Improve Opteron 6200's performance (and PPD) by using TurionPowerControl

Well I found the problem with my 2P 6274 box - wasn't quite what I expected:

molex-eps-sm.jpg


I guess the extra power draw was too much for the molex to 8pin eps. I replaced it with a dual molex to 8pin eps for the time being. But probably won't be messing with the pstate change again until I get a pci-e to 8-pin eps adapter. I'm just thankful all the hardware still works.
 
Last edited:
There has been speculation that 8101 favours the faster memory thats available with the 62xx chips but it was only talked about for a couple of days and i haven't seen any more on it for a while now.

My 6272 4P rig has 32GB of Kingston's KHX1866C11S3P1K2/4G installed running at 1600MHz, but seems to be slower than quickz
4P 6272 box. He has a TPF of 12:48 while my at the moment is doing 13:00 as fastest on 8101 (first WU after update with TPC.) I do not know the type of memory he have.

I belive my memory has timings as CL 9 - 9 - 9, but can have 10 -10 -10 when one running 1600MHz. I am not sure if I can adjust it as so on my Tyan S1288 mobo? Link to my memory: http://www.valueram.com/datasheets/KHX1866C11D3P1K2_4G.pdf

Update after reset: I can not remember I have done it, but the timings was already set to CL 10- 10 - 10 - 28
 
Last edited:
My RAM is not fast and it only runs at a standard 1333MHz CL 9-9-9-24, which is confirmed by 'tpc -dram'.
 
This has been working perfectly since I was able to get a pci-e to eps adapter to use on the 2nd eps connector.
 
:)Hi, I have acquired a pair of 6274 retail chips, the question is should I run these on an Asus KGPE-D16 or as 2 CPU's on a 4P SMH8QGi+-F? With both boards running standard Bios, will they take advantage of the TPC talked about inn the thread? All of my other motherboards are currently SM. regards, Charlie
 
it'll only work on the standard bios because the overclock modified bios only works for magny cour processors. but if you plan to go 4p 6200's i'd go with the 4p board even if you are only running 2 processors.
 
:)Hi, I have acquired a pair of 6274 retail chips, the question is should I run these on an Asus KGPE-D16 or as 2 CPU's on a 4P SMH8QGi+-F? With both boards running standard Bios, will they take advantage of the TPC talked about inn the thread? All of my other motherboards are currently SM. regards, Charlie

If you use the SM H8QGi+-F the TPC will work, and I assume it will work with 2 CPUs as it does with four.
 
TPC acceleration for Opteron 6200 should work on any G34 mobo, no matter 2P or 4P board it is.
However, I'm not sure if it would work for Opteron 4200 and other kind of Bulldozer chips, could anyone perform a test to confirm it?
 
Acording to TurionPowerControl, the Opteron 6272 CPU has 7 pstates:
core 0 pstate 0 (pb0) - En:1 VID:19 FID:14 DID:0.00 Freq:3000 VCore:1.3125
core 0 pstate 1 (pb1) - En:1 VID:35 FID:8 DID:0.00 Freq:2400 VCore:1.1125
core 0 pstate 2 (p0) - En:1 VID:41 FID:5 DID:0.00 Freq:2100 VCore:1.0375
core 0 pstate 3 (p1) - En:0 VID:41 FID:5 DID:0.00 Freq:2100 VCore:1.0375
core 0 pstate 4 (p2) - En:0 VID:41 FID:5 DID:0.00 Freq:2100 VCore:1.0375
core 0 pstate 5 (p3) - En:0 VID:41 FID:5 DID:0.00 Freq:2100 VCore:1.0375
core 0 pstate 6 (p4) - En:1 VID:49 FID:12 DID:1.00 Freq:1400 VCore:0.9375

But it seems that '-psmax 0' is invalid, after such a setting I find following infomation in the result of 'TurionPowerControl -l':

Processor Maximum PState: unsupported feature
Processor Startup PState: 6
Processor Maximum Operating Frequency: 3000 MHz

Since the Processor Startup PState is shown as 6 (for my 6272), I think the default setting might be 'TurionPowerControl -psmax 6'.

Hi quickz,

Can you tell me where I can find this data? I am interested in the 6274 retail and ES variants. Best regards, Charlie
 
Wanted to comment on something I'm pretty sure I've found.

I manually enter the tpc commands when I reboot (don't do it very often).

A few days ago I noticed that I was getting consistently lower tpf's (average on 8101 had dropped 20 seconds below my long term running average).

I remembered that when I had rebooted last week, I made a typo when I used tpc -set, instead of tpc -set ps 1, i had typed tpc set -ps1, without the space it programs ps0. I noticed it, and didn't bother to reset ps0 since it wouldn't ever get to that bin when folding (I correctly reset ps 1).

I wondered though if perhaps it does, occasionally, so I reset ps0 a few days ago(using less voltage than the default, something I'd never bothered with before figuring it was wasted effort). Low and behold TPF's are back down, and load power has also dropped slightly.

I say I'm pretty sure I've found this as my sample since resetting ps0 is now up to a n of 10, and the difference is statistically significant, but won't be totally confident for a while.

If this continues to hold, it seems the cpu is occasionally, very briefly (i sat staring at tpc -CM for a bit and never saw a zero) entering ps0, enough to have a slight impact on TPF.
 
Hmm, did you mean -psmax0 there? -set ps0 (by itself) only selects a pstate to operate on
-- it needs to be followed by something that actually alters pstate settings, so... <confused/>

I'm about to release 0.44-rc2 which has improved command line parsing so now we get something along these:
Code:
$ sudo ./TurionPowerControl -set ps0
TurionPowerControl 0.44-rc2 (trunk-r144)
Turion Power States Optimization and Control - by blackshard

ERROR: unknown set sub-command -- ps0

$ sudo ./TurionPowerControl -psmax0
TurionPowerControl 0.44-rc2 (trunk-r144)
Turion Power States Optimization and Control - by blackshard

ERROR: invalid argument -- -psmax0
$
 
sorry, truncated rest of line, i.e...

sudo tpc -set ps 1 vcore 1.150 freq 3000

my typo was...

sudo tpc -set ps1 vcore 1.150 freq 3000

which sets ps0 to 1.150/3000
 
I see, so that's just like not specifying it at all (it defaults to P-state 0).

So... basically your P-states 0 and 1 have now identical settings? (vide: TurionPowerControl -l | grep 0.pstate.[01] | head -2)
 
I see, so that's just like not specifying it at all (it defaults to P-state 0).

So... basically your P-states 0 and 1 have now identical settings? (vide: TurionPowerControl -l | grep 0.pstate.[01] | head -2)

They did when it was running slower. I reset ps0 to it's default freq (3300), with more voltage then my ps1 setting, but less than the default ps0 voltage, and now it appears as if it is back to running as before (faster), but with slightly less power draw.
 
I see. Interesting!

If you run sudo TurionPowerControl -CM for >30 seconds, I take it that the only ps 0 and ps 1 are reported in the stats? (needs tpc 0.44-rc1)
 
I see. Interesting!

If you run sudo TurionPowerControl -CM for >30 seconds, I take it that the only ps 0 and ps 1 are reported in the stats? (needs tpc 0.44-rc1)

I let it run for a couple of minutes while folding and never saw a zero (all 1's), which is part of why I'm not totally convinced yet (even though statistically, the odds of the improved tpf being chance are pretty small already...)
 
Thanks a lot for this tip. I purchased two opteron 6272 CPUs last year and have had them working at 2,1 GHz all the time. I downloaded TurionPowerControl and executed the command -psmax 1. Now my CPUs are working at 2,4 GHz. I have made a benchmark test and found it gives a 10% performance gain over the default configuration.
 
Hey All..

I am currently trying to overclock 2 AMD 6272 16 cores.

I can use the Turion software to place the chips running at 2.4ghz, but can't seem to find out how to lock them at 3.. if I try to change the state to 0 it doesn't do anything..

Would someone be able to give me a quick heads up if there is something I'm missing? How do you set a chip at turbo speed and lock it there.. or is this even possible?
 
"psmax" trick works only for P-state 1 (all-core turbo bin) -- in your case -- 2.4 GHz.

Unfortunately, no method of forcing P-state 0 ("some cores" turbo bin -- 3.0 GHz in your case)
in production chips has been identified...
 
"psmax" trick works only for P-state 1 (all-core turbo bin) -- in your case -- 2.4 GHz.

Unfortunately, no method of forcing P-state 0 ("some cores" turbo bin -- 3.0 GHz in your case)
in production chips has been identified...

I see.. ok well thank you for the information anyway :)
 
Hey All..

I am currently trying to overclock 2 AMD 6272 16 cores.

I can use the Turion software to place the chips running at 2.4ghz, but can't seem to find out how to lock them at 3.. if I try to change the state to 0 it doesn't do anything..

Would someone be able to give me a quick heads up if there is something I'm missing? How do you set a chip at turbo speed and lock it there.. or is this even possible?

Not sure which board you're using, but I had to leave Boost (CPB) enabled in BIOS and disable via tpc in order to overclock my 6282 extra spicy chips on the H8QGL board. Did you leave Boost enabled? Have you tried overclocking your 6272's beyond the stock all-core turbo of 2.4GHz?

Here's my example "run-tpc" script:
Code:
#!/bin/sh
if [ "`id -u`" != 0 ]; then
  echo "You must run this script with sudo"
  exit 1
fi
TPC=TurionPowerControl
FREQ=3400	#works great at 1.1875
VCORE=1.1875
#  $TPC -nbfid 7 for 2200MHz, then soft reboot
#  $TPC -nbfid 8 for 2400MHz, then soft reboot, only 40GB RAM seen
$TPC -boostdisable
$TPC -fo 3
$TPC -set ps 2 vcore $VCORE freq $FREQ
$TPC -set ps 1 vcore $VCORE freq $FREQ
$TPC -set ps 0 vcore $VCORE freq $FREQ
sleep 1
$TPC -fo 0

My old post about this is here.
 
Dear tear or musky
You guys helped me earlier with this problem, can you please do it again?

I had to reinstall Ubuntu 12.04 because I made a mistake. This is a 4P rig with one Tyan S8812 motherboard and factory produced Opteron 6272 chips. The problem now is that I do not get TurionPowerControl installed correctly. I come this far with the installation, then comes the error message for this line:

vidar@G346272:~/tpc-0.44-rc2/src$ sudo make install
install -ps TurionPowerControl /usr/bin
ln -sf TurionPowerControl /usr/bin/tpc
ln: «/usr/bin/tpc/TurionPowerControl»: Can not overwrite directory
make: *** [install] Error 1
vidar@G346272:~/tpc-0.44-rc2/src$

I thought that when I used "sudo make install" so I should not get such messages, so what am I doing wrong? I have used this guide: http://hardforum.com/showthread.php?t=1669698

Edit: I managed to find the solution, so TPC is working again!
 
Last edited:
i've probably missed something important so apologies if i have but im looking into this as i may be getting a G34 setup soon cheap (pair of 6272's along with a supermicro H8DG6-F and 16gb ecc ram at 1600MHz for like £140) but from what i see there is a fair bit of linux stuff in here, does this 'overclock' (or constant turbo or whatever it's to be called) work within windows? if so then ill possibly look into it if i have the power to do so but if not then im stuck at stock till i can source some other higher clocked opterons or some ES ones

sorry to have dug up an old post too XD
 
Last edited:
i've probably missed something important so apologies if i have but im looking into this as i may be getting a G34 setup soon cheap (pair of 6272's along with a supermicro H8DG6-F and 16gb ecc ram at 1600MHz for like £140) but from what i see there is a fair bit of linux stuff in here, does this 'overclock' (or constant turbo or whatever it's to be called) work within windows? if so then ill possibly look into it if i have the power to do so but if not then im stuck at stock till i can source some other higher clocked opterons or some ES ones

sorry to have dug up an old post too XD
It will not work in Windows, only within Linux.
 
It will not work in Windows, only within Linux.

ah right, thanks for the reply, i had my hopes up too much it seems XD looks like ill be scouring the net for ES or higher clocked opterons when i have the money then :p

thanks anyways though :)
 
Windows binaries of TPC work just fine.

It *does* work in Windows with no issues whatsoever.

tpc-2012r2.png

how'd you get that tear? if i can get the constant all core boost then that'll help lots XD ill have to be making notes since ill have a while before i can afford my optys but still, be nice to know, many thanks in advanced
 
Like OP explains: TurionPowerControl.exe -psmax 1

Just make sure to run it from admin cmd.exe window.
 
Like OP explains: TurionPowerControl.exe -psmax 1

Just make sure to run it from admin cmd.exe window.

ill need to experiment with this a bit, thanks, ill attempt it with my 8320 for now since that's all i have for the minute but if this works then that's grand
 
I know this is a necro, but I need some help with this.

I've got a Dell quad G34 system with 6272s. They're running at 2100Mhz I believe and I've installed TPC and ran the command. I got the "done." after running the command, but when I run lscpu the CPUs appear to still be running at 2100Mhz. Is there something else I need to do?
 
I know this is a necro, but I need some help with this.

I've got a Dell quad G34 system with 6272s. They're running at 2100Mhz I believe and I've installed TPC and ran the command. I got the "done." after running the command, but when I run lscpu the CPUs appear to still be running at 2100Mhz. Is there something else I need to do?

This command isn't changing the clock speed of the CPUs, but rather changing the power profile that they're operating in. PSMAX = 1 basically runs them full tilt under load, whereas in the default configuration, there may be some power/thermal regulation that holds back performance.
 
I thought it forced the CPUs to run at their turbo speed which for the CPUs I have should be 2.4Ghz. Right?
 
Hey, Skillz. I was unaware that Turion Power Control worked with motherboards other than selected Supermicro G34 4Ps. When we used TPC with the Supermicros, we also installed Tear's custom motherboard BIOS.
 
Hey, Skillz. I was unaware that Turion Power Control worked with motherboards other than selected Supermicro G34 4Ps. When we used TPC with the Supermicros, we also installed Tear's custom motherboard BIOS.
^^^^^^^^ This. I don't believe you're getting any clock enhancement using TPC on a Dell MB. And the MBs that tear worked on are very specific and unlikely to port.
 
hi folks,

i've been playing around with some Opteron 6328s as these have the highest frequency for the G34 6300 series.

The -psmax 1 trick does help.

i've not been able to get it into pstate 0 however to hit 3.8 GHz.

When the OS is handling this it occasionally does hit 3.8 however.

Any tips here?

Power States table:
-- Node: 0 Core 0
core 0 pstate 0 (pb0) - En:1 VID:19 FID:22 DID:0.00 Freq:3800 VCore:1.3125
core 0 pstate 1 (pb1) - En:1 VID:27 FID:19 DID:0.00 Freq:3500 VCore:1.2125
core 0 pstate 2 (p0) - En:1 VID:35 FID:16 DID:0.00 Freq:3200 VCore:1.1125
core 0 pstate 3 (p1) - En:1 VID:38 FID:12 DID:0.00 Freq:2800 VCore:1.0750
core 0 pstate 4 (p2) - En:1 VID:43 FID:7 DID:0.00 Freq:2300 VCore:1.0125
core 0 pstate 5 (p3) - En:1 VID:46 FID:3 DID:0.00 Freq:1900 VCore:0.9750
core 0 pstate 6 (p4) - En:1 VID:51 FID:12 DID:1.00 Freq:1400 VCore:0.9125
 
Back
Top