The stress test results show if these are within safe maximum temperatures.
The stress test results show if these are within safe maximum temperatures.
I increased my i5 6600k to 4.6ghz using 1.35 voltage and performed a realbench stress test for 15 minutes with 8GB memory. The images below show the high temperatures, reaching up to 90 degrees, with an average of around 80-85 degrees. This is normal for the performance level achieved. Is this safe or should I be concerned?
My setup:
i5-6600K 3.5 GHz
Cooler Master Hyper 212 EVO
Asus Z170-A ATX LGA1151
Kingston HyperX Fury Black 16 GB (2 x 8 GB) DDR4-2133 Memory
Crucial BX100 250 GB 2.5" SSD
Western Digital Caviar Blue 1 TB 3.5" 7200RPM HDD
MSI Radeon R9 390 8 GB Video Card
NZXT S340 ATX Mid Tower Case
SeaSonic S12II 620 W 80+ Bronze Certified PSU
Check the links for more details: https://imgur.com/F00osEi, https://imgur.com/pM4Mph5, https://imgur.com/nQcMLLu
In the best scenario, a 2-4c difference is possible; in the worst, you might lose $10 on paste. Don't misunderstand, I personally use tg, but it doesn't decide everything. The 212 is what's limiting this project, according to me. The voltage looks high too, but maybe just a silicon variation. My 6600k could handle 4.6 at 1.25, but it needs over 1.4 for stable 4.7. If you can afford a cooler, get one better. For air cooling, the d15 is unmatched, though it's expensive. You might want to consider the artic freezer 34 duo.
What type of heatsink or CPU cooler are you employing? If you're using the original Intel part, it's not recommended to overclock it. An aftermarket option would be more suitable for this task.
Cooler Master Hyper 212 EVO 82.9 CFM Sleeve Bearing CPU cooler with improved performance
90C is definitely comfortable at just 4.6 GHz, though I'm more used to checking the temperatures on a DH-15 than those of a 212... Is the core voltage (1.34) necessary, or should it be adjusted manually? Or was the mainboard configured with an OC preset? It seems 1.34V might be more than what's needed, which adds a lot of heat; try lowering it to 1.3V then 1.27V and see how stability and temperatures respond.
Purchase a thermal grizzly kryonaut that is non-conductive, then thoroughly clean the CPU die and Heatsink copper using isopropyl alcohol. Allow it to dry completely before applying the thermal paste (ensure it is super clean and shiny). Apply with moderate mounting pressure. Conduct additional thermal tests.
In the best scenario, a 2-4c difference is possible; in the worst, you might lose $10 on paste. This doesn't mean I rely on tg, it's just not critical. The 212 model is what’s limiting this project, according to me. The voltage level seems high, but maybe it's just a matter of silicon quality. My 6600k could handle 4.6 at 1.25, but stability would need over 1.4. If budget allows, get a better cooler. For air cooling, the d15 is top-notch but expensive. You might consider the artic freezer 34 duo—it's around $35 and offers 200W compared to the 212’s 150W. If lower voltage is an option, go for it. But if that’s not feasible, look for a cooler with better specs.
It really depends on whether he chose the most affordable materials or installed it poorly; there could be a ten-cent difference. I think the new cooler is a good idea, as I've seen many of them with air coolers and they're much cheaper while still providing similar cooling unless you go for an extremely custom build that costs a lot.