Moving To The Dark Side Of Cooling - Home Made Dice Pot

Witchdoctor · Sep 1, 2015

Sounds like you are all set

in future a quick test is to pull the pot up and see if it is making contact prior to tightening it down

if not you have something holding you off the die you can adjust now
You should not have to rely on tightening to make contact with the die
tightening is simply to spread TIM evenly and eliminate the pot from moving

MaddMutt · Sep 1, 2015

I still don't think I had it with full contact/pressure as I was only able to get the CPU to -25c

I was afraid to push it to hard and blow a VRM, I hit 4.8 @ 1.525v and 5 @ 1.6v. I could hit 5.2 @ 1.73v but it was unstable to run any benches

I know I could of had 5.5+ with 2.0v but < also comes with being able to take the CPU to a lower temp

EarthDog · Sep 1, 2015

So... you didn't check the mount before you benched it?

You are using subambient cooling and are afraid to push it hard? That is like buying a lambo, taking it to the track, and driving on the track at freeway speeds!

Madmutt... come on man... :-/

. Make sure your money, time, and effort are worth it. Check your mount first as we said before you started. And if you are afriad to kill something, you probably shouldn't be benching under DI/LN2. Go get em man!

Witchdoctor · Sep 1, 2015

2.v istoo much for dice , probably not great even at ln2 temps

as I said above it should contact the die easily with no pressure

Rinse and repeat , 1.7 to 1.75 ish would be just about all I would do @ -80

you have to get the mount straightened out or it will be a moot point

MaddMutt · Sep 2, 2015

I know you GUYS are wanting pic's of my FIRST DICE run. I have to resize them from the camera before I can up load them

I could not get the pot to release the CPU during my tear down, I thought that it was still frozen to the CPU. I took it off about 1hr ago and the pot was properly seated on the CPU, I pulled the CPU out of the socket getting the pot unmounted

I know that I had to have been below -25c ( Cheap a$$ K probe ) as I had the IPA into a jelly for the last 1hr of my benching

I was able to hit -25c ( From the Cheap a$$ K probe ) { @ ~ 9:30am, from then until ~ 11:50am } but close to the end I could not keep the cold??????
I don't know if I was wanting to much from the CPU, Ram, or Board but I could not run the speed I was hoping to hit

I was hoping to hit :
CPU @ 5.0+
Ram @ DDR-1600 7-8-7-15 _ 24 1T ( Crucial Ballistix Elite @ DDR-2133 9-10-9-24 _ 33 1T )
CPU/NB @ 3200 1.375v < was still listed as being in the safe range.
I would set the board up through AISuit to what I wanted to run, then try to run the bench and would 90% of the time get a BSOD?????

EarthDog · Sep 2, 2015

Perhaps it is the probe, perhaps it is not mounted in the right spot... Anyway, if there isn't a CB/CBB, then you don't really need to know the temps honestly.

Just makes sure the DI is crushed so there is more surface area. Some can get it in pellets, but if you get it in block form, take a hammer to it and pulverize it to mostly powder with a few chunks.

MaddMutt · Sep 2, 2015

Here is a shot of my SuperPI run while I still had the cold

I had to save my CPU-Z as a .cvf, I'm trying to get a validation for getting to 5.1GHz w/ 1100t

MaddMutt · Sep 2, 2015

Here is a pic of my pot at work

I needed a second person just to keep the pot full of DICE :rofl:

MaddMutt · Sep 2, 2015

I just got conformation from AMD that my FX-8350 and FX-8320 where both BAD. Thanks to Dolk for letting me know about the 8350 or I might have taken out more boards

For some reason, benching with DICE was not as fun as I was hoping it to be

It might have been that I was expecting more from the 1100t that what I got

Or because I had not do my homework on the Memory and CPU/NB settings that I got to many BSODs and could not go from one Bench run to another

Mandrake4565 · Sep 2, 2015

Maddmutt, having expectations is tough because you never know what the chip will do. Sometimes you get a real good one and sometimes its a dud. I will always try and learn a chip to see what it likes benching before I'll take it cold. It makes for a lot less frustration. Otherwise, it's your first time and you didnt kill anything. :thup:

EarthDog · Sep 2, 2015

Mandrake4565 said:
Maddmutt, having expectations is tough because you never know what the chip will do. Sometimes you get a real good one and sometimes its a dud. I will always try and learn a chip to see what it likes benching before I'll take it cold. It makes for a lot less frustration. Otherwise, it's your first time and you didnt kill anything.

This.

To add to that, and with utmost respect, you are attempting some pretty advanced things as far as overclocking goes but don't really have a grasp of the basics. Sub-ambient adds a whole different can of worms to your existing can of worms and that can only mean a mess. Your drive to do it, I am totally impressed with, but I just think the cart is before the horse in this case.

MaddMutt · Sep 3, 2015

I made up a little chart to help me out while benching

If you want I can send you the .xml file ( 28kb )

Mandrake4565 · Sep 3, 2015

That's great MaddMutt, I do have a suggestion, break it down for a single bench. If you really want to learn what your Cpu likes on a particular bench take one bench and then run it over and over again. See how far the chip will go on what volts and then do the same for the the other changeable parameters. So on AMD that will include the Memory, Nb Freq and Ht link. Every bench is different and will respond differently to the different inputs. Additionally if you go 3d which will be fruitless for points on AMD before you go cold, learn what drivers are best for the bench and well as LOD settings and such.

MaddMutt · Sep 4, 2015

Do you reuse the same IPA from bench to bench or after x days switch it out for new IPA?????? I still have half a F1 container of IPA from my last bench runs, it should have evaporated by now????? I will go slower the next time and not think I'm KING and can get close to WR run. The best we have right now is 4500 on water, so I'll try and shoot for 4600. I just got a e-mail yesterday ( I was down sick all day ) from ASUS approving my RMA for my CHVFZ and Sabertooth :thup:

Just a rehash for FYI, AMD has approved my RMA for my FX-8350 & FX-8320 :thup:

When I get them back, they are going so cold

My memory is still fresh from when I used them in the RR to win

I'm still not ( memory ) up on my PII 955 & PII 1100t as I should be

I'm currently short boards to test them on before moving them to the cold board

IIRC the PII 955 just cares about power and not balancing the Memory, CPU/NB, NB, SB, Ect,Ect.

I Thank You All For Your Help :thup:

Janus67 · Sep 4, 2015

I always dump the pot after a session

Also, don't know if you were doing it, but use some chopsticks or something to help stir the pot while the dice+acetone/alcohol are in there.

MaddMutt · Sep 4, 2015

Janus67 said:
I always dump the pot after a session

Also, don't know if you were doing it, but use some chopsticks or something to help stir the pot while the dice+acetone/alcohol are in there.

Thanks for the FYI on the IPA

stir the pot... :facepalm:

That was not mentioned in CLASS - GOING COLD 101. The teacher must have forgotten that crucial piece of information :rofl:

I've got IIRC like 10 different teachers, HUM...which one forgot that part????

Blaylock · Sep 4, 2015

I use acetone which remains a liquid well below -80C. No need for stirring, just keep feeding the hungry hungry pot.

Mandrake4565 · Sep 4, 2015

Janus67 said:
I always dump the pot after a session

Also, don't know if you were doing it, but use some chopsticks or something to help stir the pot while the dice+acetone/alcohol are in there.

X2

MaddMutt said:
Just a rehash for FYI, AMD has approved my RMA for my FX-8350 & FX-8320 When I get them back, they are going so cold My memory is still fresh from when I used them in the RR to win
I'm still not ( memory ) up on my PII 955 & PII 1100t as I should be I'm currently short boards to test them on before moving them to the cold board IIRC the PII 955 just cares about power and not balancing the Memory, CPU/NB, NB, SB, Ect,Ect.

I Thank You All For Your Help

The AMD chips will respond positively to Memory, Nb Freq and in some benches Ht Link spd. Cpu speed is always king but if you're looking for the last few point/seconds they will be crucial.

MaddMutt · Sep 4, 2015

Witchdoctor said:
2.v istoo much for dice , probably not great even at ln2 temps

as I said above it should contact the die easily with no pressure

Rinse and repeat , 1.7 to 1.75 ish would be just about all I would do @ -80

you have to get the mount straightened out or it will be a moot point

Right now I'm running Super PI 32m @ 4615.4 with 1.6v. I can get the short runs @ 4815.1 with 1.750v but it will crash if I try the long runs

I tried out my Crucial memory from my last run. CPU-Z said that I had 4GB installed but in single channel

I tested each one and one was BAD

When the computer booted it said I had 2GB installed. Looks like another RMA for me

Using my backup set of G.Skill Ripjaws X DDR3-2133 @ 9-11-10-28_38 2T, I have them set @ 8-8-8-20_38 2T.
Got to go back to my benching computer now :thup:

Later

Dr. McCoy · Sep 4, 2015

MaddMutt said:
Right now I'm running Super PI 32m @ 4615.4 with 1.6v. I can get the short runs @ 4815.1 with 1.750v but it will crash if I try the long runs

I tried out my Crucial memory from my last run. CPU-Z said that I had 4GB installed but in single channel I tested each one and one was BAD When the computer booted it said I had 2GB installed. Looks like another RMA for me Using my backup set of G.Skill Ripjaws X DDR3-2133 @ 9-11-10-28_38 2T, I have them set @ 8-8-8-20_38 2T.
Got to go back to my benching computer now
Later

The 1100T you are running is about on target for what to expect from a Thuban on DICE.
My 1100T does close to that itself so you're in the ballpark at least, however be careful, close to 1.8v's is flirting with disaster even on DICE for one of those chips. With mounting problems still a possibility I'd stay away from anything over 1.7v's for now and be sure you have those problems solved before going higher with that. Excessive voltage can also cause booting problems too, sometimes not using so much voltage is the answer if trying to push things really hard.

I mean once you lose the chip, it's gone so don't go there if you can help it.

Note I didn't need that much voltage to get mine up for a decent run so be wary of what's going on with temps while doing a run.

Moving To The Dark Side Of Cooling - Home Made Dice Pot

Member

Member

Gulper Nozzle Co-Owner

Member

Member

Gulper Nozzle Co-Owner

Member

Member

Member

Mr. Clean Senior Member

Gulper Nozzle Co-Owner

Member

Mr. Clean Senior Member

Member

Benching Team Leader

Member

"That Backfired" Senior Member

Mr. Clean Senior Member

Member

Member

Similar threads