The temperatures during overclocking for the Intel i7 7700k are excessively high.
The temperatures during overclocking for the Intel i7 7700k are excessively high.
Earlier, I shared a concern about my processor temperatures reaching excessively high levels at 5.15Ghz with 1.35 volts, recording readings between 87-88°C.
I attempted to push the CPU to 5.0Ghz, which succeeded in achieving a stable overclock at 1.35v. However, the Cinebench r15 results still showed temperatures around 88°C. I adjusted my target back to 5.0Ghz for a more manageable result, though I still saw 87°C.
I own a Kraken x62 AIO water cooler, which makes me wonder if voltage settings or cooling efficiency are the problem. My idle temps range from 35-40°C, and ambient stays between 20-24°C. I expected better performance from the cooler initially but wasn’t sure if it was faulty or just underperforming.
Recently, I managed a steady overclock of 5.05Ghz at 1.28v, but temperatures remained at 88°C. I rely mainly on Cinebench r15 for testing because I don’t want to run long stress tests like Ida64, which could risk damaging the chip. I’ve purchased an Artic Silver 5 cooler to replace the one that came with the Kraken, which is the only option I see as viable besides lowering the clock speed.
I’m also considering adjusting my air conditioner temperature by 2°f, but this has led to freezing conditions.
I don't believe you must "underclock"; simply lower your overclock settings. You'll need a rare chance to get a chip that can consistently run a 4.8ghz or higher boost on any of the recent Intel designs, especially without a top-of-the-line custom loop. Your temperatures are clearly too high—definitely not acceptable. At least, this ensures the processor lasts longer. All relevant details about Intel CPU temperatures are available here:
I don't believe you must "underclock"; simply lower your overclock settings. You'll need a rare chance to get a chip that can consistently run a 4.8ghz or higher boost on any of the recent Intel designs, especially without a top-of-the-line custom loop. Your temperatures are clearly too high—definitely not acceptable. At least, this ensures the processor lasts longer. All relevant details about Intel CPU temperatures are available here:
No, the results don't seem satisfactory. Temperatures exceeding 85°F are problematic, and going above 75°F is not ideal. I'm unsure where you're coming from, but this seems incorrect, particularly if it's not during an extreme thermal test with Prime or another component.
This information comes directly from the Anandtech testing on the 7700k:
At 4.8 GHz, the Core i7-7700K performed well in POV-Ray, but reaching 1.400 volts caused the processor to hit 95°C during OCCT and its mixed AVX tasks. That prompted me to stop it, as the CPU was consuming 122W from idle to load. This low power draw of 122W is surprisingly low—normally we'd expect closer to 160W for similar i7 overclockable models.
The second set of data shows stability at 4.8 GHz and 4.9 GHz, but at 5.0 GHz and 1.425 volts, the CPU entered thermal recovery modes, reflected in the lower POV-Ray scores.
According to what we've heard online, our CPU sample is average to poor for overclocking performance. Some colleagues at motherboard makers are reporting 5.0 GHz at 1.3 volts (with AVX offset), though I'm certain they're not discussing a stable, long-term solution.
The CPU tends to get quite warm naturally, as Intel has stopped using solder connections between the IHS and the CPU die. Doing well without delidding and using liquid metal is a good sign. This is the highest temperature I’ve achieved with my CPU, just a few seconds after the test completed. I’m running it on a Noctua U-14S with an air cooler, which caused the temperatures to fall rapidly. On a very demanding benchmark, my CPU at 5.2GHz ran around 1.47 volts and reached between 74 and 77 degrees Celsius. That voltage is higher than what you used, so your cooler should keep it cooler than mine if you follow the recommended steps from Thermal Grizzly. There are several YouTube guides on this method; some are more effective than others. Thermal Grizzly offers detailed instructions for their product.
It’s worth noting that cutting a small section at the tip of the applicator helps control the flow better. For tighter control, hold the plunger between your thumb and finger at the tube’s “wings” – this gives excellent flow regulation. Otherwise, excess liquid can spray out, which you definitely don’t want. Use only a minimal amount.
Additionally, protect the resistors on the PCB (which will be covered by the IHS) with a thin layer of liquid tape or regular black electrical tape. You can cut a small piece with scissors if needed.
I use delid and relid tools from Rocket Cool
https://rockitcool.myshopify.com/products/rockit-88
I also employ a very thin line of head gasket maker or a high-temperature silicone with an extended applicator tip in the package. I apply it only on two sides of the bottom of the die, not covering the entire edge. This ensures the CPU stays secure while allowing gas to escape through uncovered areas – a key reason Intel leaves a small gap without sealant.
For detailed guidance, refer to Rocket Cool’s instructions available on YouTube.com.
Without using liquid metal, lowering the temperature is unlikely. Be patient, follow the steps carefully, and you should succeed.
EDIT: I forgot to include an image.
No matter how it's done, it remains risky and isn't advised for regular users. If you're familiar with this process professionally or have experience with it, let someone with expertise handle it. For others, it could lead to serious problems.
darkbreeze,
Someone with experience in computer hardware can handle CPU delid and relid tasks confidently. If they follow instructions carefully, Rockit Cool’s guidance will ensure a smooth process. Sticking to the steps is essential.
You’re right that certain individuals shouldn’t proceed with delid. Some online videos demonstrate incorrect methods that could cause issues.
There are also reliable tutorials on YouTube that explain proper procedures. The main challenge lies in distinguishing between correct and incorrect actions, which can be risky for those lacking knowledge.
For those who need clarity, this video is straightforward and accurate: https://www.youtube.com/watch?v=uGSF2mIzSxA
We both agree what counts as someone who has worked with computer hardware enough to feel confident. The average teenager might think they can handle it after swapping a graphics card or installing fans, but that’s unlikely to be true. We’ve talked about this privately before, and many people who have been working with hardware for 20 to 25 years say they wouldn’t feel comfortable taking the risk themselves. Personally, I’m sure I could manage it too, but I prefer to let professionals handle it unless it involves a critical component that could be lost.