CaptainXCrunch Posted August 26, 2019 Share Posted August 26, 2019 Hey ! I installed my fully Corsair custom loop yesterday . Very satisfied with the result at the moment . My EVGA RTX 2080 is sitting at around 26C idle and 45C in Superposition . But i'm a little bit concerned with my 8700K temps . I tried to overclock it at 5GHZ on all core @ 1.33V , seem stable but temps goes to 90C while benchmarking and sit around 40C idle , which obviously i'm not comfortable at . I haven't overclocked it before with my H150i , so i don't have any reference to compare result with . At stock setting in the bios temps sit at 30C idle / 60C benchmarking . I'm thinking i maybe haven't applied thermal paste correctly , but i'd like to see what result some of you get before going in . At the moment i've dialed 4.6GHZ @ 1.2V / 36C idle/ 75C benchmarking which i can live with , but i'd really like to go for the 5GHZ ... which i may have to delid the CPU to get the result i'm expecting :P https://imgur.com/gallery/aPQN2JF Link to comment Share on other sites More sharing options...
skyteam78 Posted August 26, 2019 Share Posted August 26, 2019 (edited) Hi, XC7 contains TIM so you shouldn't apply thermal paste on it. I had no issues to run my 8700K stable on 4.8 GHz all cores with my previous 110I GTX cooler (it was hitting ~77C during stress test). How you run your benchmarking? What tool are you using for that? With XC7 and delidded 8700K I have much better results. Edited August 26, 2019 by skyteam78 Link to comment Share on other sites More sharing options...
CaptainXCrunch Posted August 26, 2019 Author Share Posted August 26, 2019 Hi, XC7 contains TIM so you shouldn't apply thermal paste on it. I had no issues to run my 8700K stable on 4.8 GHz all cores with my previous 110I GTX cooler (it was hitting ~77C during stress test). How you run your benchmarking? What tool are you using for that? With XC7 and delidded 8700K I have much better results. Just to clarify , i removed the pre applied thermal paste on the XC7 and replaced it with the Corsair TM30 . Tube is brand new but seemed a little bit weird when i applied the first time , it did the same as a bottle of mustard which has not been use for a while , like if the mix inside it had split ... So i'm not quite sure if i had a bad tube . At the moment , before i run intensive benchmarking i use Cinebench R20 just to see if the OC would atleast hold up for a little and see what kind of temperature i get around . And i tried to run Prime 95 just for the sake of it and temp skyrocketed to 100C @ 5GHZ . I'm probably gonna end up applying Thermal Grizzly Kryonaut when i'll drain the loop in the upcoming week . I hope temps go back down . I'm pretty sure my H150i would have got better temps if i had dialed in a 5.0GHZ OC with it :/ Link to comment Share on other sites More sharing options...
Corsair Employee Corsair GregX Posted August 26, 2019 Corsair Employee Share Posted August 26, 2019 Hi, Looks really nice! According to your GPU idle and load temperatures it seems the loop works really good. How is your coolant temperature under load? Are you utilizing the Hydro X profiles or how are you managing fans and pump speed? Things to consider when comparing AIO for CPU and custom cooling for CPU and GPU. With AIO, you have a dedicated radiator for a single heat source. In your case, a 360 radiator for CPU only. With custom cooling there is additional heat source (GPU) which adds to the same radiator. It's normal that the CPU temperature is a tad higher compared to AIO cooling. Even with your system that has a 420 radiator for CPU and GPU. I would recommend to check current imprint of thermal paste, do take a few pictures so we can check. Then reapply TM30. Mix it if it looks like it has separated. 8700K with stock TIM between die and IHS does run really hot. With 5G+ and additional voltage, it's a mini inferno. If aiming for such clocks I would consider to delid it. Link to comment Share on other sites More sharing options...
CaptainXCrunch Posted August 26, 2019 Author Share Posted August 26, 2019 Once again , hi Greg ahah ! Currently i'm running pump and fans at a fixed 50-60% cause i don't want to hear the system ramping up and down . Coolant sit around 29-30 . I am well aware that i have another heat source in the loop , that's why i went with a XR7 which is much THICCER than the one that was on my H150i ahah ! I'm not concerned about the idle temp that much , but i feel like it wasn't spiking that high when applying a load on the CPU with the 8700k . I'm also aware that the 8700k is almost like the surface of the sun :P Like mentionned in my other post for the XG7 , i will drain the loop next weekend to change my fluid . I will take picture about all the process if you guys want ! Link to comment Share on other sites More sharing options...
c-attack Posted August 26, 2019 Share Posted August 26, 2019 Pump speed/Radiators/Fan speed all affect the coolant temperature (or delta T). 29-30C is about as low as most people can get in Summer in a 23-26C room. The increase in coolant will depend on the above factors. Is it one 360mm radiator? I can't quite see. CPU temp vs Coolant Temp is the measure of voltage, conductivity, other stuff you can't control, but it is close to a constant will give you an idea of the maximum coolant temperature you can reach before you get into CPU max temp trouble. Most standard 8700Ks are going to have a CPU to Coolant differential of +50C at 1.30v. So if coolant is reaching 40C, that would explain both the idle cpu temp (coolant temp = minimum possible CPU temp) and the maximum temp (40C + 50C = 90C). Link to comment Share on other sites More sharing options...
CaptainXCrunch Posted September 2, 2019 Author Share Posted September 2, 2019 I think i may have found what was wrong . For some reason the thermal paste didn't spread out all the way across the cpu . Went for a X pattern this time ... should be plenty with this application ... The picture is from the previous application , didn't take any of this one since it always end up as a debate , will update you guys with how much improvement i'll get https://imgur.com/a/v5JdC8C Link to comment Share on other sites More sharing options...
Vegan Posted September 2, 2019 Share Posted September 2, 2019 looks like the interface pressure between the block and the CPU was uneven I use AMD but Intel cooler mounts are fairly solid Link to comment Share on other sites More sharing options...
CaptainXCrunch Posted September 2, 2019 Author Share Posted September 2, 2019 I don't really know what happened there , i suspect i tightened one side too much and thermal paste went to the other side and couldn't get back on the corner i tightened first . But now my problem is gone , 79C 5.0Ghz 1.33V on Prime95 with spike in the 85C territory , i'm very satisfied ! Idling at 30C with all core locked at 5.0Ghz :) Link to comment Share on other sites More sharing options...
Antixs Posted September 4, 2019 Share Posted September 4, 2019 I have the same CPU, I have never seen a CPU spike so fast in temps. Temps bounce around between 75c/85c coolant reaches 43c. I think best bet would be delid. Link to comment Share on other sites More sharing options...
skyteam78 Posted September 4, 2019 Share Posted September 4, 2019 I have the same CPU, I have never seen a CPU spike so fast in temps. Temps bounce around between 75c/85c coolant reaches 43c. I think best bet would be delid. Yes. From personal experience, delidding worth it. I saw ~ 20 degrees difference under the same testing conditions while testing my overclocked system to 4.8 GHz with Corsair 110i GTX AIO cooler. Link to comment Share on other sites More sharing options...
c-attack Posted September 4, 2019 Share Posted September 4, 2019 The delid on the Coffee Lake 8000 series CPUs is about as high a return in terms of cost/degree of temp reduction there is. No other change you can make will come close to that level of reduction. Link to comment Share on other sites More sharing options...
skyteam78 Posted September 5, 2019 Share Posted September 5, 2019 I tested my fresh hydro x build yesterday on few stress tests and results are very good. I managed to overclock to 5.2 GHz on all cores without any issues and temps are great. Cpu (8700K delided) reached maximum 64C and GPU (EVGA RTX 2080ti) reached maximum 40C. I will continue to test it but for now results are quit impressive. Link to comment Share on other sites More sharing options...
Recommended Posts