rig down

[64dragon]

tonight i'll be swapping in the new 9505 and will work on the OC over the next week or two depending on how things go between prime sessions since i won't see failures untill i get home from work.

[INFRNL]

Thats unacceptable; we need your computing power now, right now. j/k. I Hope it goes well and you get back to help soon

I need to do the same thing, but I am in a decent spot for production and do not want to take from the team. Maybe when I build my next system; I will take more time to get higher clocks to see if helps production and then slowly work on all the rigs 1 at a time.

good luck

[Jaco]

Let us know how that Q9505 clocks.
I'd like to compare with a regular Q9550.

[64dragon]

chip is in and running my 2nd OCCT. i boot up and found VID was 1.287 :eek: i am now trying at 1.23v in bios which came up as 1.21v idle and curently showing 1.16v load in Everest. hopefully it will play nice. also, temps were hitting 45C on my TRUE Black

[64dragon]

at stock (2.83ghz) boinc bench is:
2878 Whetstone
8382 Dhrystone

which beat out my X3330 at 3ghz, was:
3143 Whetstone
6531 Dhrystone

both on version 6.2.19 Vista 64bit

[Kondik]

Sorry for intrusion but I have a side question

Hm Hm

I don't get your Xeon's Whetstone

I myself have this

2988 floating point MIPS (Whetstone) per CPU
9809 integer MIPS (Dhrystone) per CPU

your FPMIPS is higher
and IMIPS lower

I think I should stop doing drugs

[jcool]

at stock (2.83ghz) boinc bench is:
2878 Whetstone
8382 Dhrystone

which beat out my X3330 at 3ghz, was:
3143 Whetstone
6531 Dhrystone

both on version 6.2.19 Vista 64bit


That score on the Xeon is weird, it looks like its from a 32bit OS/BOINC client.

Unfortunately Whetstone is the important one so your X3330 was def. faster.

[64dragon]

Yea, i was surprised to see how different the scores were. What i meant by the 9505 being faster is based of D_A's equation (Whet + Dhry) x .05 x # of cores. The 9505 will do 2252 Boinc points per day and the 3330 would do 1934 Boinc ppd.

I'm not 100% sure that the 3330 score was on 64bit but as far as i can remember i've only run 64bit vista...unless the trial period that i did in the begining wasn't and thats what the screenshots i had were from. I do know both results were from 6.2.19

I made it through 8hrs of small fft on Orthos at the 1.23v in bios last night which makes me happy. It is currently crunching since i'm at work but tonight i'll start to OC.

[64dragon]

so far, the chip is doing pretty good, i was a little worried in the begining based on the VID. i went to 3.1ghz @ 1.237v bios and ran OCCT for over an hr fine set to just CPU stress. I just jumped up to 3.3 and 1.25v bios (1.19v load based on Everest) and have been running OCCT for 1/2 hr with no issues...other than the 47C temps (which isn't an issue). if i can get 100% stable on my given voltages then the chip OC's better volts wise than my Xeon did but has higher temps.

[64dragon]

oh how the tides turn... i'm at 3.5 now but needed to go up to 1.30v in bios to be 8hrs Orthos small FFT stable. I'm testing large FFT's now, VTT and NB are both at 1.32v, core 4 failed last night so i had to up it a notch. I am a little surprised/disappointed that i needed to add .05v to go from 3.3ghz to 3.5ghz, this puts the cpu voltage at the same as my Xeon needed.

I worked on finding max FSB last night and 6x480 (2.88ghz) needed 1.38 vcore and the rest of the voltages were on auto. Everest read VTT and vNB as 1.38v. 7x480 (3.36ghz) wouldn't boot at the same voltages. I know the board can handle it, i had the Xeon at 8x487 (3.9ghz) just to see what it could do in the past.
So, does that mean the 9505 is limited to 4ghz (ignoring that VTT and vNB are too high for daily setting)?

[Kondik]

oh how the tides turn... i'm at 3.5 now but needed to go up to 1.30v in bios to be 8hrs Orthos small FFT stable. I'm testing large FFT's now, VTT and NB are both at 1.32v, core 4 failed last night so i had to up it a notch. I am a little surprised/disappointed that i needed to add .05v to go from 3.3ghz to 3.5ghz, this puts the cpu voltage at the same as my Xeon needed.

I worked on finding max FSB last night and 6x480 (2.88ghz) needed 1.38 vcore and the rest of the voltages were on auto. Everest read VTT and vNB as 1.38v. 7x480 (3.36ghz) wouldn't boot at the same voltages. I know the board can handle it, i had the Xeon at 8x487 (3.9ghz) just to see what it could do in the past.
So, does that mean the 9505 is limited to 4ghz (ignoring that VTT and vNB are too high for daily setting)?

Don't be sad, I have to go from 1.225 V to 1.3 V to go from 3333 MHz to 3666, I think I have the sadest little I7 when it comes to V. :hehe:

[64dragon]

to the guys better at OCing than me, what does it mean when the same core keeps failing? when i ran small FFT and large, it is always core 4 that fails. does that mean i need more vcore or VTT since the most recent was during large FFT 320k test?

[Kondik]

VCore or QPI V is low. Strange thing is when I oc is allways a BSOD :p: I think I should play less with the OC in summer.
Wait C2Q or I7 ? I'm kinda lost , if it's I7 then it's V core and QPI.
If it's C2Q then North Bridge Voltage + Vtt + Core

[64dragon]

since last post, i upped vcore and VTT a notch, ran orthos small FFT for 9hrs error free and then large FFT for 10hrs error free, but once i start crunching i come up with computational errors pretty quick.
What should I adjust or do now?

[Jaco]

That's weird. If you're primestable that long , you should have no trouble at all with WCG :confused:

[[XC] riptide]

since last post, i upped vcore and VTT a notch, ran orthos small FFT for 9hrs error free and then large FFT for 10hrs error free, but once i start crunching i come up with computational errors pretty quick.
What should I adjust or do now?

I'm gonna say its a memory issue. You can prime all you want on smalls... but it won't pick up some memory problems, and I know WCG is sensitive to memory issues... even if it doesn't use memory bandwidth all that much.. try loosening timings, maybe upping vmem a notch.

[64dragon]

That's weird. If you're primestable that long , you should have no trouble at all with WCG :confused:

that was my thought


riptide;3976211']I'm gonna say its a memory issue. You can prime all you want on smalls... but it won't pick up some memory problems, and I know WCG is sensitive to memory issues... even if it doesn't use memory bandwidth all that much.. try loosening timings, maybe upping vmem a notch.

i'm runnging this chip on the same board and ram that i have been since October, all i did was change the cpu. the ram is currently clocked lower than it was before. however, i will try when i get home from work and see what happens. at this point, any idea is appreciated

[64dragon]

STILL unstable, have been trying to tune the GTLs but haven't had any luck. everything i tried with tuning the GTL's made me less stable,

tried:
-5,-45,-5,-45 and that lasted 19minutes in Orthos (core 4 failed)
10,-35,10,-35 lasted 9 minutes (core 4 failed)
i looked at the calculation again and saw that 10,-30,10,-30 was a closer match in voltage but that lasted 6 minutes (core 4 failed)
it seemed going this way was worse so i tried going the other way
-10,-55,-10,-55 lasted 5 minutes (core 1 failed)

then i tried positive values:
50,10,50,10 and failed in 10 minutes on core 4
70,30,70,30 failed in 21 minutes on core 4 (my longest time up to this point)
80,40,80,40 failed in 5 minutes on core 4 (a bit worse than above run)
90,50,90,50 failed after 1hr 13min on core 4 (new record for longest run)
100,60,100,60 failed at 41 min on core 4 but with a fatal error instead of a rounding error like the rest. not sure what that means in relation to GTL tuning

so, i went back to 90,50,90,50 since that was the best i've had and then upped my vNB a notch and failed in 34 minutes. even setting VTT and GTLs to Auto is unstable now, i fail in 40 minutes.
Anyone have any more ideas?

It seems i got a bad apple, i'll be giving this chip to my dad. Maybe the Xeon will clock better in a UD3P and i'll consider using that and give the Maximus II Formula to my dad.

[64dragon]

i've given up on this chip. i was at 3.5ghz for a couple days but then got a BSOD so i dropped down to 3.4ghz at the same voltages and a few days later got another BSOD (saying something about a processor thread).

Last night i just got the rig back up with the Xeon back in, i'm going to work on the OC with this since i learned a bit about tuning the gtl's and moving to the 400 strap while messing with the 9505. With some work, i think i can get it running at 3.8ghz (up from 3.5ghz).

[[XC] Hicks121]

Mine has the same VID as you. Must have got them from the same place ;)

Its crunching away for the last couple of weeks @ 3.7. What mobo are you running. Im using my Max Formula x38. & really want to swap it out & get it into the ud3p as that board will do 500+ fsb.

[64dragon]

i was running it on my Maximus II. i plan on getting a UD3P for my dads build cause they're a great price and seem to be dependable since so many people are buying them. I've been thinking of keeping the UD3 for myself but the blue doesn't go with my theme (black and red).

[Sadasius]

Crunching away on my setup without any problems. Turned my RAM down and was fine. Had it at 1800 but then turned it down to 1440. CPU still going nuts at above 4ghz no problem. Using the Maximus Extreme which is very similar to that formula.

[64dragon]

i'm down again. got the UD3P so it is time to start assembling dad's pc but gonna test my Xeon in it first and see how it is compaired to my Maximus II Formula. i'll have a few points here and there while testing the OC