are the ES 6234 chips compatable with Asus KGPE MB?

Status
Not open for further replies.
S

Shadowmeph

Limp Gawd
Joined
Dec 3, 2011
Messages
156
I am not sure if I am posting in the proper place and if not please move to the right place.
I have an Asus KGPE-D16 mother board and it works perfectly with my Opteron 6128 chips. I picked up some ES 6234 chips and have installed them onto this board but I get reboots when this board is stressed at all. the funny thing about that is that the reboots Vary in time .

these is my second set of these chips that I have tried because I originally thought that the chips were flaky so I contacted the person which sold me these chips and they swapped them. also I have RMAed this board thinking that the Board it self was messed up . But the exact same thing happens.

I have ran Mem tests and even replaced all the non ecc ram and tested them which came back with no errors at all.

I have the newest Bios and played around with different settings in the Bios but still same thing happens.
So now I am wondering if it is the motherboards them selves that don't like these ES chips.
 
Does the support page on the asus site list the 6234 as compatible. If it does have you tried just one cpu to see if the board works with just one.

Another thought - what PSU are you using?
 
One item you will need to give the experts here is ALL the information on the PSU.
Then you will need to tell what application you are running.
I can't help you but these are questions that will be asked.
 
I don't know about the ES chips. 6234s should be compatible. I ran 6272s for a while. Make sure you reset the BIOS to the defaults and try it with no other cards in the MB.

I'd also check to make sure you aren't running a beta version of the BIOS.
 
Smells like the cpu's might need a little more voltage or your psu is too weak for these chips.
 
Rebooting when under load is a classic example of not enough vcore voltage to the chips. If you are trying to run the chips are 'stock' settings, then it is most likely a PSU issue.... been there... done that.
 
 
1. Verify that the PSU is adequate as fellow teammates already suggested :)

2. What is the part number of your CPUs? Early samples (*G43) aren't known for reliable operation.

3. You could try to speculatively raise the vcore using TPC (EDIT: make sure to use 0.44-rc2) and see
  what happens (also disable PowerNow/CPB/Turbo)

  (EDIT: and the opposite as well -- keep everything enabled and force CPUs to first bin turbo by means
   of TurionPowerControl -psmax 1 )

4. You could try to lower ambient temps and see how it affects the issue (esp. if it's as
  simple as opening the window)

5. If you're running (or can install) Linux, we could try to learn more about the failure.
  If you feel like spending time on it you could do... this:

  - install mcelog: sudo apt-get install mcelog (afterwards -- make sure it's running: ps auxw | grep [^]]mcelog)

  - enable verbose process crash reporting: http://hardforum.com/showpost.php?p=1039390775&postcount=264
   (not ultimately necessary but may come handy later)

  - set up and start folding

  - wait for the machine to crash (the longer time it takes to crash, the better chance
   of capturing useful information)

  - once machine reboots, examine output of dmesg and contents of /var/log/messages
   for information like: Machine Check Exceptions/HARDWARE ERRORS and RCU
   stalls

  - report your findings
 
Last edited:
ok well I have tried a couple of different PSUs the original PSU I was using was a seasonic SS-850 HT which during one of my stress tests I used a multi meter to make sure there wasn't any jumps in current and found nothing, but to be cautious I bought a new OCZ z850 PSU and redid all of the test but same thing happens, I am new to multiple CPU and well even OCing CPUs .
As for the numbers I am not sure which number is the correct one to post on one of the chips the first number is zs 202043tcg43 I am thinking that the last digets of g43 are what you are refering to?
when this first started to happen way back in OCT I pulled the board and have had it since sitting out in the open so I can get at everything easier.
these chips at idle are at 18c underload using something like OCCT they get up to 48 c
at this time it is all running on the windows 7 because I am more familiar with WIndows 7 . but when I am folding I use linux
I really need for someone to wlak me through the changing of the vcore
when I had the other es chips I was using this to start out
cd ~/tpc/bin/Win AMD64
TurionPowerControl -boostdisable
TurionPowerControl -fo 4
sleep 1
TurionPowerControl -set pstate 0 freq 3000 vcore 1.200
TurionPowerControl -set pstate 1 freq 3000 vcore 1.200
TurionPowerControl -set pstate 2 freq 3000 vcore 1.200
TurionPowerControl -set pstate 3 freq 3000 vcore 1.200
sleep 1
TurionPowerControl -fo 0
now I have know idea if this was correct but I was using this as a basis which was given to me for the factory settings of the other es chips
udo modprobe msr
sudo modprobe cpuid
cd ~/tpc/bin/Ubuntu-x86_64
sudo ./TurionPowerControl -boostdisable
sudo ./TurionPowerControl -fo 4
sleep 1
sudo ./TurionPowerControl -set pstate 0 freq 2000 vcore 1.0875
sudo ./TurionPowerControl -set pstate 1 freq 2000 vcore 1.0875
sudo ./TurionPowerControl -set pstate 2 freq 2000 vcore 1.0875
sudo ./TurionPowerControl -set pstate 3 freq 2000 vcore 1.0875
sleep 1
sudo ./TurionPowerControl -fo 0
 
Both of those PSU have strong enough 12v rails. I would:

- install a single CPU
= run memtest to verify the memory is good
= run a FAH run at stock speeds to verify stablility

if above work good, swap CPUs and repeat. if that is good add the second CPU and repeat again.

ES can be squirrelly sometime. best to reduce the number of variables and find a stable state and then expand.
 
402blownstroker said:
Both of those PSU have strong enough 12v rails. I would:

- install a single CPU
= run memtest to verify the memory is good
= run a FAH run at stock speeds to verify stablility

if above work good, swap CPUs and repeat. if that is good add the second CPU and repeat again.

ES can be squirrelly sometime. best to reduce the number of variables and find a stable state and then expand.
Click to expand...
TY for suggestion.
I have done this already but the same thing happens so I installed the 6128 chips back installed the 6128s and been folding since. but I really want to get these ES chips going so I can get back to folding with some decent numbers once again.
I guess I am going to switch the 6234 ES chips again and reinstall Linux .
 
originally I was using 8 x 2 gb Kingston KVR1333 ram non ecc then I bought new ram which 8x1GB crucial
 
I'm a dope :eek: just looked at your name. ;)

I got my D16 board from MU_Engineer and have only tried 6128's in it, haven't had the time or coin to try 62xx's in it.

If I may ask...where did you get your ES's

tear said:
Early samples (*G43) aren't known for reliable operation
Click to expand...
Good to know.
 
From a forum member at the Hardware Canucks his name is Deadthings, he actually might even be a member at these forums also
 
Shadowmeph said:
I really need for someone to wlak me through the changing of the vcore
Click to expand...

0. Use Linux

1. Download and install TPC 0.44-rc1 per musky's guuide

2. Do not overclock, forget all your "OC commands", power-cycle the box so we start with clean sheet

3. Set up mcelog and verbose crash reporting as suggested before

4. First, record (and post here) your current settings: sudo TurionPowerControl -l

5. I'd expect PowerNow/CPB/Turbo to be typically enabled so initially do your testing having
  performed sudo TurionPowerControl -psmax 1 then verify with sudo TurionPowerControl -CM
  that all cores are exclusively in P-state 1

6. Perform your testing

7. If no joy -- we'll check the logs [EDIT] and move to vcore bump (while at stock speed)

EDIT:
Also, don't guess about memory. If you haven't done so -- memtest (as 402 already suggested). Shuffling good memory is a waste of time.

Yeah, I was referring to part number ending in G43. They are very YMMV thing.
 
Last edited:
tear said:
 
1. Verify that the PSU is adequate as fellow teammates already suggested :)

2. What is the part number of your CPUs? Early samples (*G43) aren't known for reliable operation.

3. You could try to speculatively raise the vcore using TPC (EDIT: make sure to use 0.44-rc2) and see
  what happens (also disable PowerNow/CPB/Turbo)

  (EDIT: and the opposite as well -- keep everything enabled and force CPUs to first bin turbo by means
   of TurionPowerControl -psmax 1 )

4. You could try to lower ambient temps and see how it affects the issue (esp. if it's as
  simple as opening the window)

5. If you're running (or can install) Linux, we could try to learn more about the failure.
  If you feel like spending time on it you could do... this:

  - install mcelog: sudo apt-get install mcelog (afterwards -- make sure it's running: ps auxw | grep [^]]mcelog)

  - enable verbose process crash reporting: http://hardforum.com/showpost.php?p=1039390775&postcount=264
   (not ultimately necessary but may come handy later)

  - set up and start folding

  - wait for the machine to crash (the longer time it takes to crash, the better chance
   of capturing useful information)

  - once machine reboots, examine output of dmesg and contents of /var/log/messages
   for information like: Machine Check Exceptions/HARDWARE ERRORS and RCU
   stalls

  - report your findings
Click to expand...

well bing fairly noobish to all of this I installed one es 6234 and then installed crunchbangx86_64 linux updated everything , downloaded TPC 0.44-rc2
extracted the folder, cd bin/Ubuntu-x86_64,
did the sudo modprobe msr
sudo modprobe cpuid ( I had to look at my old PMs on the HC forums./
sudo ./TurionPowerContol -l which gave me some numbers now I am not sure where to go from here.

I also installed mce log but when I did the ps auxw | grep [^]]mcelog command
I recieve a grep[^]]mcelog:command not found
also not sure of how to do this - enable verbose process crash reporting: http://hardforum.com/showpost.php?p=1039390775&postcount=264
 
Paste results of TurionPowerControl -l here.

There's a space after grep: ps auxw | grep [^]]mcelog
 
I only have one of the ES chips installed . also I didn't see the space after the drep lol ty
Code: 
-- Node: 1 Core 5
core 5 pstate 0 (pb0) - Boost PState Disabled
core 5 pstate 1 (pb1) - Boost PState Disabled
core 5 pstate 2 (p0) - En:1 VID:38 FID:4 DID:0.00 Freq:2000 VCore:1.0750
core 5 pstate 3 (p1) - En:1 VID:43 FID:2 DID:0.00 Freq:1800 VCore:1.0125
core 5 pstate 4 (p2) - En:1 VID:46 FID:14 DID:1.00 Freq:1500 VCore:0.9750
core 5 pstate 5 (p3) - En:1 VID:48 FID:8 DID:1.00 Freq:1200 VCore:0.9500
core 5 pstate 6 (p4) - En:1 VID:51 FID:4 DID:1.00 Freq:1000 VCore:0.9125

 --- Node 0:
Processor Maximum PState: 7
Processor Startup PState: 5
Processor Maximum Operating Frequency: No maximum defined. Unlocked multiplier.

Minimum allowed VID: 123 (0.0125V) - Maximum allowed VID 0 (1.5500V)
Processor AltVID: 51 (0.9125V)

 --- Node 1:
Processor Maximum PState: 7
Processor Startup PState: 5
Processor Maximum Operating Frequency: No maximum defined. Unlocked multiplier.
 
Did you get that after fresh boot? Looks like Turbo has been disabled. Go ahead and enable it in the BIOS then re-run the command please...

So, did ps | grep ... line return running mcelog?
 
yes mce is running , I am not sure where turbo is in my Bios
Whoa what happened here , I rebooted and am in my Bios and the time is 6 hrs ahead
 
Last edited:
Enable PowerNow, then you should see CPB (Core Performance Boosting) or Turbo option.
 
And no Turbo/CPB or anything? That's strange.


Ok, forget testing first turbo bin, then.
Can you just boot it for now and double check that pb0 and pb1 are still disabled? (TurionPowerControl -l)
 
core 5 pstate 0 (pb0) - Boost PState Disabled
core 5 pstate 1 (pb1) - Boost PState Disabled
I take it that the above means that they are disabled?
 
Yup. For some reason board isn't enabling Turbo. Weird. Let's ignore that for now.

To bump volts do this (paste the output too, please):
Code: 
sudo  ./TurionPowerControl  -fo  1
sudo  ./TurionPowerControl  -set  ps  2  vcore  1.1250
sudo  ./TurionPowerControl  -fo  0

Once done, run:
Code: 
sudo  ./TurionPowerControl  -l
to confirm 1.1250V on the CPU

And:
Code: 
sudo  ./TurionPowerControl  -CM
to confirm all cores are in P-state 2 (Ctrl+C to interrupt).

Once confirmed, run FAH and see what transpires :)
 
sudo ./TurionPowerControl -fo 1
PState set to 1
Done
sudo ./TurionPowerControl -set ps 2 vcore 1.1250
TurionPowerControl 0.44-rc2 (tpc-0.44-rc2-r144)
Turion Power States Optimization and Control - by blackshard

All nodes all cores pstate 2 - set core voltage to 1.1250
*** -set parsing completed

Done.
sudo ./TurionPowerControl -fo 0
TurionPowerControl 0.44-rc2 (tpc-0.44-rc2-r144)
Turion Power States Optimization and Control - by blackshard

PState set to 0

Done.

sudo ./TurionPowerControl -l

-- Node: 1 Core 5
core 5 pstate 0 (pb0) - Boost PState Disabled
core 5 pstate 1 (pb1) - Boost PState Disabled
core 5 pstate 2 (p0) - En:1 VID:34 FID:4 DID:0.00 Freq:2000 VCore:1.1250
core 5 pstate 3 (p1) - En:1 VID:43 FID:2 DID:0.00 Freq:1800 VCore:1.0125
core 5 pstate 4 (p2) - En:1 VID:46 FID:14 DID:1.00 Freq:1500 VCore:0.9750
core 5 pstate 5 (p3) - En:1 VID:48 FID:8 DID:1.00 Freq:1200 VCore:0.9500
core 5 pstate 6 (p4) - En:1 VID:51 FID:4 DID:1.00 Freq:1000 VCore:0.9125

--- Node 0:
Processor Maximum PState: 7
Processor Startup PState: 5
Processor Maximum Operating Frequency: No maximum defined. Unlocked multiplier.

Minimum allowed VID: 123 (0.0125V) - Maximum allowed VID 0 (1.5500V)
Processor AltVID: 51 (0.9125V)

--- Node 1:
Processor Maximum PState: 7
Processor Startup PState: 5
Processor Maximum Operating Frequency: No maximum defined. Unlocked multiplier.

Minimum allowed VID: 123 (0.0125V) - Maximum allowed VID 0 (1.5500V)
Processor AltVID: 51 (0.9125V)

sudo ./TurionPowerControl -CM
I am not sure what that out put means But I am going to paste it in here
Ts:1464908 -
Node 0 c0:ps6 - c1:ps2 - c2:ps2 - c3:ps2 - c4:ps6 - c5:ps6 - Tctl: 26
Node 1 c0:ps6 - c1:ps6 - c2:ps6 - c3:ps6 - c4:ps6 - c5:ps6 - Tctl: 21
Node0
C0: 0 0 166 1 3 3 363 C1: 8 0 208 1 3 2 322
C2: 8 0 341 1 1 2 167 C3: 32 0 386 1 1 2 123
C4: 31 0 290 0 2 3 246 C5: 3 0 313 0 2 4 222
Node1
C0: 3 0 188 0 2 0 353 C1: 1 0 189 0 2 0 353
C2: 0 0 66 0 2 1 474 C3: 1 0 166 0 2 1 374
C4: 1 0 71 0 2 0 467 C5: 4 0 82 0 2 0 456
MinTctl:18 MaxTctl:27
 
Oh, I see, PowerNow is enabled so that's why you're in low P-states (ps6 mostly).

I have just one more suggestion before starting FAH, then --
Code: 
sudo  -s
for  i  in  /sys/devices/system/cpu/cpu[0-9]*  ;  do  echo  performance  >  $i/cpufreq/scaling_governor  ;  done
exit

Then re-check sudo ./TurionPowerControl -CM. You should see ps2 across the board (instead of ps6 + fluctuations).

Once confirmed, launch FAH and see what happens.
 
ok now it is showing in node 0 and node 1 ps2
Ts:2274042 -
Node 0 c0:ps2 - c1:ps2 - c2:ps2 - c3:ps2 - c4:ps2 - c5:ps2 - Tctl: 20
Node 1 c0:ps2 - c1:ps2 - c2:ps2 - c3:ps2 - c4:ps2 - c5:ps2 - Tctl: 19
Node0
C0: 0 0 574 0 0 0 0 C1: 12 0 579 0 0 0 0
C2: 7 0 582 0 0 0 0 C3: 4 0 582 0 0 0 0
C4: 4 0 585 0 0 0 0 C5: 1 0 586 0 0 0 0
Node1
C0: 0 0 583 0 0 0 0 C1: 3 0 586 0 0 0 0
C2: 0 0 586 0 0 0 0 C3: 0 0 586 0 0 0 0
C4: 0 0 586 0 0 0 0 C5: 0 0 586 0 0 0 0
MinTctl:18 MaxTctl:21
this kind of thing also shows me how little I know lol
I am going to have to save all of these posts so I can keep looking at them to try to understand what is going onand why is it doing this lkol
 
Umm yes, but... hmm... there were cases of CPU going to ps0.... (see stats -- 12 here, 4 there etc.).
Very... strange.

Ok, let's do something beyond that:
Code: 
sudo  ./TurionPowerControl  -fo  1
sudo  ./TurionPowerControl  -boostdisable
sudo  ./TurionPowerControl  -fo  0
sudo  ./TurionPowerControl  -CM

This time you should only see non-zero numbers for ps2... I hope.
 
Code: 
shadowmeph@crunchbang:~/tpc/tpc/bin/Ubuntu-amd64$ sudo  ./TurionPowerControl  -fo  1
[sudo] password for shadowmeph: 
TurionPowerControl 0.44-rc2 (tpc-0.44-rc2-r144)
Turion Power States Optimization and Control - by blackshard

PState set to 1

Done.
shadowmeph@crunchbang:~/tpc/tpc/bin/Ubuntu-amd64$ sudo  ./TurionPowerControl  -boostdisable
TurionPowerControl 0.44-rc2 (tpc-0.44-rc2-r144)
Turion Power States Optimization and Control - by blackshard

Boost Lock Disabled.  Unlocked processor
Fid, Did, Vid, NodeTdp, NumBoostStates and CStateBoost can be edited
Boost disabled
APM disabled

Done.
shadowmeph@crunchbang:~/tpc/tpc/bin/Ubuntu-amd64$ sudo  ./TurionPowerControl  -fo  0
TurionPowerControl 0.44-rc2 (tpc-0.44-rc2-r144)
Turion Power States Optimization and Control - by blackshard

PState set to 0

Done
shadowmeph@crunchbang:~/tpc/tpc/bin/Ubuntu-amd64$ sudo  ./TurionPowerControl  -CM
Ts:3267596 - 
Node 0	c0:ps2 - c1:ps2 - c2:ps2 - c3:ps2 - c4:ps2 - c5:ps2 - Tctl: 28
Node 1	c0:ps2 - c1:ps2 - c2:ps2 - c3:ps2 - c4:ps2 - c5:ps2 - Tctl: 28
Node0
 C0:     0     0   588     0     0     0     0       C1:     2     0   588     0     0     0     0
 C2:     2     0   590     0     0     0     0       C3:     0     0   590     0     0     0     0
 C4:     0     0   590     0     0     0     0       C5:     0     0   590     0     0     0     0
Node1
 C0:     0     0   590     0     0     0     0       C1:     0     0   590     0     0     0     0
 C2:     0     0   573     0     0     0     0       C3:    17     0   573     0     0     0     0
 C4:    17     0   590     0     0     0     0       C5:     0     0   590     0     0     0     0
MinTctl:28	 MaxTctl:28

^Cshadowmeph@crunchbang:~/tpc/tpc/bin/Ubuntu-amd64$
 
Crazy.

Ok, let's test anyway. We can't do much more short of enabling Turbo from TPC (and I'd like to do that later).

Launch FAH, see what gives :-o
 
ok ty for all of your time tear :)
I just want to get tghis thing stable so that I can get some waterblocks OC this thing and see my fah number grow :)
 
Just crashed .
I was getting my hopes up then I checked the cpu usage and realized like an idiot I forgot to run ./fah6 -smp 12 for over 3 hrs
 
How much time did it take the machine to crash after starting fah6?

What does this return (type it in verbatim -- pretty important):
Code: 
sudo  egrep  -i  'machine.check|hardware.error'  /var/log/{messages,syslog}
 
I appogize for my delay right now it is 7:15 AM :)
ok with fah6 runing and stressing out the CPU it took about 45mins-1 hr.

AANd the output of sudo egrep -i 'machine.check|hardware.error' /var/log/{messages,syslog} =
Code: 
/var/log/messages:Jan 20 06:54:44 crunchbang kernel: [    3.255738] [Firmware Warn]: GHES: Poll interval is 0 for generic hardware error source: 1, disabled.
/var/log/messages:Jan 20 08:55:20 crunchbang kernel: [    2.723605] [Firmware Warn]: GHES: Poll interval is 0 for generic hardware error source: 1, disabled.
/var/log/messages:Jan 20 14:11:47 crunchbang kernel: [    2.739601] [Firmware Warn]: GHES: Poll interval is 0 for generic hardware error source: 1, disabled.
/var/log/messages:Jan 20 22:16:28 crunchbang kernel: [  299.816187] [Hardware Error]: Machine check events logged
/var/log/syslog:Jan 20 06:54:44 crunchbang kernel: [    3.255738] [Firmware Warn]: GHES: Poll interval is 0 for generic hardware error source: 1, disabled.
/var/log/syslog:Jan 20 08:55:20 crunchbang kernel: [    2.723605] [Firmware Warn]: GHES: Poll interval is 0 for generic hardware error source: 1, disabled.
/var/log/syslog:Jan 20 14:11:47 crunchbang kernel: [    2.739601] [Firmware Warn]: GHES: Poll interval is 0 for generic hardware error source: 1, disabled.
/var/log/syslog:Jan 20 22:16:28 crunchbang kernel: [  299.816187] [Hardware Error]: Machine check events logged
 
Status
Not open for further replies.
Back
Top