot and cold" chips have to do with the quality/solder of the IHS? (But no one has been brave enough to try removing one ever since the failures with core i7's...)[/Q Good point and I am sure there is some truth to that, I remember those daz.:D There are so many potential variables in play here with there current design that can effect surface radiation of heat plus the board socket and block clamp design of certain hardware just adds to the fire. I want a board socket mount made from perhaps a granite surface plate.:cool:
I've had 2x B and 2x C and C is much better for 24/7 clock needs about .05V less for same clock.
Why are most of these new batches such duds compared to the earlier batches?
This seems like Core 2 Duo and Penryn all over again...