F5F Stay Refreshed Power Users Overclocking MSI 970 Won't Overclock

MSI 970 Won't Overclock

MSI 970 Won't Overclock

D
Dover_Demon
Junior Member
21
05-20-2024, 05:31 AM
#1
Hello everyone. I’m just a few months away from assembling my first PC, and only started experimenting with overclocking yesterday, so please forgive any oversight if I’ve missed something obvious. I also searched the forums without success, so I hope this hasn’t been covered before.

I followed the OC settings from this article:
http://www.hardocp.com/article/2014/09/2...2vaq7grKUn
And here’s my parts list:
http://pcpartpicker.com/list/xGbMGX

I configured a profile in Afterburner using the exact settings from the "Overclocking Without Voltage" section, applied it, and GPU-Z now shows the same increased GPU clock and boost as described in the article when that profile is active. However, after running Unigine Heaven (1920x1080, Ultra quality, Extreme tessellation, x8 AA), GPU-Z displays full-load values significantly lower than what the article reported for Hardocp. Below are my current results:

GPU Core Clock: 1468.9
CPU Memory Clock: 2000.7
GPU Temperature: 69
Fan Speed: 58%
Fan Speed (RPM): 1345
Memory Used: 1332
GPU Load: 100%
Memory Controller Load: 67%
Video Engine Load: 0%
Bus Interface Load: 20%
Power Consumption: 95.3%

PerfCap Reason: Util
VDDC: 1.2120

What I’m observing is that my memory usage is only about 64% of what the tests showed, my memory controller load is much lower, and the voltage seems higher than expected. The most concerning thing is that my power draw is only 95.3%, which is quite low.

Given my basic understanding, it appears that with Afterburner’s profile set to 110% Power Limit, I should see a Power Consumption around that level in GPU-Z, as shown in the article. Since mine is below 100%, it might suggest I’m not overclocking at all if I also notice the higher clock speeds recorded by GPU-Z.

Could this be because of the reduced memory usage? If so, would my RAM be faulty? Or could the slightly higher VDDC indicate a failing power supply? Also, I see GPU-Z lists “PerfCap Reason: Util”; I’m not sure what that means, but if it relates to Nvidia Control Panel’s Power Management Mode, it’s currently set to “Prefer Maximum Performance.”

Any advice or clarification would be greatly appreciated. I’ve read many overclocking guides already and can’t spot any steps I missed. I’m still in the dark here. Thank you all for your help!
D
Dover_Demon
05-20-2024, 05:31 AM #1

Hello everyone. I’m just a few months away from assembling my first PC, and only started experimenting with overclocking yesterday, so please forgive any oversight if I’ve missed something obvious. I also searched the forums without success, so I hope this hasn’t been covered before.

I followed the OC settings from this article:
http://www.hardocp.com/article/2014/09/2...2vaq7grKUn
And here’s my parts list:
http://pcpartpicker.com/list/xGbMGX

I configured a profile in Afterburner using the exact settings from the "Overclocking Without Voltage" section, applied it, and GPU-Z now shows the same increased GPU clock and boost as described in the article when that profile is active. However, after running Unigine Heaven (1920x1080, Ultra quality, Extreme tessellation, x8 AA), GPU-Z displays full-load values significantly lower than what the article reported for Hardocp. Below are my current results:

GPU Core Clock: 1468.9
CPU Memory Clock: 2000.7
GPU Temperature: 69
Fan Speed: 58%
Fan Speed (RPM): 1345
Memory Used: 1332
GPU Load: 100%
Memory Controller Load: 67%
Video Engine Load: 0%
Bus Interface Load: 20%
Power Consumption: 95.3%

PerfCap Reason: Util
VDDC: 1.2120

What I’m observing is that my memory usage is only about 64% of what the tests showed, my memory controller load is much lower, and the voltage seems higher than expected. The most concerning thing is that my power draw is only 95.3%, which is quite low.

Given my basic understanding, it appears that with Afterburner’s profile set to 110% Power Limit, I should see a Power Consumption around that level in GPU-Z, as shown in the article. Since mine is below 100%, it might suggest I’m not overclocking at all if I also notice the higher clock speeds recorded by GPU-Z.

Could this be because of the reduced memory usage? If so, would my RAM be faulty? Or could the slightly higher VDDC indicate a failing power supply? Also, I see GPU-Z lists “PerfCap Reason: Util”; I’m not sure what that means, but if it relates to Nvidia Control Panel’s Power Management Mode, it’s currently set to “Prefer Maximum Performance.”

Any advice or clarification would be greatly appreciated. I’ve read many overclocking guides already and can’t spot any steps I missed. I’m still in the dark here. Thank you all for your help!

S
Sheray
Member
218
05-20-2024, 05:31 AM
#2
nnickers :
Hi everyone. I'm only a few months out from building my first PC, and only began to attempt OCing yesterday, so apologies in advance if I'm missing anything obvious. I also tried searching the forums with no luck, so I hope this hasn't been addressed before.
I took my OC settings from this article:
http://www.hardocp.com/article/2014/09/2...2vaq7grKUn
And here is my parts list:
http://pcpartpicker.com/list/xGbMGX
I set up a profile in Afterburner with the exact settings they use in the "Overclocking Without Voltage" section, applied it, and GPU-Z does reflect the same increased GPU clock and boost as the article above when this profile is active. However, after...
S
Sheray
05-20-2024, 05:31 AM #2

nnickers :
Hi everyone. I'm only a few months out from building my first PC, and only began to attempt OCing yesterday, so apologies in advance if I'm missing anything obvious. I also tried searching the forums with no luck, so I hope this hasn't been addressed before.
I took my OC settings from this article:
http://www.hardocp.com/article/2014/09/2...2vaq7grKUn
And here is my parts list:
http://pcpartpicker.com/list/xGbMGX
I set up a profile in Afterburner with the exact settings they use in the "Overclocking Without Voltage" section, applied it, and GPU-Z does reflect the same increased GPU clock and boost as the article above when this profile is active. However, after...

J
JinxCracuda
Junior Member
22
05-20-2024, 05:31 AM
#3
Are you sure they used Heaven to load up the card? I can't find any info on what they are using for this.
J
JinxCracuda
05-20-2024, 05:31 AM #3

Are you sure they used Heaven to load up the card? I can't find any info on what they are using for this.

F
flameboy101
Member
173
05-20-2024, 05:31 AM
#4
SinxarKnights :
Are you certain they used Heaven to load the card? I don't have any details about what they're doing here. Actually, I just chose Heaven since overclocking benchmarks are often mentioned as the go-to test. Do you think Heaven might not be challenging enough for my system to reach full capacity? I've been testing at 1920x1080 to match my screen, but I could try a higher resolution later if it helps determine the problem. I wouldn't want to look foolish if it's something as basic as that :/
F
flameboy101
05-20-2024, 05:31 AM #4

SinxarKnights :
Are you certain they used Heaven to load the card? I don't have any details about what they're doing here. Actually, I just chose Heaven since overclocking benchmarks are often mentioned as the go-to test. Do you think Heaven might not be challenging enough for my system to reach full capacity? I've been testing at 1920x1080 to match my screen, but I could try a higher resolution later if it helps determine the problem. I wouldn't want to look foolish if it's something as basic as that :/

T
136
05-20-2024, 05:31 AM
#5
Heaven isn't the best, but it's good for locating items. Remember to check Valley and 3Dmark too (try both firestrike and firestrike extreme). Don't go for Furmark.
T
TurboN_GGA9000
05-20-2024, 05:31 AM #5

Heaven isn't the best, but it's good for locating items. Remember to check Valley and 3Dmark too (try both firestrike and firestrike extreme). Don't go for Furmark.

F
FladaHD
Member
222
05-20-2024, 05:31 AM
#6
As you are aware, operating an OC game is your own responsibility and depends on chance. The card only works reliably at factory-set clocks and as it was originally purchased. Some people might say they achieved impressive results, while others may not get half as much or not at all. That’s just the nature of things. I view all comments with a bit of skepticism, similar to how some reviews are cherry-picked rather than representing the full range of cards available. People often boast online without considering the inconsistencies. Good luck!
F
FladaHD
05-20-2024, 05:31 AM #6

As you are aware, operating an OC game is your own responsibility and depends on chance. The card only works reliably at factory-set clocks and as it was originally purchased. Some people might say they achieved impressive results, while others may not get half as much or not at all. That’s just the nature of things. I view all comments with a bit of skepticism, similar to how some reviews are cherry-picked rather than representing the full range of cards available. People often boast online without considering the inconsistencies. Good luck!

P
Poulpee
Junior Member
15
05-20-2024, 05:31 AM
#7
nnickers :
Hi everyone. I'm only a few months out from building my first PC, and only began to attempt OCing yesterday, so apologies in advance if I'm missing anything obvious. I also tried searching the forums with no luck, so I hope this hasn't been addressed before.
I took my OC settings from this article:
http://www.hardocp.com/article/2014/09/2...2vaq7grKUn
And here is my parts list:
http://pcpartpicker.com/list/xGbMGX
I set up a profile in Afterburner with the exact settings they use in the "Overclocking Without Voltage" section, applied it, and GPU-Z does reflect the same increased GPU clock and boost as the article above when this profile is active. However, after running Unigine Heaven for some time (1920x1080, Ultra quality, Extreme tessellation, x8 AA), GPU-Z show my full-load at far lower levels than what they saw at hardocp. Here are my maximum results below:
GPU Core Clock: 1468.9
CPU Memory Clock: 2000.7
GPU Temperature: 69
Fan Speed: 58%
Fan Speed (RPM): 1345
Memory Used: 1332
GPU Load: 100%
Memory Controller Load: 67%
Video Engine Load: 0%
Bus Interface Load: 20%
Power Consumption: 95.3%
PerfCap Reason: Util
VDDC: 1.2120
What I'm noticing is that my memory used is only about 64% of what their tests produced, my memory controller load is much lower, my voltage is for some reason higher, and what seems worst to me is that
my power consumption reaches only 95.3%.
With an admittedly basic understanding of all this, it seems that with my Afterburner OC profile set to 110% Power Limit I should be seeing Power Consumption around there in GPU-Z, as they do in the article. With my Power Consumption reflecting <100%, I would have interpreted that as my card not overclocking at all if I weren't also seeing the increased clocks registered in GPU-Z.
Is this due to the lower memory usage, in which case would my RAM likely be faulty? Or my slightly higher VDDC, and would that suggest a bad PSU? I also notice GPU-Z lists "PerfCap Reason: Util;" I'm not entirely sure what that means, but if it pertains to the Nvidia Control Panel Power Management Mode, I have mine set to "Prefer Maximum Performance."
Any help or suggestions here would be hugely appreciated. I've read probably a dozen tutorials on overclocking now and can't find any steps I've missed, so I'm pretty in the dark at this point. Again, thank you for any help!
Memory usage could be the difference between drivers, or versions of Heaven if they were using an older one, and general hardware differences between the builds.
The voltages being different will always happen between cards....the maximums are set in the bios, so it'll grab what it needs to run at the clocks you're telling it to run at.
Power consumption....you don't want the power consumption to go above 100%, or even 110%, regardless of where you have the power limit target set at. IF it does, the card will power limit throttle and your core clocks will drop.
Just because the power consumption doesn't say 100% or 110% doesn't mean that your card isn't overclocking, it means that your card isn't pulling as much of the power limit that's set in the bios....that's actually a good thing. No worries. When you start seeing PerfCap REason: PWR, that's when you have problems.
Every card is different, and using settings from some article isn't a good idea. FIgure out what YOUR card can do, and run with it. But remember, that EVERY program is just a little bit different, and what works for one, won't work for another.
Heaven, btw, is a good benchmark to use for testing, because it doesn't use a lot of CPU, where as Valley is heavily CPU dependent. OR, use FS Ultra, graphics test 1 and 2. Pay no mind to the physics test or the combined test. Just run graphics test 1 and 2.
P
Poulpee
05-20-2024, 05:31 AM #7

nnickers :
Hi everyone. I'm only a few months out from building my first PC, and only began to attempt OCing yesterday, so apologies in advance if I'm missing anything obvious. I also tried searching the forums with no luck, so I hope this hasn't been addressed before.
I took my OC settings from this article:
http://www.hardocp.com/article/2014/09/2...2vaq7grKUn
And here is my parts list:
http://pcpartpicker.com/list/xGbMGX
I set up a profile in Afterburner with the exact settings they use in the "Overclocking Without Voltage" section, applied it, and GPU-Z does reflect the same increased GPU clock and boost as the article above when this profile is active. However, after running Unigine Heaven for some time (1920x1080, Ultra quality, Extreme tessellation, x8 AA), GPU-Z show my full-load at far lower levels than what they saw at hardocp. Here are my maximum results below:
GPU Core Clock: 1468.9
CPU Memory Clock: 2000.7
GPU Temperature: 69
Fan Speed: 58%
Fan Speed (RPM): 1345
Memory Used: 1332
GPU Load: 100%
Memory Controller Load: 67%
Video Engine Load: 0%
Bus Interface Load: 20%
Power Consumption: 95.3%
PerfCap Reason: Util
VDDC: 1.2120
What I'm noticing is that my memory used is only about 64% of what their tests produced, my memory controller load is much lower, my voltage is for some reason higher, and what seems worst to me is that
my power consumption reaches only 95.3%.
With an admittedly basic understanding of all this, it seems that with my Afterburner OC profile set to 110% Power Limit I should be seeing Power Consumption around there in GPU-Z, as they do in the article. With my Power Consumption reflecting <100%, I would have interpreted that as my card not overclocking at all if I weren't also seeing the increased clocks registered in GPU-Z.
Is this due to the lower memory usage, in which case would my RAM likely be faulty? Or my slightly higher VDDC, and would that suggest a bad PSU? I also notice GPU-Z lists "PerfCap Reason: Util;" I'm not entirely sure what that means, but if it pertains to the Nvidia Control Panel Power Management Mode, I have mine set to "Prefer Maximum Performance."
Any help or suggestions here would be hugely appreciated. I've read probably a dozen tutorials on overclocking now and can't find any steps I've missed, so I'm pretty in the dark at this point. Again, thank you for any help!
Memory usage could be the difference between drivers, or versions of Heaven if they were using an older one, and general hardware differences between the builds.
The voltages being different will always happen between cards....the maximums are set in the bios, so it'll grab what it needs to run at the clocks you're telling it to run at.
Power consumption....you don't want the power consumption to go above 100%, or even 110%, regardless of where you have the power limit target set at. IF it does, the card will power limit throttle and your core clocks will drop.
Just because the power consumption doesn't say 100% or 110% doesn't mean that your card isn't overclocking, it means that your card isn't pulling as much of the power limit that's set in the bios....that's actually a good thing. No worries. When you start seeing PerfCap REason: PWR, that's when you have problems.
Every card is different, and using settings from some article isn't a good idea. FIgure out what YOUR card can do, and run with it. But remember, that EVERY program is just a little bit different, and what works for one, won't work for another.
Heaven, btw, is a good benchmark to use for testing, because it doesn't use a lot of CPU, where as Valley is heavily CPU dependent. OR, use FS Ultra, graphics test 1 and 2. Pay no mind to the physics test or the combined test. Just run graphics test 1 and 2.

9
992x
Senior Member
506
05-20-2024, 05:31 AM
#8
Vellinious :
nnickers :
Hi everyone. I'm only a few months out from building my first PC, and only began to attempt OCing yesterday, so apologies in advance if I'm missing anything obvious. I also tried searching the forums with no luck, so I hope this hasn't been addressed before.
I took my OC settings from this article:
http://www.hardocp.com/article/2014/09/2...2vaq7grKUn
And here is my parts list:
http://pcpartpicker.com/list/xGbMGX
I set up a profile in Afterburner with the exact settings they use in the "Overclocking Without Voltage" section, applied it, and GPU-Z does reflect the same increased GPU clock and boost as the article above when this profile is active. However, after running Unigine Heaven for some time (1920x1080, Ultra quality, Extreme tessellation, x8 AA), GPU-Z show my full-load at far lower levels than what they saw at hardocp. Here are my maximum results below:
GPU Core Clock: 1468.9
CPU Memory Clock: 2000.7
GPU Temperature: 69
Fan Speed: 58%
Fan Speed (RPM): 1345
Memory Used: 1332
GPU Load: 100%
Memory Controller Load: 67%
Video Engine Load: 0%
Bus Interface Load: 20%
Power Consumption: 95.3%
PerfCap Reason: Util
VDDC: 1.2120
What I'm noticing is that my memory used is only about 64% of what their tests produced, my memory controller load is much lower, my voltage is for some reason higher, and what seems worst to me is that
my power consumption reaches only 95.3%.
With an admittedly basic understanding of all this, it seems that with my Afterburner OC profile set to 110% Power Limit I should be seeing Power Consumption around there in GPU-Z, as they do in the article. With my Power Consumption reflecting <100%, I would have interpreted that as my card not overclocking at all if I weren't also seeing the increased clocks registered in GPU-Z.
Is this due to the lower memory usage, in which case would my RAM likely be faulty? Or my slightly higher VDDC, and would that suggest a bad PSU? I also notice GPU-Z lists "PerfCap Reason: Util;" I'm not entirely sure what that means, but if it pertains to the Nvidia Control Panel Power Management Mode, I have mine set to "Prefer Maximum Performance."
Any help or suggestions here would be hugely appreciated. I've read probably a dozen tutorials on overclocking now and can't find any steps I've missed, so I'm pretty in the dark at this point. Again, thank you for any help!
Memory usage could be the difference between drivers, or versions of Heaven if they were using an older one, and general hardware differences between the builds.
The voltages being different will always happen between cards....the maximums are set in the bios, so it'll grab what it needs to run at the clocks you're telling it to run at.
Power consumption....you don't want the power consumption to go above 100%, or even 110%, regardless of where you have the power limit target set at. IF it does, the card will power limit throttle and your core clocks will drop.
Just because the power consumption doesn't say 100% or 110% doesn't mean that your card isn't overclocking, it means that your card isn't pulling as much of the power limit that's set in the bios....that's actually a good thing. No worries. When you start seeing PerfCap REason: PWR, that's when you have problems.
Every card is different, and using settings from some article isn't a good idea. FIgure out what YOUR card can do, and run with it. But remember, that EVERY program is just a little bit different, and what works for one, won't work for another.
Heaven, btw, is a good benchmark to use for testing, because it doesn't use a lot of CPU, where as Valley is heavily CPU dependent. OR, use FS Ultra, graphics test 1 and 2. Pay no mind to the physics test or the combined test. Just run graphics test 1 and 2.
Thank you for the rundown! Like I'd said, this is all new to me so I really wasn't sure how to interpret most of these results. This has really helped me make sense of them though.
9
992x
05-20-2024, 05:31 AM #8

Vellinious :
nnickers :
Hi everyone. I'm only a few months out from building my first PC, and only began to attempt OCing yesterday, so apologies in advance if I'm missing anything obvious. I also tried searching the forums with no luck, so I hope this hasn't been addressed before.
I took my OC settings from this article:
http://www.hardocp.com/article/2014/09/2...2vaq7grKUn
And here is my parts list:
http://pcpartpicker.com/list/xGbMGX
I set up a profile in Afterburner with the exact settings they use in the "Overclocking Without Voltage" section, applied it, and GPU-Z does reflect the same increased GPU clock and boost as the article above when this profile is active. However, after running Unigine Heaven for some time (1920x1080, Ultra quality, Extreme tessellation, x8 AA), GPU-Z show my full-load at far lower levels than what they saw at hardocp. Here are my maximum results below:
GPU Core Clock: 1468.9
CPU Memory Clock: 2000.7
GPU Temperature: 69
Fan Speed: 58%
Fan Speed (RPM): 1345
Memory Used: 1332
GPU Load: 100%
Memory Controller Load: 67%
Video Engine Load: 0%
Bus Interface Load: 20%
Power Consumption: 95.3%
PerfCap Reason: Util
VDDC: 1.2120
What I'm noticing is that my memory used is only about 64% of what their tests produced, my memory controller load is much lower, my voltage is for some reason higher, and what seems worst to me is that
my power consumption reaches only 95.3%.
With an admittedly basic understanding of all this, it seems that with my Afterburner OC profile set to 110% Power Limit I should be seeing Power Consumption around there in GPU-Z, as they do in the article. With my Power Consumption reflecting <100%, I would have interpreted that as my card not overclocking at all if I weren't also seeing the increased clocks registered in GPU-Z.
Is this due to the lower memory usage, in which case would my RAM likely be faulty? Or my slightly higher VDDC, and would that suggest a bad PSU? I also notice GPU-Z lists "PerfCap Reason: Util;" I'm not entirely sure what that means, but if it pertains to the Nvidia Control Panel Power Management Mode, I have mine set to "Prefer Maximum Performance."
Any help or suggestions here would be hugely appreciated. I've read probably a dozen tutorials on overclocking now and can't find any steps I've missed, so I'm pretty in the dark at this point. Again, thank you for any help!
Memory usage could be the difference between drivers, or versions of Heaven if they were using an older one, and general hardware differences between the builds.
The voltages being different will always happen between cards....the maximums are set in the bios, so it'll grab what it needs to run at the clocks you're telling it to run at.
Power consumption....you don't want the power consumption to go above 100%, or even 110%, regardless of where you have the power limit target set at. IF it does, the card will power limit throttle and your core clocks will drop.
Just because the power consumption doesn't say 100% or 110% doesn't mean that your card isn't overclocking, it means that your card isn't pulling as much of the power limit that's set in the bios....that's actually a good thing. No worries. When you start seeing PerfCap REason: PWR, that's when you have problems.
Every card is different, and using settings from some article isn't a good idea. FIgure out what YOUR card can do, and run with it. But remember, that EVERY program is just a little bit different, and what works for one, won't work for another.
Heaven, btw, is a good benchmark to use for testing, because it doesn't use a lot of CPU, where as Valley is heavily CPU dependent. OR, use FS Ultra, graphics test 1 and 2. Pay no mind to the physics test or the combined test. Just run graphics test 1 and 2.
Thank you for the rundown! Like I'd said, this is all new to me so I really wasn't sure how to interpret most of these results. This has really helped me make sense of them though.

P
PemahiXD007YT
Member
120
05-20-2024, 05:31 AM
#9
No problem.
P
PemahiXD007YT
05-20-2024, 05:31 AM #9

No problem.