AMD 4P Burning and Prep for OC?

bowlinra

Limp Gawd
Joined
Apr 30, 2012
Messages
195
I've been breaking in my SM H8QGi+-F, with 4x 6166HE and the CM Hyper 212+ for the past 2 weeks. I followed the guides for the Ubuntu 10.10, FAH, TheKraken, Langouste, TPC installations and even manage to load Apache and monitor with my windows HFM.net box.

The question is, should I do anything else in prepping to load the OCNG Bios and start the Overclocking process? Any more memory test? Thoughts?

Current Stock Performance: (It seems the TPF get better as the WU is being processes)
Code:
P6901    TPF  7:45
P6903    TPF 17:12
P6904    TPF 23:33
P8101    TPF 17:15

Temps while folding P6903 with a 460w draw from the wall:
Code:
Turion Power States Optimization and Control - by blackshard - v0.41
Detected processor: Family 10h Processor
Machine has 8 nodes
Processor has 6 cores
Processor has 5 p-states
Processor temperature slew rate:9.0°C

Temperature table:
Node 0  C0:41   C1:41   C2:41   C3:41   C4:41   C5:41
Node 1  C0:40   C1:40   C2:40   C3:40   C4:40   C5:40
Node 2  C0:41   C1:41   C2:41   C3:41   C4:41   C5:41
Node 3  C0:41   C1:41   C2:41   C3:41   C4:41   C5:41
Node 4  C0:41   C1:41   C2:41   C3:41   C4:41   C5:41
Node 5  C0:40   C1:40   C2:40   C3:40   C4:40   C5:40
Node 6  C0:39   C1:39   C2:39   C3:39   C4:39   C5:39
Node 7  C0:38   C1:38   C2:38   C3:38   C4:38   C5:38

Done.
 
Last edited:
You've done what I did when I built mine, no point just dipping a toe in, you are ready for the full on plunge. If you have gotten this fair, you are ready to go [H]ard. Pull the trigger and feel the burn! Ooooh, it is a good burn ;-) For science! :cool:
 
Why would you put a gxf card in a 4P, or even hook up a monitor?

I can answer that. Because I am Grandpa and nobody has made a tutorial to show me a different way. :D

Muusssskkyyyy did you get the hint. :confused:
 
:) was just thinking power consumption is a little high for his configuration but good IPM link :cool:, sure I have no clue how it works
 
460 watts for you config..., have you an external graphic card mounted?
No just using the internal. Power consumption look to be to ranging from 445w - 460w under load of p6903. I also have the fan's set in BIOS to full.
 
I haven't use any IPMI before. Are many folks using it instead of a SSH connection?

I like IPMI... but I'm a linux user not a lover..... and I use it to fold and not much else.

So If I can sit at my windows box and use the GUI on the boxen in the rack... makes me happy.

I wait for Musky to come in and tell me I'm doing it wrong though :p
 
I haven't use any IPMI before. Are many folks using it instead of a SSH connection?

I use it for set-up - you can mount an iso as a virtual cdrom to install an OS. Beyond that, the only thing i was using it for is forced power resets or power-up in the case of an outage. With wake-on-lan, i don't really need that any more. So, while it is a cool thing, it is also not necessary.
 
I like IPMI... but I'm a linux user not a lover..... and I use it to fold and not much else.

So If I can sit at my windows box and use the GUI on the boxen in the rack... makes me happy.

I wait for Musky to come in and tell me I'm doing it wrong though :p

I do not have IMPI and I can sit on my windows box and use the GUI on the boxen :D
 
Apparently, I've lost my skills.. Attempt to load the OCNG BIOS and I can't get it to boot from USB drive.

I followed the Post from the OC thread. I use the bay-wolf instructions and created the Win 98 USB boot drive.

I get to where it boot to the USB and see "Starting Windows 98 ..." then it just hangs. I even disconnected the hard drive and cdrom. I'm using a Iomega Mini 64MB stick, if that matters.

Is Win 98 choking on see 4 CPUs or multiple cores? Thoughts?
 
Apparently, I've lost my skills.. Attempt to load the OCNG BIOS and I can't get it to boot from USB drive.

I followed the Post from the OC thread. I use the bay-wolf instructions and created the Win 98 USB boot drive.

I get to where it boot to the USB and see "Starting Windows 98 ..." then it just hangs. I even disconnected the hard drive and cdrom. I'm using a Iomega Mini 64MB stick, if that matters.

Is Win 98 choking on see 4 CPUs or multiple cores? Thoughts?

Just use unetbootin and make a FreeDOS USB stick, which is one of the choices under "Distributions". That will be easier for you.
 
Just use unetbootin and make a FreeDOS USB stick, which is one of the choices under "Distributions". That will be easier for you.

I couldn't get it to work, received some "Boot Inquiry Error" or Loader I think, anyways I tried another thumb drive and got the same error. I'm might has an issues with the LaunchU3, I couldn't seem to remove it. Anyways the short version, I pulled up a USB Boot Drive with Hiren Utilities, was able to flash the bios.
 
I have 6166HE Stock at 1.8 (200)

I started at 240 - Posted, Booted, Locked up folding about a minute in.
Rebooted, Set to 235, shutdown, powered down, restart

235 - posted, booted, Locked up folding after about 5 mins. This time I took a picture of the screen. http://img208.imageshack.us/img208/3029/p1060988y.jpg
I did notice on "top" the bottom three entries started, right before it locked up. (nautilus, gvfsd-trash, gvfs-afc-volume)
Rebooted, Set to 230, shutdown, powered down, restart

230 - posted, booted, Locked up folding after about 2 mins. Wasn't able to get a picture, I have flip the monitor to another box to check the forum.
Rebooted, Set to 225, shutdown, powered down, restarted with my Hiren Utility and running Memtest86+ (Apparently that going to run for awhile....)

I would have thought I could run any of these speeds on stock voltage. Ideas and Thoughts for the next steps?
 
Memtest86+ (looks like a DOS version) still running this morning. It's 79% Pass with 7 errors, so I'm going to let it finish today and post a picture of the result tonight. I don't know if 7+ error is OK or I'll need to RMA some of the memory.

Should I run a different test to verify the findings?
 
From what I understand, a single error can and will bork a WU.
 
It sounds like you may have a memory issue here. You may want to validate that by relaxing the memory timings with the overclock. I'll have to defer to tear for how to do that - it is a switch with the smocng.sh script.
 
From what I understand, a single error can and will bork a WU.

What also interested is I have complete 10 - 11 Bigbeta WUs without issues.. Althought I have had 2 WUs that failed to restart at the checkpoint and completely started over. Being new to Lunix / Ubuntu folding, I was told this is so what common and there is a bug with stopping the client during the checkpoint being written.
 
What also interested is I have complete 10 - 11 Bigbeta WUs without issues.. Althought I have had 2 WUs that failed to restart at the checkpoint and completely started over. Being new to Lunix / Ubuntu folding, I was told this is so what common and there is a bug with stopping the client during the checkpoint being written.

What usually happens is that the client fails to read the checkpoint files on restart, and thus starts the WU over. That does not mess up the WU, but you lose whatever percentage of the WU was completed. The initial problem is usually caused by shutting down the client while it is writing checkpoint files.

The other thing that can happen which is not usually associated with a restart is an actual client error that causes the client to stop and dump the current WU. Most times when this happens, the same WU is re-downloaded.
 
I booted from my Hiren Utility USB thumb drive and ran the Memtest86+.

It would appear to have retain the OC of 225 and Memory timing don't appear check for the G.SKILL Ripjaws X Series 8GB (4 x 2GB) 240-Pin DDR3 SDRAM DDR3 1333 (PC3 10666) Desktop Memory Model F3-10666CL7Q-8GBXH (That suppose to be a Cas 7 7-7-7-21).

Would that account for the detected errors?



Uploaded with ImageShack.us
 
Running 6166HE at 225 and here is the memory timings and stock voltage.

Memory should be running 1333+ right? Also the last two lines on the voltage (VCore:0.9750) aren't the same is that normal?

Code:
bowlinra@amd4P:~$ dmesg | grep -o Detected.*
Detected 2025.015 MHz processor.

bowlinra@amd4P:~$ sudo tpc -dram
Turion Power States Optimization and Control - by blackshard - v0.41

DRAM Configuration Status

Node 0 ---
DCT0: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=2 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=56
DCT1: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=3 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=54

Node 1 ---
DCT0: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=3 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=53
DCT1: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=3 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=54

Node 2 ---
DCT0: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=2 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=56
DCT1: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=2 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=54

Node 3 ---
DCT0: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=3 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=53
DCT1: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=3 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=53

Node 4 ---
DCT0: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=2 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=56
DCT1: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=2 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=55

Node 5 ---
DCT0: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=3 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=54
DCT1: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=3 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=53

Node 6 ---
DCT0: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=2 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=56
DCT1: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=2 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=54

Node 7 ---
DCT0: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=3 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=55
DCT1: memory type: DDR3 frequency: 1066 MHz
Tcl=6 Trcd=6 Trp=6 Tras=19 Access Mode:1T Trtp=5 Trc=27 Twr=9 Trrd=4 Tcwl=7 Tfaw=32
TrwtWB=5 TrwtTO=4 Twtr=5 Twrrd=3 Twrwr=4 Trdrd=3 Tref=2 Trfc0=0 Trfc1=2 Trfc2=0 Trfc3=0 MaxRdLatency=52


Done.
bowlinra@amd4P:~$ sudo tpc -l | grep 0.pstate.0
core 0 pstate 0 - En:1 VID:47 FID:2 DID:0.00 Freq:1800 VCore:0.9625
core 0 pstate 0 - En:1 VID:47 FID:2 DID:0.00 Freq:1800 VCore:0.9625
core 0 pstate 0 - En:1 VID:47 FID:2 DID:0.00 Freq:1800 VCore:0.9625
core 0 pstate 0 - En:1 VID:47 FID:2 DID:0.00 Freq:1800 VCore:0.9625
core 0 pstate 0 - En:1 VID:47 FID:2 DID:0.00 Freq:1800 VCore:0.9625
core 0 pstate 0 - En:1 VID:47 FID:2 DID:0.00 Freq:1800 VCore:0.9625
core 0 pstate 0 - En:1 VID:46 FID:2 DID:0.00 Freq:1800 VCore:0.9750
core 0 pstate 0 - En:1 VID:46 FID:2 DID:0.00 Freq:1800 VCore:0.9750
bowlinra@amd4P:~$
 
Running 6166HE at 225 and here is the memory timings and stock voltage.

Memory should be running 1333+ right?

TPC does not show you the memory frequency after the OC. I'm sure there a more accurate technical description...but essentially you see the settings as if the system was at 200 bclk (stock). Your ram is actually running at 1200 (1066 * 225 / 200). The reason you are seeing faster timings is that the ram is running at a slower frequency. Since it is slower, the timings can be tightened a bit and still remain within spec.
 
TPC does not show you the memory frequency after the OC. I'm sure there a more accurate technical description...but essentially you see the settings as if the system was at 200 bclk (stock). Your ram is actually running at 1200 (1066 * 225 / 200). The reason you are seeing faster timings is that the ram is running at a slower frequency. Since it is slower, the timings can be tightened a bit and still remain within spec.

Interesting, I must have missed that point in the earlier threads.
Maybe the question I should have asked, Do you think I have a memory problem and if so what the best way to track down the problem area for RMA?
 
.......Also the last two lines on the voltage (VCore:0.9750) aren't the same is that normal?

bowlinra@amd4P:~$ sudo tpc -l | grep 0.pstate.0
core 0 pstate 0 - En:1 VID:47 FID:2 DID:0.00 Freq:1800 VCore:0.9625
core 0 pstate 0 - En:1 VID:47 FID:2 DID:0.00 Freq:1800 VCore:0.9625
core 0 pstate 0 - En:1 VID:47 FID:2 DID:0.00 Freq:1800 VCore:0.9625
core 0 pstate 0 - En:1 VID:47 FID:2 DID:0.00 Freq:1800 VCore:0.9625
core 0 pstate 0 - En:1 VID:47 FID:2 DID:0.00 Freq:1800 VCore:0.9625
core 0 pstate 0 - En:1 VID:47 FID:2 DID:0.00 Freq:1800 VCore:0.9625
core 0 pstate 0 - En:1 VID:46 FID:2 DID:0.00 Freq:1800 VCore:0.9750
core 0 pstate 0 - En:1 VID:46 FID:2 DID:0.00 Freq:1800 VCore:0.9750
bowlinra@amd4P:~$[/CODE]

Yes. Normal. My 6166HE set ups have 4 different "set" VCore voltages.
I now have them all forced to 1.0000, which is a tad higher than all their built-in values.

 
Ran a full pass of Memtest86+ with no issues at Stock 1.8Ghz. Test still didn't show the 7-7-7-21 that I thought it would. Why am I have issues with the overclock speeds? What should be my next steps?



Uploaded with ImageShack.us
 
When the grub menu comes up just scroll down the list. Memtest is there.
Sorry I'm really a Noob. I couldn't find it. I tried to download it from the app store, but it say it's installed. I can't find it thru the menus and I see the grub is a bootloader, I don't think I'm running that, it a dedicated folding box following Musky's standard Ubuntu 10.10 install guide. I was able to run my test from my Hiren boot USB drive.
 
What board are you running?

Soon (in a day or two) I'll have a version of OC BIOS that
will allow testing XMP profile w/o OC (CL7/DDR3-1333
in your case).
If XMP w/o OC gives issues you most likely have bad
DIMMs (or something off between CPUand DIMMs).

Also, re-testing @ 225 (I'm assuming that's what gave
you errors) and checking whether you get errors at same
offsets as before would be a worthy exercise.
Consistent errors would mean DIMM issue. Inconsistent
errors suggest OC/tuning/board issue.

Alternatively, if you have enthusiast board that supports
XMP you could test your DIMMs in it.
 
What board are you running?

Motherboard is H8QGi+-F. I went with all the recommend hardware I had thought.

Also, re-testing @ 225 (I'm assuming that's what gave
you errors) and checking whether you get errors at same
offsets as before would be a worthy exercise.
Consistent errors would mean DIMM issue. Inconsistent
errors suggest OC/tuning/board issue.

Alternatively, if you have enthusiast board that supports
XMP you could test your DIMMs in it.
I'll have to read up on the XMP stuff. I have X58 MSI Pro-E and X-Power Big Bang, I also have some bargin bin board coming today Intel H61 ? http://www.newegg.com/Product/Product.aspx?Item=N82E16813135288 it was $15 after rebate, so it was hard to pass up.
 
Back
Top