darn sorry to hear OPB :(
oooooooh sell ya 4000+ SD to me then :)
Printable View
darn sorry to hear OPB :(
oooooooh sell ya 4000+ SD to me then :)
Quote:
Originally Posted by Onepagebook
i am guessing the drive stregnth on the memory controller. Probly an option under the memory part of the bios.Quote:
Originally Posted by bigtoe
I was just looking at the screen shots. Drive stregnth is an option if you use a64 tweaker. its on the bottom right.
If you want it to boot then i'm sure its in the bios too.
Yeah, in DFI BIOS, DRAM DRIVE STRENGTH, about 3/4 way down, try set to 2,4,6 for UTT/BH5 or 3,5 for TCCD.Quote:
Originally Posted by prc8
C
Prc8, thank you for that. But I really don't think that would be help.
cuz I tried 2,4,6,8 already. no luck.
Eva2000, you really need that?:D
replied to your pm OPB :)
also regarding drive strength it ain't set in stone what BH5/UTT vs TCCD uses.. i.e. OCZ PC3200 Gold BH-5 on 310 official bios liked Drive strength and Drive data strength of 2 and 1 respectively for 3500+ CAA2C clawhammer, while on winnie and 414-3 beta bios the same ram flies with 3 and 1.
I'm going to try 4000+ on DI tonight, that'll work better becasue you can "adjust" the cold....
C
If you're willing,
see if you can find a way to disable/terminate the temperature diode on the cpu. I'm not sure if there is any way to do this.
But apparently they changed the diode on the E4 cpus. :S
This really sucks, i hope someone finds a way to figure out this "cold bug" untill then it looks like the cascades may need to run gasses that dont boil at -100. Something like N2O that boils at -88, would probly be benificial in these situations.
Quote:
Originally Posted by charlie
DI definitely will work.
I did check the CPU temp reading under bios. when DT200 read the evap on cascade like -100c, bios temp was about -60
DI might get like -80c and CPU reading under BIOS might be -35~40
thats a pretty big diffrence?! bad contact maybe?Quote:
Originally Posted by Onepagebook
no I don't think it's bad contact.
when I used to have mach 2
it got -48 evap reading and under bios (1.48v) it only have -11 or -13
Does anyone knows what the cold bug really is? I don't think so......
Maybe you can shut down the temp sensor by isolating a pin of the cpu. The temp reading is transported by one or more pins. I know because I was trying to build a Pentium-M adapter a while ago and that's all about the pinout.
You maybe can find the pinout at AMD's website but I don't know how the temp monitoring really works in this case.
For example my 6800 card with a broken temp sensor throtled because is was reading 153 degrees Celsius all the time so maybe disabling the temp sensor will give a weird reading.
OPB if you like it I can search the pinout for u but I don't have time now.......
i think is more an issue of memory controller dont like the xtreme cold is more architecture problem rather than the temp going crazy cuz of low temps
I thought someone also defeated the cold bug by tricking the temp diode somehow via BIOS. Maybe the same thing can work here.
Oh I'm really sorry to hear that OPB, for you and for me... I already sold my FX-55... :(
I'm also for same reason, on-die memory controller + 90nm process.Quote:
Originally Posted by leviathan18
And here's why: under phase-change I cannot clock my Gskill TCCD kits (both 512MB and 1GB) much above 300MHz no matter what I do... Under air I was getting SPi 16M stable runs @ above 320-330MHz 7-3-4-2.5-1T.
This cold-bug has to be really "memory-controller" related somehow.
I wonder, If anyone have an access to some knowledgeable people at AMD and could at least shed some light onto this. I don't think that shooting an email will help here, though :(
interesting but i guess possibleQuote:
Originally Posted by bachus_anonym
Wasn't there a optimistic rumour that a future DFI BIOS rev. would deal with this ? if it's anyway related to the temp diode there should be a way to disable it. Sincerely i don't see the conection with the mem controler or the fabrication process. Some kind soul should also test the E4 rev for socket 754 and see if they behave similarly.
Bummer in my country DryIce is pretty expensive and hard to find (and if you gain acces you have to buy a truck load literaly) but LN2 is happy meal cheap and available but now AMD spoils all the fun :( :(
i think is the process of 90nm
who was it that said that the fx-55 was gonna get real hot when san diego rolled around? was it charlie. too bad he was right.
bachus you have PMQuote:
Originally Posted by bachus_anonym
holy crap 2.8ghz @ 1.4~v's nice
Quote:
Originally Posted by leviathan18
definitely. and I don't think that's what bios can solve. can't be done by trick the temp reading and work, no way. :p:
Nothing amazing. I assume you posted to up your post count. :SQuote:
Originally Posted by i found nemo
We've seen most of the V and SD cores do this. SO its not anything new.
Filling up the post with useless posts like these makes it harder for people to look through the post and find useful information. Less of these omg nice GJ i want one posts makes the thread much better.
btw nice name :rolleyes:
Back on topic,
The 754s still have the memory controller on them, so i think there will be the same problems.
OPB you have probly tried increasing the vcore. What happened when you did? Did it improve things?
Maybe if we could get some light from amd, or send them an email regarding this. It would be interesting to find out whats causing this.
ON the other hand i saw a post somewhere that s37en was "contacting" his AMD guy about this. So hopefully we can get some news from him.
hopefully soon :(
we have to wait until a fx55 or fx57 goes under that cascade