Binning celerons for CPU-Z Global pointage
 

Had an interesting thought. While CBB seems to be around -80c on the commando with 2 of the celerons I bought going to try some wacky perhaps stupid shit but the line between inovation and failure is fairly small so what the hell.

Since we all suck at cpu-z (Globally no offense guys ;) patch excluded) I thought I would try some shit out that might make it a bit easier.

Goal: Bin the 5 celerons chips I have sitting here using the least ln2 as possible.

Current Best way: Get a cheap low mass pot and mapp torch.

Idea: Use some dielectric grease mixed with your ceramic to keep the pot from freezing to the ihs (before hondacity yells at me keep in mind we are binning the chips not going for the world record speed) NO hold down, just the pot weight on the chip. When I BSOD I will remove the pot, boot the chip put the pot back on and win?

Why: Going from -180 to -80 and back takes alot of ln2 alot of torching and alot of time.


Results and pics to come, loading OS on commando as we speak. Thoughts and kind words appreciate, GO! :good:

PIC of da grease
http://overclockaholics.com/forums/a...1&d=1276655882


Used just enough to cover the base and a bit of the lip of the f1, spread it with a ruler :laughing:
http://overclockaholics.com/forums/a...1&d=1276655882


Pot was very stable, within 5c of the bios from 20c to -80c before the first pull
http://overclockaholics.com/forums/a...1&d=1276655882


Chilled down to -140c, pot was still easy to twist and easy to pull right up didnt have to force it. Pic is me pulling it up and booting to get around the CBB. Worked great, little ice around the socket because I wasnt using any paper towels or the usual black neoprene insert.
http://overclockaholics.com/forums/a...1&d=1276655882



And a pic after playing around for an hour and a half with pot around -170 the whole time. Dielectric really kept the frost at bay :thumbsup:
http://overclockaholics.com/forums/a...1&d=1276655882



In conclusion. I would use a bit more of the grease next time. After pulling 2 or 3 times temps where poor. -80c in bios with a full pot. Holding the pot down with some pressure for a minute helped by melting a bit of the ice on the edges then freezing back up in place. All and all worked right to plan, let me play with the board around getting used to aux voltages the chip likes. :thumbsup: I will use this method again for sure.


http://valid.canardpc.com/show_oc.php?id=1244534
at -80c with plenty more voltage to go. Might be a winner.

The board on the other hand is very quirky, will have to start a new thread for that.

woot i don't put the hold done for 30c benching...ummm os install rather..hahahah

monitor vid in bios ....

about the paste....i don't use paste :D

Sounds like a sound plan of attack Splave, good luck on finding the golden goose ;)

Nice idea. The -80 CBB is a serious PITA.

^ you said it patch :)

UPDATED WITH RESULTS

I like where is thread is headed and that direction FTW! :)

Nice direction there splave, but since you want to bin the chips, you plan to remove the CPU and put a new one mid-session as well? or did you already do it and I missed it? Though mid-session change of chip doesnt sound like a gud idea since there wud be then air-pockets...

Splaves pot acrobatics! I like it! :D

thats the idea, I wanted to see if I could remove the pot at -180s from the board and I could. When I bin them I wont go over -100c or so. The key is keeping the base in contact with something so it doesnt frost over. :thumbsup: Im not worried about air pockets because there will be air pockets for sure. I can always take the pot down to -180s and have the cpu at -100 or so. This is just for binning not all out for a score. :)

Nice thread I wish I had your biblical Job patience!

man, if its working then its amazing, moreso because even if something goes wrong, I think its an experiment certainly worth trying(ofcourse only for Top 50 guyz I suppose), :good:

getting a little better at the board
http://valid.canardpc.com/show_oc.php?id=1246372

Just need to get 1:2 mem strap to work. I kicked OCP in at 2.1v lol oops

Voltage over 2v on a CPU always makes me smile.

haha for sure, any trick to getting 1:2 to work mr.patch? some prociess just dont like it?
Im using the outer most ram slot. 1gb cellshock

That's quite expensive binning technique :D

Actually cedarmills are scaling quite linear. So there is no reason to waste LN2 for every single cedarmill sample. When I and Sam were binning cedar mills first thing to do was just simple air testing with 1.45V if you can't reach 5Ghz here you better step to next sample :D If cpu reach 5Ghz+ then it's qualified for next stage under DI- we are aiming here for 6.4Ghz at least, not much CPUs will be able to do this but if it can then congratulations you have 7.5Ghz+ candidate under LN2 ;)

cool technique :) I like it. DICE is more expensive than ln2 for me though! I love your 8ghz run man, very nice.

I finally got a decent one

http://hwbot.org/community/submissio...52_7559.95_mhz

347s are the way to go huh?

than it's different story :) here in banana republic it's quite expensive & I have to rent dewar :(

any cedar mill can be good- I know there are some batch theories but I never got into them
But in general from my own experience and results seen:
Costa rica I would say mostly sucks but with very rare exceptions
Malay has some rare samples really good
China has best luck to get good sample ;)

always the chinese

Nice man I want those 32 points!!!

Congrats!

Damn Splave. You and your exploits made me snag another Celly 356. Couldn't pass up a $15 server pull.

awesome! Hope she is a winner.
guy sent me a dead 347 the other day. :(

Would you do this outside of binning chips? The mass of my pot is pretty absurd (3.65kg, otternase makes em big), so cool down seems like it will take a while; i haven't used it yet so don't know but something says 8lbs will take a while. I know it's a big sucker to drop on and off, but if i can keep that pot at a lovely -190 (which should be no problem as long as i keep it full, and i can drip it right from the dewar into the pot) and pull it off for booting, it would make taking the torch to my very expensive pot less needed, even though i know it won't damage the alum or copper, just wastes ln2.

Actually the continual cycle of heat up and cool down does something to copper which makes it more responsive to ln2 over time. I've seen it for sure with my F1's. My first pot which had serious usage gradually became more responsive. When I switched to a brand new F1 I could totally see the difference. Now the new F1 is starting to get responsive after the abuse :)

K|ngp|n has talked about this wierd effect a few times now. :thumbsup:

honestly I would just strap it down, doesnt take that long to do. It was good when I had rom's koolance pot so I could thaw one while I cooled one.

Best bet is to buy a known good chip like you are doing ;)

I've heard those guys talk about that too and I don't buy it.

I call it familiarity. Use something for a while and you get comfortable with its quirks.

I also don't buy that "glazing" your pot hocus pocus. After you've been benching for a bit, you've equilibrated your pot temps with the insulation and hardware around it. Means there is less of a temp sink that absorbs changes and your pot becomes more "responsive". Reaching that point often coincides with having reached a coldbug and needing to torch. Thus the association (in my mind).

But hey, I'm not one to bash established dogma and superstition. I like benching with an anvil nearby. Whatever makes the stars align in your bench session is good. :good:

Hahaha, thx Patch, u crack me up man. :rofl

I can see it, LN2 is cryogenic temps, and cooling and warming metal through such temps changes the crystal alignments.
They do it to competition rifle barrels to line all the crystals up and get that last .1% accuracy. I don't see why it couldn't do similar things for heat transfer.

Touche!


@Kal: Just keepin it fun and real at the same time.:thumbsup:

Its the OCA way, wouldn't expect you to eat chocolate covered turds just cause I said it tasted good ;)

most of the subzero guys believe that there is such a "seasoned pot" ......blah blah blah...i don't buy it also...tried v2-v3's pot and a brand new pot....the brand new pot reacted way more predictable...

anyways...i.always reach for the critical point....don't over fill...its just a waste if you have a cb...torch it ..then you're all good...

Anyone try the spray bottle technique where you squirt the sides and get an ice layer for better temp control?

You're killing me Splave. ;)

Ice doesn't conduct temperature very well. That's why igloos work. The glaze is a barrier more than anything.

I ain't against torches. Torching is GOOD. Aside from the obvious time benefits, by rapidly warming up your pot with a torch after CB you can get back down to benching temps quicker without warming up the surrounding insulation/hardware as much as you would have just letting it sit. This also cuts down on the liquid condensation that develops outside of your pot.

ice is bad... i hate ice...

Thanks for the advice. Well i guess i will plan on torching, i debated using a heat gun, but i think flow of air will be splling the ln2, and torch is just way faster, just gotta make sure not to melt the mount.

I am looking forward to ln2, and have been for almost a year. SS is fun, but i want to get some pouring time in. I got my dewar last october, it's an 80L cryofab so transporting the sucker has been my biggest issue, as the company that will fill it for 125$ requires truck or trailer (neither of which i own). My new car i got in december (subaru outback) gave me the ability to tow, so now i just need to buy a hitch and ball and rent a trailer for a day.
Hope to be joining the -100 club asap, torch in hand :Hi:

And yeah bill and shannon, prepare for my epic post lengths when i start needing ln2 help, since you know this is a short one ;)

Its not my idea man lol Ive never tryed it, I think gomeler said he does it :Dizzy:

Good luck with the LN2 stuff! :thumbsup: