F5F Stay Refreshed Power Users Overclocking Talk about DiscussionGTX 680 Lightning/Power edition overclocking and BIOS talk.

Talk about DiscussionGTX 680 Lightning/Power edition overclocking and BIOS talk.

Talk about DiscussionGTX 680 Lightning/Power edition overclocking and BIOS talk.

Pages (2): Previous 1 2
T
TheSafeDuck
Junior Member
43
08-14-2016, 02:32 PM
#11
Nvidia max safe limit stays at 1.212v, so I’d keep it as is unless you’re experiencing very low temperatures under load, which could cause instability above it. I recall a problem where adjusting the slider in the MSI Afterburner didn’t reflect actual voltage changes; I tried EVGA Precision and used a multimeter for testing, and it showed 1.2v instead. It might have been an issue with the 780 lightning only, and after flashing new BIOS it resolved the problem.
T
TheSafeDuck
08-14-2016, 02:32 PM #11

Nvidia max safe limit stays at 1.212v, so I’d keep it as is unless you’re experiencing very low temperatures under load, which could cause instability above it. I recall a problem where adjusting the slider in the MSI Afterburner didn’t reflect actual voltage changes; I tried EVGA Precision and used a multimeter for testing, and it showed 1.2v instead. It might have been an issue with the 780 lightning only, and after flashing new BIOS it resolved the problem.

S
SimplyyRaz
Member
202
09-05-2016, 01:08 PM
#12
You're seeing improved outcomes at 1.175v compared to lower frequencies, around 1300mhz. Make sure you verify neither thermal throttling nor consistent voltage across all applications.
S
SimplyyRaz
09-05-2016, 01:08 PM #12

You're seeing improved outcomes at 1.175v compared to lower frequencies, around 1300mhz. Make sure you verify neither thermal throttling nor consistent voltage across all applications.

M
Meadras
Member
139
09-05-2016, 02:06 PM
#13
I need a really bad card. I tried it for a long time and eventually reached 1.250 volts, it worked at 1256MHz but not above that. I checked stock memory clocks and overclocked, the memory stayed okay at +500mhz.
I ran both heaven and valley benchmarks for dx11 ultra, but after some time they crashed to desktop above 1256MHz.
The GPU stayed under 63°C with full fan speed, VRM never exceeded 58°C, and the memory never went over 53°C.
Could there be an issue with the card or is it a poor build?
I might revert to the original power edition BIOS since I know it’s more reliable. The extra 90MHz overclock isn’t worth the 1.25V difference for me!
Thanks again
M
Meadras
09-05-2016, 02:06 PM #13

I need a really bad card. I tried it for a long time and eventually reached 1.250 volts, it worked at 1256MHz but not above that. I checked stock memory clocks and overclocked, the memory stayed okay at +500mhz.
I ran both heaven and valley benchmarks for dx11 ultra, but after some time they crashed to desktop above 1256MHz.
The GPU stayed under 63°C with full fan speed, VRM never exceeded 58°C, and the memory never went over 53°C.
Could there be an issue with the card or is it a poor build?
I might revert to the original power edition BIOS since I know it’s more reliable. The extra 90MHz overclock isn’t worth the 1.25V difference for me!
Thanks again

A
Alsacecoser
Junior Member
44
09-06-2016, 02:34 PM
#14
What power supply do you currently have? I recall experiencing faulty clocks on 670 OC when my power supply was unstable. Temperature issues aren't the main concern; throttling begins around 70°C and only increases slightly. If your core voltage were 1.25V, the temperature would be significantly higher with air cooling. I’m not sure if it’s actually delivering that voltage. Perhaps removing the MSI Afterburner and installing Evga Precision would help—set the voltage slider to 1.212V and run a benchmark at full fan speed.

EDIT: I remember facing similar problems with voltage on my 780 Lightning.

EDIT2: Could you confirm what maximum TDP was being reached during your benchmarks? If your power supply is the limiting factor, the "furmark" indicator would reveal it quickly, as it consumes most power in all tests. However, if your PSU isn’t reliable, I wouldn’t risk using it.
A
Alsacecoser
09-06-2016, 02:34 PM #14

What power supply do you currently have? I recall experiencing faulty clocks on 670 OC when my power supply was unstable. Temperature issues aren't the main concern; throttling begins around 70°C and only increases slightly. If your core voltage were 1.25V, the temperature would be significantly higher with air cooling. I’m not sure if it’s actually delivering that voltage. Perhaps removing the MSI Afterburner and installing Evga Precision would help—set the voltage slider to 1.212V and run a benchmark at full fan speed.

EDIT: I remember facing similar problems with voltage on my 780 Lightning.

EDIT2: Could you confirm what maximum TDP was being reached during your benchmarks? If your power supply is the limiting factor, the "furmark" indicator would reveal it quickly, as it consumes most power in all tests. However, if your PSU isn’t reliable, I wouldn’t risk using it.

D
droni
Junior Member
37
09-08-2016, 10:22 AM
#15
My PSU is the Cooler Master G750M Power supply. It should be fine, offering sufficient power and rated in bronze, almost silver level.
Review of this PSU
Afterburner, GPU-Z and HWmonitor displayed identical voltages between 1.24-1.25v after applying the overvolt in Afterburner.
I attempted Precision X1 but it shows zero milli-volts and the voltage remains unchanged when adjusting the slider. This occurred with the BIOS version 80.04.28.00.3A.
Uncertain what to do next... I'll reattempt afterburner and verify the maximum TDP at 1.15 volts, comparing it to the claimed 1.25 volts to determine if there was a change.
D
droni
09-08-2016, 10:22 AM #15

My PSU is the Cooler Master G750M Power supply. It should be fine, offering sufficient power and rated in bronze, almost silver level.
Review of this PSU
Afterburner, GPU-Z and HWmonitor displayed identical voltages between 1.24-1.25v after applying the overvolt in Afterburner.
I attempted Precision X1 but it shows zero milli-volts and the voltage remains unchanged when adjusting the slider. This occurred with the BIOS version 80.04.28.00.3A.
Uncertain what to do next... I'll reattempt afterburner and verify the maximum TDP at 1.15 volts, comparing it to the claimed 1.25 volts to determine if there was a change.

P
PVPwithpat
Junior Member
40
09-08-2016, 12:40 PM
#16
I just tested furmark and noticed the voltage is shifting a bit:
TDP drops from 95% to 105% when it goes from 1.15V to 1.2V, and the temperature jumps up significantly to 69°C from 64°C.
It was interesting to see that when TDP exceeded 100%, the perfCap reason switched to power limited. I adjusted the power limit from 100% to 133% and it returned to the original reason.
I'll continue testing.
P
PVPwithpat
09-08-2016, 12:40 PM #16

I just tested furmark and noticed the voltage is shifting a bit:
TDP drops from 95% to 105% when it goes from 1.15V to 1.2V, and the temperature jumps up significantly to 69°C from 64°C.
It was interesting to see that when TDP exceeded 100%, the perfCap reason switched to power limited. I adjusted the power limit from 100% to 133% and it returned to the original reason.
I'll continue testing.

M
martims098
Member
193
09-09-2016, 06:55 AM
#17
Turning MSAA off in furmark results in the core clock dropping to 1150MHz regardless of adjustments, and the voltage falls to 1.15V. When MSAA is enabled at 8x, the voltage rises to 1.248V with a core frequency of 1.241MHz. I verified gpuz and confirmed the memory controller was fully loaded. The TDP varied between about 75% when MSAA is on and 105% when it’s off. Can you clarify this? Is it because the memory is limiting the core clock?
M
martims098
09-09-2016, 06:55 AM #17

Turning MSAA off in furmark results in the core clock dropping to 1150MHz regardless of adjustments, and the voltage falls to 1.15V. When MSAA is enabled at 8x, the voltage rises to 1.248V with a core frequency of 1.241MHz. I verified gpuz and confirmed the memory controller was fully loaded. The TDP varied between about 75% when MSAA is on and 105% when it’s off. Can you clarify this? Is it because the memory is limiting the core clock?

O
onska
Member
65
09-11-2016, 06:13 AM
#18
I would maintain it at 1.212v for continuous use if you need longer periods. I'm unsure about higher voltages because it's a lightning edition—it might handle more than lower quality cards. If the FPS difference justifies the extra voltage, go for it; there are mods to reach 1.35v but that comes with higher risk.
I tested my 780 Lightning at 1.212v for six months and another at 1.25v for six months too—no noticeable performance drop. But card behavior varies. Nvidia says 1.212v is the safest maximum.
O
onska
09-11-2016, 06:13 AM #18

I would maintain it at 1.212v for continuous use if you need longer periods. I'm unsure about higher voltages because it's a lightning edition—it might handle more than lower quality cards. If the FPS difference justifies the extra voltage, go for it; there are mods to reach 1.35v but that comes with higher risk.
I tested my 780 Lightning at 1.212v for six months and another at 1.25v for six months too—no noticeable performance drop. But card behavior varies. Nvidia says 1.212v is the safest maximum.

Pages (2): Previous 1 2