Last edited:
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Trying to tune 16 core 62xx ES
- Thread starter thinklet
- Start date
I see it, at 3.3GHz the TPF is 7:50, if scaling it to the "magic number" of 6:42 (TPF matches to 1M PPD) it would need at least 3.85GHz! 
So your previous plan of reaching the magic number seems to be a mission impossible.
So your previous plan of reaching the magic number seems to be a mission impossible.
With 3.6GHz & 8101 the TPF was down to 8:51 as shown in #110 . This is 1 sec better then my 4P E5-4650 that can manage 6:52 on 8102. If one could manage the 4P 6276 ES to 3.7GHz that will bring down TPF to around 8:35 for 8101. In my experience every 100MHz up make the TPF 15 - 18sec down for 8101. I beleive this indicate that this rig can run 8102 with TPF around 6:39 or better, at 3.7GHz?
18 (or 15)sec for 8101 is 15(or 12)sec for 8102 if my calculations is correct.
If I turn the fans on the radiators that cools the CPUs around to take air from outside I beleive they will do a better job cooling the chips. And I have read some place that this chips can take VCore as high as 1.5V voltage. If this is about right I think it can be done. What do you think, is it worth a try?
18 (or 15)sec for 8101 is 15(or 12)sec for 8102 if my calculations is correct.
If I turn the fans on the radiators that cools the CPUs around to take air from outside I beleive they will do a better job cooling the chips. And I have read some place that this chips can take VCore as high as 1.5V voltage. If this is about right I think it can be done. What do you think, is it worth a try?
I see it, at 3.3GHz the TPF is 7:50, if scaling it to the "magic number" of 6:42 (TPF matches to 1M PPD) it would need at least 3.85GHz!
So your previous plan of reaching the magic number seems to be a mission impossible.
Opteron 6200 is quite good at 8101 but less good at 8102, so your calculations based on 8101 might be inaccurate.
In my estimation, 3.7GHz is probably insufficient for you, unless you could shorten your 8102's TPF @3.3GHz from 7:50 to about 7:30. (You could try flashing memory sticks though, it might help)
BTW, the maximum vcore of 6200 ES chips is 1.55V, which could be found in tpc's output. You won't need it unless you want to try above 4GHz (with 2000W power draw).
In my estimation, 3.7GHz is probably insufficient for you, unless you could shorten your 8102's TPF @3.3GHz from 7:50 to about 7:30. (You could try flashing memory sticks though, it might help)
BTW, the maximum vcore of 6200 ES chips is 1.55V, which could be found in tpc's output. You won't need it unless you want to try above 4GHz (with 2000W power draw
).Thanks, I did not think of that, but can see now that you are right, it is not accurate to compare this rig with E5-4650 TPF for TPF. To make 3.85GHz on 6276 ES is maybe possible, but I do not think the motherboard will survive that much power.
Update on OC 6276ES with P8102
I am doing a test on 3.4GHz with P8102, and it seems that the scaling is very much better then we thought. TPF went from 7:51 down to 7:18. That is a improvement of 33 sec for 100MHz up. If this is the case for the next 200MHz the TPF for 3.6GHz will be 6:12. What do you think quickz? The power draw is now 944W.
This is crasy! The next TPF went down to 5:39 an so up to 7:20. What is going on in my rig?
I will update you with some pictures later today.
Some pictures:
Uploaded with ImageShack.us
If this is true for the next 200GHz up to 3.6GHz it will be made a record for 1 million on 1 client!
After the 1.13mill TPF the power draw went down to 925W. TPF is now 7:16 after 68%.
Update on OC 6276ES with P8102
I am doing a test on 3.4GHz with P8102, and it seems that the scaling is very much better then we thought. TPF went from 7:51 down to 7:18. That is a improvement of 33 sec for 100MHz up. If this is the case for the next 200MHz the TPF for 3.6GHz will be 6:12. What do you think quickz? The power draw is now 944W.
This is crasy! The next TPF went down to 5:39 an so up to 7:20. What is going on in my rig?
I will update you with some pictures later today.
Some pictures:
Uploaded with ImageShack.us
If this is true for the next 200GHz up to 3.6GHz it will be made a record for 1 million on 1 client!
Code:
vidar@G34Rigg3:~$ sudo TurionPowerControl -temp
Turion Power States Optimization and Control - by blackshard - v0.43
Detected processor: Family 15h (Bulldozer/Interlagos/Valencia) Processor
Machine has 8 nodes
Processor has 8 cores
Processor has 7 p-states
Processor has 2 boost states
Processor temperature slew rate:9.0°C
Temperature table:
Node 0 C0:36 C1:36 C2:36 C3:36 C4:36 C5:36 C6:36 C7:36
Node 1 C0:35 C1:35 C2:35 C3:35 C4:35 C5:35 C6:35 C7:35
Node 2 C0:49 C1:49 C2:49 C3:49 C4:49 C5:49 C6:49 C7:49
Node 3 C0:39 C1:39 C2:39 C3:39 C4:39 C5:39 C6:39 C7:39
Node 4 C0:37 C1:37 C2:37 C3:37 C4:37 C5:37 C6:37 C7:37
Node 5 C0:37 C1:37 C2:37 C3:37 C4:37 C5:37 C6:37 C7:37
Node 6 C0:36 C1:36 C2:36 C3:36 C4:36 C5:36 C6:36 C7:36
Node 7 C0:36 C1:36 C2:36 C3:36 C4:36 C5:36 C6:36 C7:36
Done.
vidar@G34Rigg3:~$ sudo TurionPowerControl -l | grep "pstate 2"
core 0 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 1 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 2 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 3 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 4 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 5 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 6 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 7 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 0 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 1 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 2 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 3 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 4 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 5 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 6 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 7 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 0 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 1 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 2 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 3 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 4 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 5 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 6 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 7 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 0 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 1 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 2 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 3 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 4 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 5 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 6 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 7 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 0 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 1 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 2 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 3 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 4 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 5 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 6 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 7 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 0 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 1 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 2 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 3 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 4 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 5 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 6 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 7 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 0 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 1 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 2 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 3 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 4 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 5 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 6 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 7 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 0 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 1 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 2 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 3 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 4 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 5 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 6 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
core 7 pstate 2 (p0) - En:1 VID:28 FID:18 DID:0.00 Freq:3400 VCore:1.2000
vidar@G34Rigg3:~$
vidar@G34Rigg3:~$ sudo ht-retries.sh
L0S0 L1S0 L2S0 L3S0 L0S1 L1S1 L2S1 L3S1
Node 0 0000 0000 0000 0000 0000 0000 0000 0000
Node 1 0000 0000 0000 0000 0000 0000 0000 0000
Node 2 0000 0000 0000 0000 0000 0000 0000 0000
Node 3 0000 0000 0000 0000 0000 0000 0000 0000
Node 4 0000 0000 0000 0000 0000 0000 0000 0000
Node 5 0000 0000 0000 0000 0000 0000 0000 0000
Node 6 0000 0000 0000 0000 0000 0000 0000 0000
Node 7 0000 0000 0000 0000 0000 0000 0000 0000
vidar@G34Rigg3:~$
Last edited:
Update:
I will be away from late this evening until sunday afternoon. I'm leaving on a sea voyage along the seaside of Norway. I will anyway be online from time to time to keep myself updated in this tread, and monitor the machines back home. My son will stay home and is able to restart if any of them should go down. This is incredibly exciting, so I'm looking forward to being back for further testing late Sunday.
I will be away from late this evening until sunday afternoon. I'm leaving on a sea voyage along the seaside of Norway. I will anyway be online from time to time to keep myself updated in this tread, and monitor the machines back home. My son will stay home and is able to restart if any of them should go down. This is incredibly exciting, so I'm looking forward to being back for further testing late Sunday.
Last edited:
402blownstroker
[H]ard|DCer of the Month - Nov. 2012
- Joined
- Jan 5, 2006
- Messages
- 3,242
Subscribing and re-reading thread
It's interesting to see that your performance/frequency at 3.4GHz was significantly better than 3.3GHz. Did you get DLB enabled for both 3.3 and 3.4GHz? As for your crazy TPF of 5:39, I cant find a reasonable explanation. I never met it on my rigs.
And yes, if you could keep your current folding efficiency to 3.7GHz (not 3.6GHz), you would probably see the magic number of 1M ppd.
A simple math:
Now your TPF at 3.4GHz is 7:18 (=438s), while your goal is 6:42 (=402s).
So with a linear scaling you would need 3.4*(438/402)=3.704GHz to realize it.
And yes, if you could keep your current folding efficiency to 3.7GHz (not 3.6GHz), you would probably see the magic number of 1M ppd.
A simple math:
Now your TPF at 3.4GHz is 7:18 (=438s), while your goal is 6:42 (=402s).
So with a linear scaling you would need 3.4*(438/402)=3.704GHz to realize it.
Dumb question... I did sudo:
TurionPowerControl -set ps 2 vcore 1.075 freq 2600
TurionPowerControl -set ps 1 vcore 1.075 freq 2600
TurionPowerControl -set ps 0 vcore 1.075 freq 2600
TurionPowerControl -fo 0
Want to see actual speed of the chips. Anyway to do this easily? I did sudo TurionPowerControl -l | grep "pstate 2" (and other pstates)... and I know TurionPowerControl -CM... but how can I verify actual clock speed other than cross referencing and hoping?
TurionPowerControl -set ps 2 vcore 1.075 freq 2600
TurionPowerControl -set ps 1 vcore 1.075 freq 2600
TurionPowerControl -set ps 0 vcore 1.075 freq 2600
TurionPowerControl -fo 0
Want to see actual speed of the chips. Anyway to do this easily? I did sudo TurionPowerControl -l | grep "pstate 2" (and other pstates)... and I know TurionPowerControl -CM... but how can I verify actual clock speed other than cross referencing and hoping?
This is a feature of newer BIOSes so it may or may not be present in yours.
It offers no advantages to FAH (removes "lowest" boosted P-state IIRC) so it's
better to leave it disabled which, at least on Supermicro boards, is the default.
Voltage regulation on GL is superior to any other 4P G34 board. Not sure if efficiency
follows but I would expect that it does.
I'd say this is pretty good. I'm 10:25-10:40 at 3.0 and 1600/6-8-6.
Now running 3.3GHz, TPF = 9:36, VCore = 1.1750V, Power Draw = 880W
Update: Now running 3.4GHz, TPF = 9:29, VCore = 1.2000V, PD = 950W. I have expected more increase in TPF from +100MHz. TPF down to 9:24 after 5 - 6%. After 63% is TPF down to 9:20.
I will switch clocksource=hpet on next time I reboot.
Update: Where do I "switch to clocksource=hpet"?
Update: Now running 3.4GHz, TPF = 9:29, VCore = 1.2000V, PD = 950W. I have expected more increase in TPF from +100MHz. TPF down to 9:24 after 5 - 6%. After 63% is TPF down to 9:20.
I will switch clocksource=hpet on next time I reboot.
Update: Where do I "switch to clocksource=hpet"?
Last edited:
- Set your clocksource to hpet - this method of overclocking does not actually mess with the system clock, but there are other things that work better with the clocksource set to hpet.
Instructions for Ubuntu Desktop:
Instructions for Ubuntu Desktop:
Code:
1. sudo nano /etc/default/grub
2. Find this line: GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"
3. Change it to: GRUB_CMDLINE_LINUX_DEFAULT="quiet splash clocksource=hpet"
4. Save your changes - Ctrl-X - Y - Enter
5. sudo update-grub
6. Restart the machineInteresting note... 1.100v @ 2.8GHz TSC = 2.8GHz clock staying steady at 66C. When I go to 2.9GHz it drops the clock to 1.4GHz... hmmm...
402blownstroker
[H]ard|DCer of the Month - Nov. 2012
- Joined
- Jan 5, 2006
- Messages
- 3,242
OK, I have a set of 62xx 16core chips showing up tomorrow. Could of quick questions to get started with:
1) ONG or stock BIOS?
2) Turn off PowerNow/CPB/Turbo in the BIOS right?
3) If the ONG BIOS is not used, how can the memory timings be tweaked?
4) Anyone have rough voltage to frequency mappings?
1) ONG or stock BIOS?
2) Turn off PowerNow/CPB/Turbo in the BIOS right?
3) If the ONG BIOS is not used, how can the memory timings be tweaked?
4) Anyone have rough voltage to frequency mappings?
Can only answer part of your question. I have done some notes of my own testing of my 4P SM H8QGI-F & 4 x Opteron 6276es. Memeory is 16 x 4GB Corsair Vengeance® Low Profile Blue — 4GB Dual Channel DDR3 Memory Kit (CML4GX3M2A1600C9B) The memory is working only at 1333MHz.
Testing OC for 4P Opteron 6276es pr. 21.11.2012
P8101 3.0GHz VCore 1.1000V TPF = 10:32 PPD = 508K Power Draw = 780W stable
P8101 3.1GHz VCore 1.15000V TPF = 10:37 PPD = 502K Power Draw = 835W stable
P8101 3.2GHz VCore 1.16125V TPF = 10:10 PPD = 535K Power Draw = 855W stable
I can not explain the big different in TPF between 3.2 - 3.3GHz!
P8101 3.3GHz VCore 1.1750V TPF = 9:36 PPD = 583K Power Draw = 880W stable
Do not overclock these cards longer than this without heatsinks on all components, the reason why not,
you can read about here: #162
P8101 3.4GHz VCore 1.2000V TPF = 9:20 PPD = 609K Power Draw = 950W stable
P8101 3.5GHz VCore 1.2500V TPF = 9:04 PPD = 635K Power Draw = 990W stable
P8101 3.6GHz VCore 1.2750V TPF = 8:48 PPD = 665K Power Draw = 1035W stable
P8101 3.7GHz VCore 1.3000V TPF = 8:xx PPD = xxxK Power Draw = 1150W unstable
P8102 3.4GHz VCore 1.2000V TPF = 7:18 PPD = 880K Power Draw = 905W stable
P6958 3.6GHz VCore 1.2750V TPF = 36sec PPD = 237K Power Draw = 940W stable
In my experience the Power DRaw vary from reboot to another reboot on the same WU, and from WU to WU!
Update: Added 3.0 - 3.2GHz values.
Update 29.12.2012, 3.0GHz values.
Testing OC for 4P Opteron 6276es pr. 21.11.2012
P8101 3.0GHz VCore 1.1000V TPF = 10:32 PPD = 508K Power Draw = 780W stable
P8101 3.1GHz VCore 1.15000V TPF = 10:37 PPD = 502K Power Draw = 835W stable
P8101 3.2GHz VCore 1.16125V TPF = 10:10 PPD = 535K Power Draw = 855W stable
I can not explain the big different in TPF between 3.2 - 3.3GHz!
P8101 3.3GHz VCore 1.1750V TPF = 9:36 PPD = 583K Power Draw = 880W stable
Do not overclock these cards longer than this without heatsinks on all components, the reason why not,
you can read about here: #162
P8101 3.4GHz VCore 1.2000V TPF = 9:20 PPD = 609K Power Draw = 950W stable
P8101 3.5GHz VCore 1.2500V TPF = 9:04 PPD = 635K Power Draw = 990W stable
P8101 3.6GHz VCore 1.2750V TPF = 8:48 PPD = 665K Power Draw = 1035W stable
P8101 3.7GHz VCore 1.3000V TPF = 8:xx PPD = xxxK Power Draw = 1150W unstable
P8102 3.4GHz VCore 1.2000V TPF = 7:18 PPD = 880K Power Draw = 905W stable
P6958 3.6GHz VCore 1.2750V TPF = 36sec PPD = 237K Power Draw = 940W stable
In my experience the Power DRaw vary from reboot to another reboot on the same WU, and from WU to WU!
Update: Added 3.0 - 3.2GHz values.
Update 29.12.2012, 3.0GHz values.
Last edited:
So what I did was
sudo TurionPowerControl -set ps 0 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 1 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 2 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 3 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 4 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 5 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 6 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 7 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 8 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 9 vcore 1.075 freq 2800
Seems like that keeps temps below 70C even using Dynatron A1's with silent fans (I do need to swap back to muskymod fans)
sudo TurionPowerControl -set ps 0 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 1 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 2 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 3 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 4 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 5 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 6 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 7 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 8 vcore 1.075 freq 2800
sudo TurionPowerControl -set ps 9 vcore 1.075 freq 2800
Seems like that keeps temps below 70C even using Dynatron A1's with silent fans (I do need to swap back to muskymod fans)
That's excessive. Read this to get general idea on what to do: http://hardforum.com/showpost.php?p=1039116320&postcount=7
CPUs slowing down could also be related to overheating. Examine output of (sudo) TurionPowerControl -htc to verify.
402,
(shall things go south). Details: http://www.amdzone.com/phpbb3/viewtopic.php?f=521&t=138490&start=225#p218762
General flow is:
1. Read SPDs
2. Modify SPDs
3. Flash SPDs
Let know if you have further Qs
Be aware that this may not fly with typical chips. Some of mine needed that voltage to do mere 3.0...
Latest stock BIOS. No OCNG for these yet...
That is correct.
The only way is flashing SPDs. Make sure to try it with one bank (CPU) first not to hose whole system
(shall things go south). Details: http://www.amdzone.com/phpbb3/viewtopic.php?f=521&t=138490&start=225#p218762
General flow is:
1. Read SPDs
2. Modify SPDs
3. Flash SPDs
Let know if you have further Qs
402blownstroker
[H]ard|DCer of the Month - Nov. 2012
- Joined
- Jan 5, 2006
- Messages
- 3,242
Let know if you have further Qs
Did you really need to say that
Hopefully these chips will have a nice improvement over the 6166HE chips and lessen my poor attitude.
Dead Things
Limp Gawd
- Joined
- Jul 8, 2011
- Messages
- 152
You can set freq to whatever value you like, but it will be rounded to the nearest 100 MHz step because the multiplier can only change by increments of 0.5.
402blownstroker
[H]ard|DCer of the Month - Nov. 2012
- Joined
- Jan 5, 2006
- Messages
- 3,242
If some really smart guy could modify the BIOS to set the BLCK rate, then that would be possible
Dead Things
Limp Gawd
- Joined
- Jul 8, 2011
- Messages
- 152
I am anxiously looking forward to "mid-December"!If some really smart guy could modify the BIOS to set the BLCK rate, then that would be possible
According to tpc's document:
So a step of 50MHz is possible if you are using "the hard way" for frequency tuning and setting DID to 1.
However, the highest frequency we could get is limited by the maximum value of FID. For ES 6100s it's only 31 and we could only get at most 2350MHz in this way, don't know if ES 6200s allow higher FIDs.
Code:
FREQ=(100*(FID+Const.))/(2^DID)However, the highest frequency we could get is limited by the maximum value of FID. For ES 6100s it's only 31 and we could only get at most 2350MHz in this way, don't know if ES 6200s allow higher FIDs.
I have tested with several DDR3 1600 DIMMs and only found that G.Skill 12800CL6 could work at 1600MHz (but CL9) in default.
You might have to flash SPDs (with tear's tools) to get your memory work at its advertised frequency and CL timing.
You might have to flash SPDs (with tear's tools) to get your memory work at its advertised frequency and CL timing.
Thanks quickz
On my Tyan S8812 I use Kingston HyperX http://www.kingston.com/dataSheets/KHX1866C11D3P1K2_4G.pdf & 4 X 6272 Retail chips, and I am able to run the memory at 1600MHz without any adjustment. Maybe I should try this memory on the SM board & 6276 ES chips. Do you think that they also will work on the SM H8QGI-F board in 1600MHz by default?
I could of course also try tear's tool and try to flash the Corsair memory, but I am afraid I could damage the memory sticks, is that possible? It looks quite difficult to flash the memory? Could I do it on another computer, or must it be done on the mobo it would be used? And finally, do I have to flash all 16 sticks at the same time?
On my Tyan S8812 I use Kingston HyperX http://www.kingston.com/dataSheets/KHX1866C11D3P1K2_4G.pdf & 4 X 6272 Retail chips, and I am able to run the memory at 1600MHz without any adjustment. Maybe I should try this memory on the SM board & 6276 ES chips. Do you think that they also will work on the SM H8QGI-F board in 1600MHz by default?
I could of course also try tear's tool and try to flash the Corsair memory, but I am afraid I could damage the memory sticks, is that possible? It looks quite difficult to flash the memory? Could I do it on another computer, or must it be done on the mobo it would be used? And finally, do I have to flash all 16 sticks at the same time?
This one might be able to run at 1600MHz on SM boards as well, since it has the JEDEC profile of DDR3-1600 @1.5V.
As for flashing memory SPDs, you could do it on another computer, and you don't need to flash all sticks at the same time.
However, flashing SPDs would only improve the performance very slightly (usually on the order of 1%). I don't think you need to do it right now if you feel it's difficult.
And did you find any reason for the strange temperature of node2? You could try to reinstall the heatsink, exchange CPU positions, or buy another CPU to replace it.
As for flashing memory SPDs, you could do it on another computer, and you don't need to flash all sticks at the same time.
However, flashing SPDs would only improve the performance very slightly (usually on the order of 1%). I don't think you need to do it right now if you feel it's difficult.
And did you find any reason for the strange temperature of node2? You could try to reinstall the heatsink, exchange CPU positions, or buy another CPU to replace it.
Thanks for reasonable answer´s.
As for the special temperature on node2, I have not done anything yet. I guess this is CPU socket2, or am I mistaken?
As for the special temperature on node2, I have not done anything yet. I guess this is CPU socket2, or am I mistaken?
And did you find any reason for the strange temperature of node2? You could try to reinstall the heatsink, exchange CPU positions, or buy another CPU to replace it.
Yes, but be careful, the real CPU position might be not the same as that shown in the manual.
I found this odd issue on my H8QG6 board:
When I installed 4 Opteron 6100 chips on the board, the positions of these CPUs are:
which are the same as that shown in the manual.
However, when I installed 4 Opteron 6200 chips on the same board, the positions of these CPUs are proven to be:
Very interesting, you could see CPU2 and CPU3 exchanged their positions!
I found this odd issue on my H8QG6 board:
When I installed 4 Opteron 6100 chips on the board, the positions of these CPUs are:
Code:
CPU1 CPU3
CPU2 CPU4However, when I installed 4 Opteron 6200 chips on the same board, the positions of these CPUs are proven to be:
Code:
CPU1 CPU2
CPU3 CPU4More than that -- http://hardforum.com/showthread.php?t=1647656
Thanks!
This is the same as for mine as for the SuperMicro H8QGi+-F in http://hardforum.com/showthread.php?t=1647656
There is a connector for connection of the power for the cpu fan near each CPU, there it is written a CPU socket number, that also correspond with the numbers that tear wrote.
I have also measured the temperature of each CPU manually with a digital thermometer, and it varies only by 2 degrees between CPUs. I think that there must be something wrong with the sensors on the system board that shows a difference as 12 degrees. I can also by touching each water cooler block feel that temperatures are approximately equal for all the water blocks.
This is the same as for mine as for the SuperMicro H8QGi+-F in http://hardforum.com/showthread.php?t=1647656
There is a connector for connection of the power for the cpu fan near each CPU, there it is written a CPU socket number, that also correspond with the numbers that tear wrote.
I have also measured the temperature of each CPU manually with a digital thermometer, and it varies only by 2 degrees between CPUs. I think that there must be something wrong with the sensors on the system board that shows a difference as 12 degrees. I can also by touching each water cooler block feel that temperatures are approximately equal for all the water blocks.
Yes, I did the test last week on a H8QG6 board.
My method is to stop the CPU fans in turn for a short time, and watching the CPU temperature change by 'tpc -mtemp'. The mapping for Opteron 6100s is the same as the manual, but the mapping for Opteron 6200s is found to be different.
My method is to stop the CPU fans in turn for a short time, and watching the CPU temperature change by 'tpc -mtemp'. The mapping for Opteron 6100s is the same as the manual, but the mapping for Opteron 6200s is found to be different.