blimpboy3 Posted September 7, 2020 Share Posted September 7, 2020 I got a replacement H100i a few months ago and things were running great. The pump speed at quiet was in the 1800s while performance was in the 2800s. Now I've noticed that the pump speed has increased to 2000s and 3000s respectively, along with an increase in CPU temperature. Has this happened to anyone else? I've tried rotating the entire system to clear any blockage but this hasn't helped. Could there be some permeation causing the liquid level to drop and reduce the resistance for the pump, resulting in the increased speed? Link to comment Share on other sites More sharing options...
c-attack Posted September 9, 2020 Share Posted September 9, 2020 How much of an increase in CPU temperature? Does this show up as a coolant temperature increase as well (H100i Temp)? Link to comment Share on other sites More sharing options...
blimpboy3 Posted September 10, 2020 Author Share Posted September 10, 2020 Approximately 6-10C increase. Coolant temp is still roughly the same, and all other system parameters are the same including ambient temperature. There's not a spec of dust on the radiator either. This was tested using an 8700k at 5ghz, 1.32v with pump at max speed and fans at 80% while running P95 blend no AVX. Link to comment Share on other sites More sharing options...
c-attack Posted September 10, 2020 Share Posted September 10, 2020 Sometimes you see a increase in pump speed when there is a mild blockage, prior to the point where it becomes so bad the cooler is unusable. However, I would expect to see a coolant temp rise to match the CPU temp rise if there was a direct connection. I also would be careful with using Prime 95 as a diagnostic tool for cooler performance. It really puts the limit on the CPU/voltage side. It's possible there is an actual flow rate slow down through the CPU block resulting in higher temperatures, but difficult to distinguish from any of the dozen other things that can cause a change on the other side. Do your normal stuff, test with something mild and constant load like the CPU-Z 'Bench' stress test so you can keep it going to assess coolant temp rise. If there really is a flow issue, it's going to be present in all of these situations and the coolant temp changes more clearly distinguished over a longer period of time. Link to comment Share on other sites More sharing options...
blimpboy3 Posted September 13, 2020 Author Share Posted September 13, 2020 Ended up getting an RMA. Link to comment Share on other sites More sharing options...
Recommended Posts