F5F Stay Refreshed Power Users Overclocking Core clock vs Memory Clock.

Core clock vs Memory Clock.

Core clock vs Memory Clock.

D
Denaos5208
Member
68
08-26-2016, 06:39 AM
#1
Hello, I frequently encounter these discussions before, yet sometimes I struggle to understand them. It has been quite challenging for me to create a new character with my card right now. I currently own an Asus Turbo GTX 960 with 4GB RAM, and I play on a wide-screen 2560 x 1080 monitor. Occasionally, I experiment with different settings to achieve better graphics for games that require lower specifications, using supersampling and aiming for a 5k virtual display. Recently, I've been attempting to overclock my card frequently, but I consistently encounter varying outcomes when adjusting core and memory speeds. I've read that maintaining a stable core clock is beneficial, yet sometimes it's advantageous to allocate more power to memory for higher resolutions. I'm unsure about these adjustments because my performance doesn't always improve, even when increasing core by 150 and memory by 200. In contrast, with games like Witcher, my performance can drop to around 78 on the witcher benchmark, while it stays under 70 on Unigine Heaven. I apologize for the lengthy explanation and hope you can provide some guidance from the community on this issue. Also, please accept my apologies for any English errors—I'm not fluent in this language.
D
Denaos5208
08-26-2016, 06:39 AM #1

Hello, I frequently encounter these discussions before, yet sometimes I struggle to understand them. It has been quite challenging for me to create a new character with my card right now. I currently own an Asus Turbo GTX 960 with 4GB RAM, and I play on a wide-screen 2560 x 1080 monitor. Occasionally, I experiment with different settings to achieve better graphics for games that require lower specifications, using supersampling and aiming for a 5k virtual display. Recently, I've been attempting to overclock my card frequently, but I consistently encounter varying outcomes when adjusting core and memory speeds. I've read that maintaining a stable core clock is beneficial, yet sometimes it's advantageous to allocate more power to memory for higher resolutions. I'm unsure about these adjustments because my performance doesn't always improve, even when increasing core by 150 and memory by 200. In contrast, with games like Witcher, my performance can drop to around 78 on the witcher benchmark, while it stays under 70 on Unigine Heaven. I apologize for the lengthy explanation and hope you can provide some guidance from the community on this issue. Also, please accept my apologies for any English errors—I'm not fluent in this language.

O
o0_nawaf_0o
Junior Member
5
08-27-2016, 09:23 AM
#2
You should expect improved performance by focusing on the GPU core clock instead of the memory. If you examine the pre-OCed cards, you'll notice they prioritize the GPU core clock over the memory.
O
o0_nawaf_0o
08-27-2016, 09:23 AM #2

You should expect improved performance by focusing on the GPU core clock instead of the memory. If you examine the pre-OCed cards, you'll notice they prioritize the GPU core clock over the memory.

M
MarianAngela
Junior Member
19
09-09-2016, 05:43 PM
#3
It requires a mix of the GPU core clock and memory clock. The process involves gradual adjustments, testing one at a time, to avoid overloading either the memory or GPU, which could harm performance.
M
MarianAngela
09-09-2016, 05:43 PM #3

It requires a mix of the GPU core clock and memory clock. The process involves gradual adjustments, testing one at a time, to avoid overloading either the memory or GPU, which could harm performance.

M
Magister_Magi
Member
151
09-17-2016, 01:36 AM
#4
Tradesman1:
The balance between GPU core clock and memory clock is crucial. It requires gradual testing, adjusting one at a time to avoid strain on either component. If one exceeds its limit, it can negatively impact the other. Thank you for your advice. It seems a balance is needed, but I wonder if prioritizing one over the other would be better for my specific setup. For an ultrawide, I aim for about 20-25% more pixels, which might mean focusing more on memory. Also, I’m curious about the voltage and whether it affects either part more than the other. I’m also checking if better airflow improves performance.
M
Magister_Magi
09-17-2016, 01:36 AM #4

Tradesman1:
The balance between GPU core clock and memory clock is crucial. It requires gradual testing, adjusting one at a time to avoid strain on either component. If one exceeds its limit, it can negatively impact the other. Thank you for your advice. It seems a balance is needed, but I wonder if prioritizing one over the other would be better for my specific setup. For an ultrawide, I aim for about 20-25% more pixels, which might mean focusing more on memory. Also, I’m curious about the voltage and whether it affects either part more than the other. I’m also checking if better airflow improves performance.

T
Tigrio
Member
54
09-17-2016, 03:06 PM
#5
You should expect improved performance by focusing on the GPU core clock instead of the memory. If you examine the pre-OCed cards, you'll notice they prioritize the GPU core clock over the memory.
T
Tigrio
09-17-2016, 03:06 PM #5

You should expect improved performance by focusing on the GPU core clock instead of the memory. If you examine the pre-OCed cards, you'll notice they prioritize the GPU core clock over the memory.