• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

what cpu to get for best results in a subzero environment

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Warlord2

Member
Joined
Mar 19, 2001
Location
Bakersfield,CA
Im making a whole new computer and it will be cooled using cascade propane setup. Im looking for -50c LOAD temps or maybe better. I want a proccessor that will get me best performance and overclocking ability. I have been looking at the xp 1700 B cpus along with the barton 2500. Price rangs is around $130 max but the less I spend on the cpu the more I will have for a gfx card. I hear I can get maybe 3ghz out of the xp 1700 and about 2.7ghz maybe out of the barton 2500. Im guessing the barton is better, but by how much? because there is about a $70 price difference and I could use the money for a r9700 instead of a r9500 pro. There will be 1 gig of ddr 3500 ram a raid 0 setup of 60gb WD harddrives.
 

zabomb4163

Member
Joined
Apr 27, 2002
1700+ will most likely get close to 3gig with that kind of cooling (assuming you can pull that stuff off)

getting to 3gig is possible with bartons as well, it just takes extreme cooling. some guy on vr-zone got one up to 3gig with a prometeia and several on vr-zone have had Bartons at 3.4 with l2n. so it seems the limit is your cooling.

on a sidenote, could you link me to any of your plans for your cooling system. just sounds practically impossible. It'll be a new record on the forums if it exist though.
 

juliendogg

Member
Joined
Nov 29, 2001
Location
NC, USA
yeah details on the cooling setup please?!!!!
about the chip, i think you might be better to find a t-bred b that is NOT one of the dlt3c's, they don't do well with high voltages at all, despite extrememe cooling, maybe a good stepping 2100xp t-bred b? the dlt3c may get you close 2 3ghz true, but you also may get one that will not post @ voltages above 1.8. that would kinda defeat the purpose of your cooling yeah? if i could swing it i'd love to toss the Barton into a cooling setup like that. :D.

good luck and keep posting!

J.
 

Gautam

Senior Benchmark Addict
Joined
Feb 4, 2003
Location
SF Bay Area
Check here and here for the highest achievers. Either get a week 248 or 251 AIUHB 2100+, or a 3000+ Barton. The latter pretty much guarantees 3 ghz with the kind of cooling you're going to have...if you really can pull it off; I know very little about extreme cooling, but i wouldn't think that setup like that get you temps nearly that low.
 

zabomb4163

Member
Joined
Apr 27, 2002
I'd put my bet with the 2500+ Barton. Gautam he said 130$ max. you recommended a processor several hundred dollars over his max, next time be more considerate.
 
OP
W

Warlord2

Member
Joined
Mar 19, 2001
Location
Bakersfield,CA
I think I will get the barton. I might as well go all out sense Ive come this far. The cooling is going to be a cascade phase-change unit using propane for high stage and low. The boiling point of propane is -43.74F at 0 psi but sense its going to be in a cascade Im HOPING for a boiling point of -85.05F (-65c) at 20" vacuum. Here is a link to the project. At the end of the thread you can find some pics. The problem is, no one that I know of has tryed this type of cooling method so I really dont have a clue what my temps will be at during full load. Also thermal pastes get REALLY crappy at low temps so that will hurt me a lot.
 

hitechjb1

Senior Member
Joined
Feb 2, 2003
I am impressed with your effort, regardless of whether you can go to 3 GHz, or 4 GHz, ... It builds knowledge and know how.

For this project, probably one of the goal is highest possible clock frequency, I think a Tbred B 1700+ DLT3C and 2100+ would have better leverage on clock frequency than the current Barton.

The 1700+ has the advantage of higher transistor speed but potential higher leakage to withstand high Vcore, but low temperature would reduce the leakage concern. It is a tradeoff and "luck" of the pick. The 2100+ is more all around for oc, and better chance to stand higher Vcore, but in general slower transistor characteristic than the 1700+.

Also you may end up having to try a few CPU and pick the best one. So cheaper price CPU would let you to get few more choices. I would get at least one 1700+ and one 2100+ with good stepping. Both would cost about the same price as a 2500+ Barton.

Keep us informed.
 
Last edited:

Gautam

Senior Benchmark Addict
Joined
Feb 4, 2003
Location
SF Bay Area
Whoa, my apologies; I didn't notice that you had a $130 budget. Your system looks absolutely amazing. I'll second a 2500+ Barton in that case. 3 ghz could be very attainable for you.
 

altec

polka dot ninja
Joined
Dec 23, 2002
Location
Doylestown, PA
That propane thing seems a little dangerous to me, but you can try it if you want. Tell us what happens (like if your house survives)
 

Lord_MiL

Member
Joined
Jan 29, 2003
LOL
Propane isn't explosive in it's liquid state, and in a closed, phase change cooling setup like he has, there is very, very little risk of it igniting. However, it's funny to see people's reactions to such an idea.

I'm not sure if you'll quite be able to hit -50 even with the cascade system for reasons that you've already hit on. When you build a cascade system like that, your inefficiencies continue to increase and thermal compound also begins to crapify. I certainly think that something near that temp is possible, but be prepared for a great deal of tooling in order to get perfect contact, increase pressure, etc.
Good luck, I'm sure when it's all done, you'll be very pleased with your system!