San Diego Cold Bug??

Joined
Nov 2, 2004
Messages
825
Ok fellow [H]'ers...Here is my situation and question. I know that with the Winchester cores there was issues with the cold bug and having them work with a good phase change deep freeze..So im reading some reviews on the newer San Diego cores and low and behold I see them getting them to work with phase change units and get some pretty great oc's out of the chips. But.... on the other hand I hear about stories of people trying to oc with them and getting no where. I have an FX57 on the way to go with my DFI SLI ultra board and plan on phase change cooling with temps easilly as low as -60 to -70 below.(for now and colder later) so... with that in mind. what info or experinces do ya all have and if there is a deep freeze issue is it something that could be related to a board issue that can be fixed by a bios. Or is it totally CPU problem. just looking to wreck havock Ocing and eventually go with a dual casscade and have some ocing fun. But makes it hard if this problem exist with a standard phase unit. thanks for any help and suggestions!!!

the Cpt.
 

(cf)Eclipse

Freelance Overclocker
Joined
Feb 18, 2003
Messages
30,027
i don't have personal experience, but it seems that all the ones testest so far can do at least -40c to -50c

i think it's a 90nm problem, memory controller related :(
 
Joined
Nov 2, 2004
Messages
825
Prolly very very possible to achive a real use-able Oc of 3.4 to 3.5.. at least with FX57 anyhow(yes i know FX55 has hit these areas before, But not normally use-able except for some and i do mean very few benches and screenshots of a high oc but not usable.) maybe even higher. one can hope.Hmmm instead of going real cold i guess i will just have to fine tune my phase unit and keep it around the lower end of the happy area of cold.. thanks!!!!!
 

BigDaddy85

Gawd
Joined
May 13, 2005
Messages
876
(cf)Eclipse said:
i don't have personal experience, but it seems that all the ones testest so far can do at least -40c to -50c

i think it's a 90nm problem, memory controller related :(
:( so ... ur tellin us that all 90nm core or just winnie?
 

(cf)Eclipse

Freelance Overclocker
Joined
Feb 18, 2003
Messages
30,027
it's called WINCHESTER! :p

and i'm fairly certain it's a bug that was introduced with 90nm, which the first of happened to be winchester. amd has been supposedly 'working' to eliminate this bug, but i definitly have seen a few revE's with the cold bug
 

BillR

Born Again Cynic
Joined
Feb 17, 2002
Messages
18,540
I have noticed a few bios updates (MSI in particular) that address the cold bug. Not sure if it’s a CPU issue or a bios issue though.
 

(cf)Eclipse

Freelance Overclocker
Joined
Feb 18, 2003
Messages
30,027
well some people have made note of memory overclockability going down as the core gets really cold, which makes me suspect the memory controller failing at the cold temps
 
Top