Temps are too high or stable ?
Temps are too high or stable ?
I just swapped the bigger Intel stock cooler with a noctua NH-c14s large down draft model on my 7700k. I set it in dual fan mode, using the top fan connected through a low-noise adapter after reading a recommendation or hearing it was quieter. In terms of noise, it’s significantly better than the original cooler—gaming used to be very loud, which prompted the upgrade. Now that I have a better cooler, I’m considering a modest stable overclock. My system has an UD3L 270 gigabyte motherboard, and I adjusted the overclock via the auto setting in BIOS after clicking “Upgrade 4.8” on my 7700k. I think this made some pre-set changes, but I didn’t modify anything else in BIOS. My temperatures are around 95°C at 26.6°C, and the Intel testing tool shows a max of 93°C—most experts say staying under 95°C is ideal, as 100°C is the point where throttling starts. Are there any other BIOS settings I should change manually? (It’s up to date). Is this a typical temperature, or should I try another configuration? My case is a quiet DPIC with two small intake fans behind a closed panel, a single small exhaust fan in the back, similar to the Fractal R series design. Would replacing the stock fans with better ones be worthwhile? Could it just mean my chip isn’t performing optimally? Anyone using this cooler on a 7700k and wanting a comparison would be helpful. I experience quick spikes during gaming—chips tend to jump 20-30°C for a second before returning. Generally, I run at 60-80°C while playing Overwatch at 1440p. From what I see online, these temperatures might be a bit high for this cooling setup. On stock settings, I was around 75°C in Prime 95. I’m wondering if someone more experienced can advise on BIOS tweaks, like changing case fans or remounting, or if this is just normal for a chip that’s not up to the task. I don’t want to risk damaging my board by jumping to conclusions—if it’s just a “bad 7700k,” I’m fine at 4.8 and not aiming for 5GHz. I’d prefer to keep temps in the 80s during stress tests at 4.8. Thanks for your time.
I verified the voltage is adjustable and varies with load in prime95 and realbench; it generally stayed between 1.26 and 1.272, reaching a peak of around 1.308. Some users reported values as high as 1.4V on forums. In both Intel Extreme Utility and diagnostic tool versions of Prime95, I found temperatures up to 75-80°C. During the 18-minute stress test on Realbench, I stopped it at 98°C, which caused the back panel airflow to become very hot. The internal components and heat sinks also became warm. Removing the low-noise adapter from the top fan helped maintain similar temperatures. I also changed the BIOS settings to 4.6 and kept stock at 4.5, but temperatures remained around 75-80°C maximum.
what voltages are you operating at? higher voltage leads to increased temperatures.
gaming temperatures appear normal. you're overclocked and using an air cooler. anything above 80 degrees is, in my view, too high for gaming; i'd say it's acceptable if the temperatures stay mostly in the 60s.
the temperature readings are extremely high during stress tests. consider running another stress test like realbench. prime 95 tends to cause high temps on kabylake chips.
I will check when I get home I didn't pick a voltage I just loaded the preset from gigabytes bios for a 4.8g overclock on a 7700k assuming they were at the very least a good jumping point and since I wasn't pushing it however since I'm not happy with the amount of heat I'll see if it's got high voltage. I'm honestly not so much worried about overheating issues in games as I don't think that will happen. I mainly used p95 as that's what I could find the most comparison numbers online for but I can test some of the other programs. My main concerns are as follows. Did I mess up the install on the cooler since my temps are higher than others? Are my case fans providing poor airflow inside my case? Are my bios settings poorly optimized causing the processor to produce more heat under Hess tests than it should? I'll check my volts and try and find some comparisons online and see if the preset from gigabyte is giving it too much juice.
I'll review things when I get back home. I didn't select a voltage because I just imported the preset from gigabytes bios for a 4.8g overclock on a 7700k, hoping it's a decent starting point. Since I'm not pushing it much, and I don't want excessive heat, I'll check if it supports higher voltages. I'm not too concerned about overheating during games, as I believe it won't happen. I mostly used p95 because that's what I found the most comparisons for online. I can try other programs too. My main worries are: did I install the cooler incorrectly because my temps are higher than others? Are my case fans providing insufficient airflow inside the case? Are my BIOS settings not optimized, causing the processor to generate more heat during Hess tests than necessary? I'll verify the voltage and look for comparisons online to see if the gigabyte preset is delivering too much power. First, let's check the voltage—it will give us a clear direction to work from. I don't think anything went wrong with the installation since my gaming temperatures seem normal. Also, try a different stress test besides prime95. If those temperatures stay high, we can shift focus to the fans and other aspects. If they improve with another test like realbench, then everything is okay. For now, just check the voltage and run another stress test.
I examined the voltage, which is adjustable and varies depending on the load. In prime95 and realbench, it remained around 1.26 to 1.272 most of the time... it reached up to 1.308, which is not unusually high. Some discussions on forums indicate values as high as 1.4v. While I see readings between 75-80 in both Intel Extreme Utility and diagnostic tool when using Prime95 with 95C+, I disabled the stress test at 18 minutes, reaching 98°C. At that point, the air from behind the PC became extremely hot. Following that, both the interior and heat sinks were also very warm. When I removed the low-noise adapter on the top fan of the heat sinks, the temperatures stayed similar. I changed the BIOS settings to 4.6 and kept the stock at 4.5 temperatures across all components, which remained between 75-80°C maximum. Back to 4.8 and 90s again in both Prime95 and realbench. It appears the heatsink is getting a lot of heat, indicating it’s working properly. The stock case fans aren’t providing much airflow when CPU and GPU are close together near the case exhaust. There’s a significant amount of hot air at the top back of my case during these tests, and I can even feel it through the case’s top. At idle, temperatures drop to 27-29°C, and the air behind the case feels cool, though the pressure doesn’t rise enough to move a tissue. Given that I don’t want to open the lid next, installing new case fans seems like the best option unless you have custom BIOS modifications. I have 32GB RAM with four 8GB sticks, all configured with an XMP profile set to 3000, which according to some research can slightly lower temperatures but also places the downflow cooler over the RAM, providing extra cooling. Still, I haven’t reached temperatures above 90°C during gaming. After two sessions of Overwatch at max settings in 1080p, core temps never exceeded 86°C, which is about 8-10 degrees higher than what I see with stock running at full settings. I felt hot air coming from behind the case at the end of a session, something I didn’t experience with the stock clocks. The voltage for stock 4.6 and 4.8 profiles seems consistent. Thoughts?
I verified the voltage is adjustable and varies with load in prime95 and realbench; it typically stayed around 1.26–1.272, reaching a peak of about 1.308. Some users reported values as high as 1.4V on forums. In both Intel Extreme Utility and diagnostic tool versions from Prime95, I reached 75–80°C in both the Extreme Utility and diagnostic tool versions of Prime95, which Prime95 pushed to 98°C at 18 minutes. At that stage, the airflow from the back of the PC became very intense, and the internal and heat sink temperatures also rose. I removed the low-noise adapter on the top fan of the heatsink, which helped keep temperatures stable. I also changed the BIOS settings to 4.6 and kept the stock at 4.5, maintaining temperatures between 75–80°C maximum. After resetting everything back to 4.8 and 90s speeds, temperatures returned to normal. The heatsink was clearly getting a lot of heat, indicating it was functioning properly. The stock case fans didn’t provide much airflow when CPU and GPU were close together near the exhaust, causing significant heat buildup at the top of the case—sometimes I could feel it through the case’s front. At idle, temperatures stayed around 27–29°C, while the air behind the case felt cool but not enough to move a tissue. Given these results, it seems the cooling system is working, though it may need some improvement. Installing additional case fans would be a good next step unless you have custom BIOS modifications. I have 32GB RAM with four 8GB sticks, all configured with an XMP profile set to 3000, which can push temperatures up a few degrees but still leaves room for improvement. The downdraft cooler is positioned over the RAM, so it should provide extra cooling. I haven’t reached 90+ during gaming sessions, and after two Overwatch matches at max settings, core temps never exceeded 86°C—still about 8–10 degrees higher than stock performance at full resolution. I felt hot air near the end of a session, which wasn’t an issue with the stock clocks. The voltage for both 4.6 and 4.8 profiles appears consistent. What are your thoughts?
Some say you might have missed out on the best performance...
Your CPU tends to run much hotter than others—very hot.
The main advice I can offer is to ensure your fan direction is optimized: cold air should enter, and hot air should be expelled. Consider adding a stronger case fan to push cooler air into the system. Also, try increasing the CPU fan speed when temperatures rise.
My final tip would be to re-seat the heatsink and apply fresh thermal paste if needed.
Good luck!
I'll check out microcenter to find some improved case fans. If that doesn't work, I'll reposition the cooler, though I doubt it's the main problem since the heatsink was still warm after testing. I've noticed few online reports of 10°C drops just by changing case fans. It seems my chip is already very hot—once I get an OC, I was hoping for a quick fix with some BIOS tweaks. What's strange is that the heatsink is getting hot, which should help move heat away from the die and into the pipes. If that's happening, then a delid might not be much of a solution, even though it seems to improve heat transfer. Maybe I'm missing something here?
This is why simply copying another Joe User's settings isn't always ideal. Even with identical hardware, each part differs slightly, not just the CPU. Are you using a consistent vCore or are you opting for Auto? It looks like Auto mode is active since your voltage shouldn’t change much if set manually. I advise reviewing discussions about overclocking and manual BIOS adjustments. Gaining insight into these steps helps with performance, hardware boundaries, and troubleshooting. Also, which Prime95 version are you employing? This significantly impacts results. Be sure to fully grasp every detail in this thread before proceeding with CPU delid, as even a correct delid can lead to similar problems if the BIOS or Prime95 version is flawed.
@ Rubix maybe I should have been more explicit about my objectives for this project, as it appears I may have misled you on several points. Initially, my primary aim is achieving a stable performance of 4.8g with a variable voltage adaptive mode activated. Additionally, I need temperatures to stay below the 80s range during stress testing. I’m not aiming to damage my chip; I was just wondering if it might be due to poor thermal conductivity in my component, and whether this aligns with what more experienced individuals are experiencing or if the fact that my heatsink appears to be functioning properly is sufficient.
I used Prime95 26.6 and didn’t think an offset was necessary, though I might have considered adding a 2-3 offset just in case.
Updates from yesterday: I swapped out the three pre-installed case fans for higher-rated models with double the CFM and set them to maximum in the BIOS. Since I was already inside the case, I also reinstalled the Noctua cooler. The case temperature decreased, and overall CPU temperatures dropped significantly at lower frequencies. On a low-voltage desktop setting, idle temps reached 26-32°C, while under load during stress tests, everything remained stable—reaching up to 80°C on a real bench and Prime95 XTU reaching similar highs, capping around 79°C. During two Overwatch games, my temperatures stayed between 40-50°C, never exceeding the low 70s based on core temperature readings after several sessions. The cooler airflow from the back of the case was noticeable. Up to 4.8g sustained performance was stable, with temperatures rarely spiking above 80°C. The cooler airflow helped a lot. For gaming at 4.8g, I experienced spikes up to 95°C, but realbench blue screens appeared only about two minutes before completing my 30-minute stress test, with most crashes around 80-86°C. Core temperatures peaked at 97°C before I encountered a blue screen.
Trying the failure Prime95 at 4.8g still resulted in high readings, and Realbench’s blue screens occurred just before finishing my test, which lasted about 30 minutes. Core 3 reached its maximum temperature during spikes, while Core 1 stayed within safe limits. Eventually, I searched for guides or videos on adjusting adaptive voltage settings, but found fewer than expected and had to stop working late.
It seems most users opt for variable voltage, which is understandable but frustrating since I want my chip to shut down smoothly at the desktop and only ramp up when necessary. My next goal is to experiment with adjusting the maximum voltage in adaptive mode to see if a slight reduction brings stability back. It was impressive to see that going from stable performance at 4.6g to 95+g could drop to just over 80°C, with only a 200MHz increase on the real bench. Core 3 didn’t improve as much at this high setting, while Core 1 hit its peak at 97°C before I encountered a blue screen. I’m planning to test whether lowering the adaptive voltage further can help stabilize performance. Also, checking the Avx offset might assist in reducing heat during the stress test.