I've got a small form factor PC that I travel with that has an i7-4790K CPU (at stock speed). It's being cooled by a Big Shuriken 2 Rev. B air cooler. When run at 100% load, the CPU has peak core temperatures of 82C (core 1), 74C (core 2), 70C (core 3), and 66C (core 4) with the steady-state values just below these temperatures--for instance, core 1 will bounce between 79C and 82C. These are the temperatures reported by HWMonitor and CAM. CAM also reports the average, which is usually around 77C.
I've tried remounting the Big Shuriken as well as other coolers on this CPU (Noctua NH-D9L, Raijintek Pallas, Cryorig C7, and Cryorig M9i both in cases and on open bench set ups) and still get this behavior--where there is quite a bit of difference in the core temperatures.
I've tried running Voffset of -0.025, -0.050, etc., and it's only stable at -0.025 V (not much difference on peak temperatures). If I disable turbo in the BIOS, I can keep the peak temperature below 80C, but still have the significant difference in core temps.
My i7-8700K doesn't exhibit this behavior... with differences in peak core temperatures less than 5C.
I'm sure that the CPU could really benefit from a delid, but I haven't gotten the nerve to do it yet. Will one core running above 80C shorten the CPU's life, or is the average more indicative of lifespan?
I've tried remounting the Big Shuriken as well as other coolers on this CPU (Noctua NH-D9L, Raijintek Pallas, Cryorig C7, and Cryorig M9i both in cases and on open bench set ups) and still get this behavior--where there is quite a bit of difference in the core temperatures.
I've tried running Voffset of -0.025, -0.050, etc., and it's only stable at -0.025 V (not much difference on peak temperatures). If I disable turbo in the BIOS, I can keep the peak temperature below 80C, but still have the significant difference in core temps.
My i7-8700K doesn't exhibit this behavior... with differences in peak core temperatures less than 5C.
I'm sure that the CPU could really benefit from a delid, but I haven't gotten the nerve to do it yet. Will one core running above 80C shorten the CPU's life, or is the average more indicative of lifespan?
Last edited: