Unofficial Opteron 939 Thread

[Verxion]

I tried each of the settings given by LithoTech, to no avail. I usually fail on double super pi within 2 loops, regardless of which of those settings I tried.

 

wardog, I tried the settings from OCZ's support page as well. I really REALLY need to up the voltage from 2.6 or it just doesn't run well. It fails within a few seconds at 2.6 volts, but at 2.8 it runs through to 10 or so loops before failing on superpi.

 

-Verxion

[wardog]

It fails within a few seconds at 2.6 volts, but at 2.8 it runs through to 10 or so loops before failing on superpi.

 

Was this with that -3(TCC(D)©(5)) file or Tony's?

 

Are these errors while OC'd? I'd be raking thru my BIOS with a fine toothed comb double checking my entries.

 

IMO a factory(OEM DFI) BIOS until I found stable settings. Something seems amiss. HT, HTT, somethings isn't adding up.

[Verxion]

wardog,

 

I am not overclocking at all. As for the BIOS, it is 632-3, which thus far seems to the most stable. I have seen people talking about HT and HTT before, but I haven't seen any settings in the BIOS at all that seem to deal with them.

 

Where do I adjust HT and HTT? What settings should they be at?

 

-Verxion

[wardog]

@ Verxion

 

If you'd like, I'll offer ya a big hand here. Other than the occasional day spent with the Wife, I get the winters off. It's not like I don't have the time.

 

I've got the HW here to duplicate yours.

 

PM'd

[Verxion]

Replied to PM. THanks a billion man.

 

-Verxion

[valic]

Hi guys, sorry I'm noob here... anyone knows about some problems with opteron 146 / DFI NF4 Infinity SLI / 2x512 MB corsair xms4400c25pt (samsung tccd i suppose)? I cannot get them run stable at default values... (random reboots before windows logon or reboots in memtest) The only settings that worked until now - and I have succesfully ran memtest at 300%- are:

 

opteron 146 @ 2500 MHz (250x10) / HT 4x / 1.45v

memory: 250 MHz (cpu 1:1) 2.5 4 4 8 1T (all others set to auto) / 2.7v

 

increasing FSB to 260/270 MHz resulted on memtest errors :eek2:

 

I have tried last 3 BIOS es from DFI but same results.

 

Before xms4400c25pt I had 2x512 Kingmax DDR400 and the system worked well with default values.

 

does anyone knows about incompatibilities between opteron and tccd memory?

 

Thanks.

[valic]

After few days of tests, I tried to play with max async latency and read preamble - I set to 9-8 and now it runs at 2830 MHz (333x8.5)/1.5v (BIOS displays 1.52v) and memory at CPU / 11 (254 MHz) /2.9v (BIOS displays 2.87v)

If I set vDIMM to 2.8 (2.77v) memtest displays errors.

Is the vDIMM 2.88 ok for tccds?

 

Today I have succesfuly run 2x512 MB Corsair XMS4400C25PT at 2.7v vDIMM (my BIOS displays 2.76v) with MAL/RP 9/7 (other settings 2.5-4-4-8-1T) on DFI nF4 Infinity SLI and opteron 146. Just in case anyone else will have this stupid problem...

Thanks for help..

[raincouver]

Noob here as well.

 

Well I upgraded from a 3200+ Venice to a 165 right after Xmas. I cleared the BIOS after I popped in the 165 and within 1/2 hour I was up to 2.9GHz and Stress Priming with no problems at all.

 

Fast forward 2 months and I am "borrowing" my brothers OCZ 2x1 GB EB Platinum Kit. Max RAM speed on this is 246 MHZ @ 2.6 ish Volts (3-3-3-8). This RAM does not like higher Voltage....Memtest errors even at lower speeds (220's). I started with Ryder OCZ's default RAM settings, relaxed the timings and started to find max CPU overclock (1:2 Divider). I get memtest errors right away.

 

Using 2.6V Vdimm I cant even boot past 275 FSB using any dividers. It will boot sometimes with 2.8Vdimm. But unfortunately not prime stable or memtest stable with that voltage. I want this RAM to work so bad but I been playing with it for 2 weeks now and it is sooo damn frustrating....have been getting a lot of no display BIOS screens when I oc too much.

 

Whats with the Opty and dividers? The Venice took all the dividers like a champ. I just wondering if I should try and find a set of PC4000. Should I try a differnt BIOS?

[raincouver]

Noob here as well.

 

Well I upgraded from a 3200+ Venice to a 165 right after Xmas. I cleared the BIOS after I popped in the 165 and within 1/2 hour I was up to 2.9GHz and Stress Priming with no problems at all.

 

Fast forward 2 months and I am "borrowing" my brothers OCZ 2x1 GB EB Platinum Kit. Max RAM speed on this is 246 MHZ @ 2.6 ish Volts (3-3-3-8). This RAM does not like higher Voltage....Memtest errors even at lower speeds (220's). I started with Ryder OCZ's default RAM settings, relaxed the timings and started to find max CPU overclock (1:2 Divider). I get memtest errors right away.

 

Using 2.6V Vdimm I cant even boot past 275 FSB using any dividers. It will boot sometimes with 2.8Vdimm. But unfortunately not prime stable or memtest stable with that voltage. I want this RAM to work so bad but I been playing with it for 2 weeks now and it is sooo damn frustrating....have been getting a lot of no display BIOS screens when I oc too much.

 

Whats with the Opty and dividers? The Venice took all the dividers like a champ. I just wondering if I should try and find a set of PC4000. Should I try a differnt BIOS?

[wizurd of awwz]

Try the 704-2bta.

 

But I wouldn't try OCing that RAM too much as it will only go so far. Check my sig.

[raincouver]

When I first put the Opty in I was still using my OCZ Gold VX. The Gold loved to overclock at high voltages. So it was kind of strange overclocking the Platinum kit at 2.6ish volts. Anyways, I already knew what settings I could run the gold at so it was a matter of finding out max OC of the CPU (2900ish at 1.39V) and finding a CPU/FSB divider (3/4).

 

I knew that the Platinum RAM overclocked to to 246MHz but when I started using dividers thats when really strange things started to happen. I took a look at the OCDB again last night and found a couple of different settings to try. I actually used Angry's settings as a starting point. The settings that I finally changed included: Undervolting the Vdimm - DRAM voltage at 2.5 with the +0.03 V enabled, CPU Vid Startup Voltage - 1.55V and proably most significant CPU Vid - 1.55V. Its just a guess but because the 2GB kit puts more stress on the memory controller the CPU needs more voltage for stability.

 

I am 8 hours 2xSP2004 (blend test) stable at 311 x 9 and currently I am running the "Large" test. Its been 3 hours and so far so good(silently praying). My problem was I was hell bent on using the roughly the same cpu voltages for hitting the same overclocks. This I guess is not so obvious when you are running on 3-4 hours of sleep a night

 

If everything works out, I would like to try to tighten the timings on the RAM, and find a stable max CPU overclock (2900ish)with less voltage. I will post an entry in the database when I finish.

 

DIY OCDB Rocks !!!

[Guest Indred]

Hey yall I got my Opty 185 and it was working ok. I've had it for about a month now and still can't seem to get it stable. I first noticed it crashing when I was playing BF2 and now it won't even pass Prime95. It reboots my computer after probably 45minutes or so. I have a strong Idea that its the CPU as I had it OC'ed when i pulled out my 64 3000+. Never had a problem with that one. I've had a few people tell me about checking the PSU. I plugged in my second PSU and pulled my 5 harddrives and Optical drive off the main PSU and put it on my second. The only thing thats attached to my main PSU is 2video cards (7800GT) MOBO of course (NF4 SLI Expert) Fans and my water pump. Should be fine for a 550W AirCool. So I'm wandering if its the CPU for sure or not before i try to Send it back or some thing. Also on load i tend to have a 15c differential between the two cores. Max temp is usually ~60C load which even for Optys its high and its water cooled to beat. I've reapplied AS5 a few different times and it didn't seem to make any change. Did i get a Goofed up one? help. Thanks.

 

EDIT: i don't know if my bios version in my sig is correct.