Folding@Home on Core i7: Bonus points for 8 core Linux VMs

bomberboysk said:
The easiest thing to do? Go set everything stock, write down the settings it automatically sets, then put those settings in and turn off auto. Then go start increasing base clock a few mhz at a time until you are no longer stable, give it a bump in vcore and continue the process until you cannot get stable any longer. It takes time to get a good overclock, but in the end the results are worth it(eg- a D0 should not have to struggle at all to get 4Ghz, especially on a board with better than average overclocking features such as power design on the evga vanilla)
Click to expand...

What's "Base Clock"? I have a C0

EDIT: Nevermind.

CPU Host Frequency - Speed of the bus i.e. Base Clock (BCLCK) or Quick Path Interconnect (QPI) speed.
Click to expand...
 
Last edited:
bomber has you on the right track ;). There are quite a few variables, but play around with it.

Getting stable at 4.0 isn't hard, but finding the best settings even for a D0 takes time. I'll have to admit...it took a weekend of messing with stuff to get it optimized at 4.1Ghz.
 
Well I think I'm screwed.

I reset to default and raised the Host Frequency to 177, left everything else Auto and it ran fine at 3.7Ghz, so I ran Prime95 and Linpack for about 20 minutes, at temps lower than the VM folding client, about 60c. Sweet.
VM was running upper 70's because I knew I needed to blow out my case.

I shut it down to blow out my case. So I blew out my case, removed my CPU fans and I wiped down all the edges of the blades because they had buildup on them, and put them back on. No big deal.

So, I plug in my power supply, which was turned off with the rocker switch, and it "snapped" when I plugged it in. Wtf? Idk if it matters.

Anyway, I turn it on and my board won't POST. I get F3, then F2, then 68, then it just goes back and forth from F3 to 68. No beep...

EDIT: So I removed my cpu, gpu, ram, reseated everything, unplugged and replugged all the power connectors, no joy.

Did I fry my cpu?
 
Last edited:
Bodaggit23 said:
Well I think I'm screwed.

I reset to default and raised the Host Frequency to 177, left everything else Auto and it ran fine at 3.7Ghz, so I ran Prime95 and Linpack for about 20 minutes, at temps lower than the VM folding client, about 60c. Sweet.
VM was running upper 70's because I knew I needed to blow out my case.

I shut it down to blow out my case. So I blew out my case, removed my CPU fans and I wiped down all the edges of the blades because they had buildup on them, and put them back on. No big deal.

So, I plug in my power supply, which was turned off with the rocker switch, and it "snapped" when I plugged it in. Wtf? Idk if it matters.

Anyway, I turn it on and my board won't POST. I get F3, then F2, then 68, then it just goes back and forth from F3 to 68. No beep...

EDIT: So I removed my cpu, gpu, ram, reseated everything, unplugged and replugged all the power connectors, no joy.

Did I fry my cpu?
Click to expand...
Simple question, but did you check to make sure the cpu power connector was fully inserted? If those are loose sometimes it can cause issues, same goes for pcie power connectors and motherboard 24pin. Afterwards pull all the memory and try to boot with one stick in the slot furthest from cpu. Next if that does not work i would go ahead and pull the cpu, see if there are any bent pins in the motherboard socket. 68 is a memory error code, which can deal with either 1- bent pins, 2- on die memory controller, 3- DIMM's themselves, or 4- Link between memory controller and memory itself. I doubt you have another X58 rig closeby to test with, but if you do test out your cpu in it, otherwise if you have a DDR3 AM3 or 775 board around, use it to test your memory dimm's.

Also, what exactly do you mean by "snapped"? Electrical sound like a spark or more of a mechanical sound?
 
Last edited:
The electrical sound was like plugging in a live cord, but the psu was
turned off, so I don't think it could have done anything to the internals.
The snap came from the plug itself when it came in contact with the psu input plug.

I completely removed and reseated the CPU and reinstalled the cooler.
When I was putting the fans back on after cleaning, I kind of jostled
the cooler so I thought that was the issue initially.

I removed my GPU and all 3 sticks of RAM and reseated.

I unplugged all the power connectors on the board, CPU, main power
connector, GPU connectors, and reseated all of them.

I have not tried booting with one stick of RAM yet, but I will.

What could I have done?

Once it cycles to 68 or 6B, whichever it is, it just stays there,
or it will go to F3 again, then back to 68 (6B?)

PS, DIMM slot 1 on the E758 is the first Green slot closest to the cpu,
but I will try each slot, with each stick of RAM.

Here's the codes:
[YT]<object width="425" height="344"><param name="movie" value="http://www.youtube.com/v/ncd7AWwGntI&hl=en_US&fs=1&"></param><param name="allowFullScreen" value="true"></param><param name="allowscriptaccess" value="always"></param><embed src="http://www.youtube.com/v/ncd7AWwGntI&hl=en_US&fs=1&" type="application/x-shockwave-flash" allowscriptaccess="always" allowfullscreen="true" width="425" height="344"></embed></object>[/YT]
 
Last edited:
What is happening there is F2->F3 is checking system memory, and 68=6B which also relates to memory. Did you remember to put the memory in the correct DIMM slots? On the 758 the green slots are the ones that should be used for a 3x DIMM configuration.
 
Ok, this is weird...I removed 2 sticks of RAM and started the pc with one stick in the number 1 slot and it POSTED, and booted into Vista.

Wtf?

I added the next stick in the next slot, it POSTED and booted...
I added the last stick and it POSTED and booted. Umm...?

Did I cause something to "trip" on my board? How?

You read what I did as far as overclocking, could it be harmful how I did it?

I reset BIOS to default, then I increased the CPU Host Frequency slowly until I got up to 3.7Ghz. I left everything else on Auto. Is that bad?
 
Bodaggit23 said:
Ok, this is weird...I removed 2 sticks of RAM and started the pc with one stick in the number 1 slot and it POSTED, and booted into Vista.

Wtf?

I added the next stick in the next slot, it POSTED and booted...
I added the last stick and it POSTED and booted. Umm...?

Did I cause something to "trip" on my board? How?

You read what I did as far as overclocking, could it be harmful how I did it?

I reset BIOS to default, then I increased the CPU Host Frequency slowly until I got up to 3.7Ghz. I left everything else on Auto. Is that bad?
Click to expand...
What it sounds to me is like you might have a bad stick of memory, i would run memtest on them, unless your board was not recognizing the memory profiles correct voltages and such which does happen sometimes.

Auto voltages tend to go far too high, i would get out LinX and Prime95 and start by going from stock voltages, will increase life of chip and run cooler at lower voltages(eg- even with a C0 3.8 should be relatively easy to attain with a halfway decent chip). Short term you might notice higher temps, but it can be unstable from the higher voltages(go figure) and long term it will shorten life of the chip.
 
I'm not sure what you just recommended. I'll run Memtest right now.

I think you're right about the memory though. Maybe I could try again, the same method,
but make sure to turn Auto OFF on anything memory related?

I did run Prime95 and Linpack (CPUTorch) for about 20 minutes before I cleaned my rig.

Why did it take so long to "reset"? I virtually did nothing. It just decided to boot again. :confused:
 
Bodaggit23 said:
I'm not sure what you just recommended. I'll run Memtest right now.

I think you're right about the memory though. Maybe I could try again, the same method,
but make sure to turn Auto OFF on anything memory related?

I did run Prime95 and Linpack (CPUTorch) for about 20 minutes before I cleaned my rig.

Why did it take so long to "reset"? I virtually did nothing. It just decided to boot again. :confused:
Click to expand...
It only booted once you put it to one stick of memory right? Sometimes if one stick of memory is bad it will not read the memory profiles correctly for timings/voltages.

Basically, when you are running at stock speed, write down the numbers it has for all your voltage settings in bios, and then change them to those settings. Then, start overclocking the base clock until it is unstable, increase vcore until it is stable, increase base clock until unstable, and increase voltage until stable, continuing until you get either the clockspeed you want or it will not overclock any higher. You may need to bump the PLL voltage up a little after 3.6Ghz or so, the other voltages shouldnt need to be adjusted as long as you arent aiming for over 4Ghz.
 
It's running right now with all sticks in place. I'll test them overnight, but I don't think I have a bad stick of memory.

If the stick was bad, why does it boot now all of a sudden? Idk...

Thanks for your help. :good:
 
Bodaggit23 said:
It's running right now with all sticks in place. I'll test them overnight, but I don't think I have a bad stick of memory.

If the stick was bad, why does it boot now all of a sudden? Idk...

Thanks for your help. :good:
Click to expand...
Could just be an issue with the memory on one stick that holds the profiles even, as long as it memtest's ok and you just input the correct voltages and timings it shouldnt cause any future issues,it could be just one of those weird occurrences sometimes that cant be explained.
 
Bodaggit23 said:
It's running right now with all sticks in place. I'll test them overnight, but I don't think I have a bad stick of memory.

If the stick was bad, why does it boot now all of a sudden? Idk...

Thanks for your help. :good:
Click to expand...

Did you reset the CMOS? If you didn't, adjusting the amount of memory you have might have automatically reset it, causing it to work again.
 
Yes, I reset the CMOS about 10 times trying to get the damn thing to boot.

I attempted to manually change my memory timings, and it refused to boot again. My BIOS has SMP Profile so maybe I'll try that.

I have it running back at 3.4Ghz, folding a big unit. My temps are back down to 60-63c running the VM client. Needed a cleaning bad. It was getting into upper 70c.

I'll mess with it more during the week. I'd do more testing but I've been not folding all day because of this, so it's gonna fold for a bit. :cool:
 
Any of you have experience with transferring files between the Linux VM and Windows? I'm trying to set up the scripts they have to backup the WU folder to the windows machine, sync the clock every hour, run it off a ramdisk, and simultaneously upload and download.
 
My first BSOD.

SYSTEM_SERVICE_EXCEPTION (0x0000003b)

Happens everytime I try to run Everest.

I think I hosed something with my failed amateur overclocking attempt, and I hope it's just a Windows issue.

Until I find out, I won't be folding. :mad:

EDIT: Guess it's not Windows related...I have some work to do.

DUAL-1.jpg
 
Last edited:
0x0000003b is usually a memory issue related BSOD. Sounds like your booting issue might have not been so random now. Have you memtested each stick individually for a few passes?
 
I have a bum stick.

You'll see the CPU-Z screenie shows 6Gigs installed, but running in Dual Channel.

I ran Memtest and it only acknowledged 4Gigs, so...

I tried each stick and sure enough, one is bad. PC will not boot with that stick by itself. Strange that it boots at all with that bad stick in.

Let's see how good Corsair customer support is...
 
Bodaggit23 said:
I have a bum stick.

You'll see the CPU-Z screenie shows 6Gigs installed, but running in Dual Channel.

I ran Memtest and it only acknowledged 4Gigs, so...

I tried each stick and sure enough, one is bad. PC will not boot with that stick by itself. Strange that it boots at all with that bad stick in.

Let's see how good Corsair customer support is...
Click to expand...
When you rma i would suggest doing an rma of all sticks, to ensure matched pairs.
 
You must log in or register to reply here.
Back
Top