StraferightDave
n00b
- Joined
- Feb 9, 2019
- Messages
- 4
I just build my first gaming PC from scratch. Before I get into the main issue, I will detail my entire build if that makes any difference.
Build: MB: Case Thermaltake View 71, Aorus Master MB, 32gb of Royal Z ram@3600mhz, i9 9900k, 2 2080ti FTW3 NVlink, Asus Thor 1200watt PSU, Krakken 360 rad AIO 3 TT case fans.
System power draw at peak (thanks to thor oled) is around 900+ watt with OC on 9900k and usually no OC on GPUs (need custom loop for temps).
OK now to my problems and yes I do admit I am a very new builder and experimenter to OC but a long time PC gamer and tech enthusiast. I tried to research as much as possible and am still ending up here.
I have tried all sorts of OC on my motherboard and have yet to have one make sense. I am currently running things stock and even that is not making any sense to me. At "stock" right now outside of enabling XMP, I just ran cinebench and it gave me a score of 2061, had all 8 cores at 5ghz, and highest core voltage was 1.28 and current 8 core CPU multiplier is 48 to 49 with 1.xx to 1.2x voltage. This is extremely efficient and optimized on paper and that doesn't make any sense!
When I tried to manually overclock it I did the following typically: multiplier 50, uncore ratio 47, AVX offset 2, diable MCT (creatively labeled "enhanced multicore performance" in the confusing af gigabyte bios), disabled c states, disabled any boost tech, disabled SVID, Voltage at 1.3 to 1.33, Loadline calibration to turbo, and set the watts up to max along with tjmax etc for stability or exclude throttling issues when testing stability. I Didn't do any negative offsets because I don't know how they work just yet and how to do it on the gigabyte MB.
Problems: never stable and from all the videos I watched, I think I did enough manipulations to achieve something proper. Even while it would be stable my cinebench would be in the 1800s to very low 1900s.
Now when I reverted to stock, I would disable MCT, enable XMP, set a voltage, disable voltage optimization and even turbo boost etc and I would get a constant 8 core 4700mhz OC. This was supposed to be the function of a basic MCT enabling but I would get it disabled. Now having it enabled it just fluctuates all over the place and I don't know why or what I can do. Sometimes the voltages can be pretty high, the OC never seems to be predictable and I need help. I do wonder if there is some sort of learning behavior because the longer I have kept this set to "stock" with MCT enabled, the voltages appear a bit lower, my cinebench has been higher and temps quite acceptable given the "stock OC." 64 degrees on CPU doing a few benches and around 68 to 70 gaming (when both GPUs enabled...1 gpu lowers temps a bit).
Attached is after I just ram cinebench to show the 50x and voltages
Thanks for reading this, I know it was long but it is my first post, I know I am an idiot noob H forum member, so want to be as detailed as possible to help those that read though this.
Build: MB: Case Thermaltake View 71, Aorus Master MB, 32gb of Royal Z ram@3600mhz, i9 9900k, 2 2080ti FTW3 NVlink, Asus Thor 1200watt PSU, Krakken 360 rad AIO 3 TT case fans.
System power draw at peak (thanks to thor oled) is around 900+ watt with OC on 9900k and usually no OC on GPUs (need custom loop for temps).
OK now to my problems and yes I do admit I am a very new builder and experimenter to OC but a long time PC gamer and tech enthusiast. I tried to research as much as possible and am still ending up here.
I have tried all sorts of OC on my motherboard and have yet to have one make sense. I am currently running things stock and even that is not making any sense to me. At "stock" right now outside of enabling XMP, I just ran cinebench and it gave me a score of 2061, had all 8 cores at 5ghz, and highest core voltage was 1.28 and current 8 core CPU multiplier is 48 to 49 with 1.xx to 1.2x voltage. This is extremely efficient and optimized on paper and that doesn't make any sense!
When I tried to manually overclock it I did the following typically: multiplier 50, uncore ratio 47, AVX offset 2, diable MCT (creatively labeled "enhanced multicore performance" in the confusing af gigabyte bios), disabled c states, disabled any boost tech, disabled SVID, Voltage at 1.3 to 1.33, Loadline calibration to turbo, and set the watts up to max along with tjmax etc for stability or exclude throttling issues when testing stability. I Didn't do any negative offsets because I don't know how they work just yet and how to do it on the gigabyte MB.
Problems: never stable and from all the videos I watched, I think I did enough manipulations to achieve something proper. Even while it would be stable my cinebench would be in the 1800s to very low 1900s.
Now when I reverted to stock, I would disable MCT, enable XMP, set a voltage, disable voltage optimization and even turbo boost etc and I would get a constant 8 core 4700mhz OC. This was supposed to be the function of a basic MCT enabling but I would get it disabled. Now having it enabled it just fluctuates all over the place and I don't know why or what I can do. Sometimes the voltages can be pretty high, the OC never seems to be predictable and I need help. I do wonder if there is some sort of learning behavior because the longer I have kept this set to "stock" with MCT enabled, the voltages appear a bit lower, my cinebench has been higher and temps quite acceptable given the "stock OC." 64 degrees on CPU doing a few benches and around 68 to 70 gaming (when both GPUs enabled...1 gpu lowers temps a bit).
Attached is after I just ram cinebench to show the 50x and voltages
Thanks for reading this, I know it was long but it is my first post, I know I am an idiot noob H forum member, so want to be as detailed as possible to help those that read though this.