the gtx 1070 core clock reduces when oc is active
the gtx 1070 core clock reduces when oc is active
I'm experiencing problems with my MSI GTX 1070 gaming, particularly noticeable when I overclock using the newest MSI Afterburner version.
I configured my OC settings as follows:
100 core voltage
126 max power limit
92°C temperature cap
80+core clock
450+ VRAM clock
Voltage is around 1.80V
My preferred benchmarking game is Project Cars, and I can push everything to its limits—even with the extra settings—and run it at 1440p with auto driving enabled.
The game runs quite intensely, dropping my FPS to around 30–48 frames per second.
Despite this, I manage to keep temperatures between 75–76°C while fans spin at 70–75%.
Typically, my clock speed is around 2025 MHz at 60fps and a minimum of 1999MHz when at 70% load.
After about an hour of continuous gaming, I noticed my voltage dropping to 1030–1000V and reaching 800V, while my clock slowed from 2000MHz down to 1900–1898MHz.
I haven’t even reached 80°C, but temperatures have stayed around 75–76°C.
Many people say I should only worry about temperatures once they reach 85°C.
What could be causing this? I’m relatively new to GPU OC settings, so any advice would be greatly appreciated.
I would attempt to modify settings in the nvidia control panel. I prefer maximum power, high quality over low quality, 2 pre-drawn frames, and numerous tabs to experiment with to determine if it makes a difference. Also, check the resolution window and ensure it matches the correct Hz; I noticed my monitor was set at 60Hz while it should be 144Hz. This aligns with Zebarjadi's experience of heat entering the GPU. Increase case fans to full speed during gaming. A cooler environment also helps. If your room is already hot, fans alone won't improve performance.
85C is a secure threshold.
the real clock relies on numerous factors. the card manages clocks and voltage based on those conditions.
thus, lower temperatures yield superior results. you don't require 1.08v to reach 2000MHz; it should be straightforward with 1.05-1.063v.
overclocking pascal cards can be enjoyable, but they offer little practical benefit. reaching 50 or even 100MHz provides only a minimal improvement at best.
As temperatures rise, the card automatically reduces its clock speed to cut down heat production. You might wonder why it’s slowing down even though you don’t feel much warmth. When first started, just the GPU gets hot. After about an hour, the entire GPU heatsink becomes full of heat and can no longer absorb more. The case also gets warmer inside. Jayztwocents even created a video on the subject. YouTube offers the best advice:
1. Apply high-quality thermal paste (5+W/M.K).
2. Add a powerful fan (60+CFM) close to the GPU. For quieter operation, choose 20-dB.
Reference: I lowered my temps by 15-20°C. Using good paste with a few fans makes a big difference. (I bought four on sale)
I would attempt to modify settings within the nvidia control panel. Prefer maximum power, high quality over low quality, 2 pre-drawn frames, numerous tabs to experiment with to determine if it affects performance. Also check the resolution window and confirm it matches the correct Hz, since I had a 144hz monitor and noticed mine was set at 60hz inside. This aligns with Zebarjadi's experience of heat entering the GPU. Increase case fans to full speed during gaming. A cooler environment also makes a difference. If your room is already hot, fans alone won't improve things.