Aaaaaargh! I'm stuck (again). I'm trying to o/c a 3570k using a Cooler Master Seidon 240V, ASRock Z77 Extreme 6 and 32GB 1866 MHz RAM (if that matters...?) but cannot seem to get it off the ground. It runs at 72-75C in Prime 95 smallFFT at stock of 3.6 GHz, I've tried manually adjusting the settings in the UEFI but have had no luck keeping it below 105C, so have resorted to trying the "Optimised CPU O/c" settings in the UEFI. 4.8, 4.6 and 4.4 wouldn't boot. 4.2 booted and ran Prime at high temps, then dropped to 84C, so I thought the problem was solved but then noticed worker #2 had stopped. So my questions are: is 105C normal when stress testing Prime? is the RAM a blocker? is it just a case of the chip not playing ball? or is the cooler not doing its job/requiring reseating and repasting? System has been built from scratch, but with second hand parts from different sources. All seems to work under normal conditions though. Grateful for any assistance!
Sounds like be a mounting problem with the cpu cooler. I had a 3570K that I ran at 4.3 on a 120mm aio and max temps would be between 60-70c depending on room temperature. What vcore and llc were you using for the overclock? Ram shouldn't be an issue, especially if the 1866 is it's xmp profile? I had a set of Ripjaws X that was 1600, but with really tight timings and I had it running at 2133 on an asrock extreme 4.
If you haven't delidded then the voltage required to get the speed you mention will have you hitting 100+ and thermal throttling, you need to back it off, once delidded and proper thermal solution between core and spreader it will handle no problem. I experienced similar when I had a 3570k, once delidded on water I could run 4.9 @ 1.5v and never see >90. In fact I am seeing the same with my daughters machine at the moment, I have a Ryzen 2400G and it seems the APUs are the only Ryzen not to get Iridium solution between core and spreader, so maxing out CPU core and GPU core it is priming at 100C (it will thermal throttle at 95) so I have accepted that I cannot do a CPU overclock and GPU overclock and also that such loads will rarely happen in real life scenarios playing kids games, so have just gone for GPU so that 1080p60 is a realistic proposition and undervolted the CPU to keep thermals in check but really its crying out for a delid and a proper cooling solution but then it’s no longer a cheap gaming system, doesn’t help that the case is only as big as the motherboard and standard cooling.
Just to close out this thread in case anyone stumbles across it in the future and can't contain themselves because there is no resolution to the tale, I thought I'd advise that it was indeed a refit of the cooler that was required! I also took the opportunity to upgrade the paste to some Thermal Grizzly, so the chip is now sitting at a steady 35C at 4.6GHz as I type (admittedly not under massive load). Burn tested in Prime, Kombustor and Unigine with the GPUs as well, plus a few games and seems good (touch wood) - no delid required. From memory, it was running ~80-82C in Prime at 1.32v. Case closed! Many thanks to all above for "chipping" in - much appreciated.
@ModSquid nice of you to revisit this thread, most of the time the internet is just full of unresolved questions
Yeah, I find that a lot! And thanks for the props - I do try to at least tie up my own "loose threads" (oh dear....) otherwise I feel we're not all helping each other. Sense of community and all that. Speaking of which, may I direct you this way, Sir, to my latest emergency about my system not shutting down in proper fashion...? ---->
