Assistance for OC I7-9700K on Aorus Z390 Pro
Assistance for OC I7-9700K on Aorus Z390 Pro
Here are the findings—using prime 95 without AVX at stock speed, the minimum temperature reaches about 35°C (room temperature around 25°C), and the maximum is 70°C. At 4.9GHz with 1.3v, the peak reaches 82°C on three cores, with the rest between 79°C and 80°C. Winter temperatures are roughly 19°C to 29°C, so it isn’t extremely cold. Ambient conditions likely influence your PC’s higher temperatures. However, you still need to consider that. Running at 4.9GHz and 82°C puts you close to the limit; others may differ. I wouldn’t increase the overclock further—stick with the current setting. Testing with prime at those parameters for several hours will confirm stability. Just don’t run it for just 20 minutes. For my own overclocks, I apply them overnight for 8 hours. Basic tests take about 20 minutes. Correction made earlier: it’s actually 1.31v, not 1.3v, and I noticed it reached 84°C on three cores... When I was at 5GHz with 1.31v, it stayed stable for 12 hours. So from 87°C to 5GHz it’s around 84°C—should I lower the frequency to get a cooler run? Is a stock 35°C at those speeds normal? Would reapplying thermal paste help?
Alright, here are the findings:
With prime 95 at stock speed and 35°C ambient, the minimum temperature is around 35°C and the maximum reaches about 70°C. At 4.9GHz with 1.3v, the peak hits 82°C across three cores, with the rest staying between 79°C and 80°C.
Winter temperatures are roughly 19°C to 29°C, so it’s not extremely cold.
Ambient conditions do influence your PC’s CPU temps, but you still need to consider that. Running at 4.9GHz and reaching 82°C puts you close to the limit. Others might have different opinions. I wouldn’t increase the overclock beyond what you’re comfortable with and stick with the current setting. Testing it for a few hours will confirm stability. For my own setup, I run prime overnight for eight hours—20 minutes is enough for basic checks.
Correction: I realized earlier it was 1.31GHz instead of 1.3GHz, which made the temperature jump to 84°C on three cores. At 5GHz and 1.31GHz, it stayed stable for 12 hours. From 87°C down to 5GHz, the temp is around 84°C.
Should I lower the frequency to reach a cooler 84°C? Is a stock 70°C at these speeds normal? Would reapplying thermal paste help?
My take: lowering it to 4.8 and adjusting voltage back to about 1.26–1.27 should keep temps under 80°C while giving room for future improvements if the cooling allows. Stock 70°C is fine for me, and a paste change won’t harm things.
Your first comment confused me—prime didn’t fail on any cores at 95°C! That seems inconsistent.
In short, you have flexibility here. Decide based on your comfort with clock speed versus voltage stability. I always aim to keep max temps under 80°C for a balanced setup.
Roland Of Gilead :
alonbl :
Roland Of Gilead :
alonbl :
here are the findings -
with prime 95 at stock speed, the minimum temperature stays around 35c (my room is roughly 25c), and the maximum reaches 70c.
at 4.9ghz with 1.3v, the peak temp hits 82c across three cores, with the rest hovering between 79c and 80c.
to be honest, winter here is between 19c and 29c, so it's not extremely cold.
so ambient conditions are influencing your PC/CPU temperatures. BUT, you still need to consider that.
at [email protected] and reaching 82c, you're right at the threshold for me. Others might have different views. I wouldn't increase the overclock beyond this point, and stick with what you have. Running tests for a few hours will confirm stability. Don't just test for 20 minutes - for my overclocks, I run them overnight for 8 hours. Basic checks take about 20 minutes.
sorry, I corrected myself earlier – it's actually 1.31v, not 1.3v, and now I see it hit 84c on three cores...
When I was at 5ghz @ 1.31v, it was stable for 12 hours.
so from 87c to 5ghz it's around 84c. Should I lower the frequency to get a cooler run?
Is 70c at stock speeds (prime95) typical? Would reapplying thermal paste help?
This is my take. I'd lower it to 4.8 and voltage back to about 1.26-1.27. This aligns with what most people use (voltage-wise). It should keep temps under 80c while giving you some buffer to possibly improve further if the thermal performance allows.
70c at stock on prime95 is normal for me. You can definitely try reapplying paste – it won't hurt.
Looking back at your first message, I'm a bit puzzled. The main point is that prime didn't fail on any cores at 95c! That doesn't add up.
Anyway, I think you're on the right track. Your final choice between clock speed and voltage really depends on you.
I always aim to keep my maximum overclock below 80c for a balanced approach of voltage, speed, and stability.
1.26v-1.29v with prime95
stable at 1.3v, but I see 84c on one core and 83c on two cores – the rest around 80c.
it doesn't seem like a significant difference compared to 5gz at 1.31v 87c.
when I remove the front panel of the case, I get:
prime95 without axv at 5ghz 1.31v reaches 80c.
prime95 with axv at 5ghz 1.31v reaches 90c
It's really high for me, but it won't damage your processor. Watch the temperatures closely; if things become unstable or crashes happen, turn off the overclocking. Keep in mind that boosting all cores only helps in certain situations.
The gap between 4.8 and 4.9 is almost insignificant when considering real FPS gains in games.
Just ensure your final settings are correct. 80-83c feels quite high for me, though it won't harm your processor. Watch the temperatures closely; if things become unstable or crashes occur, turn off the overclock. Keep in mind that boosting to 4.9 is already beneficial. An all-core overclock only helps in certain situations.
The gap between 4.8 and 4.9 is minimal for real FPS gains in games. I thought a 1.3v boost at 4.8ghz with 84c on two cores would be typical.
Whether it matters if you run 5ghz at 1.31v versus 4.8ghz at 1.3v with 84c is unclear—3 more GHz for 3 cores seems like a reasonable trade-off.
Roland Of Gilead :
Just double-check your final configuration settings. 80-83c feels quite high for me, though it won’t damage your processor. Just monitor the temperatures and if things become unstable or cause crashes, turn off the overclock. Remember, your CPU can still boost to 4.9 anyway—having an all-core overclock only helps in certain situations.
The gap between 4.8 and 4.9 is minimal when it comes to actual FPS gains in games. I expected a need for 1.3v at 4.8ghz with 84c on two cores, isn’t that typical?
Is it really important to run at 5ghz with 1.31v and 87c compared to 4.8ghz at 1.3v and 84c? It seems like a 3-core boost at 5ghz could be a better option.
I think you might have missed the main point. If your goal is to reach the highest clock speeds, go ahead and fine-tune whatever setting feels right for you. For a 100-200mhz increase, plus temperature and voltage changes, it probably isn’t worth it. The key is testing both speeds—you’ll likely see almost no real performance difference. You might see slight improvements in some benchmarks, but that’s about it. So, is this worth risking a potentially unstable system where any gains could be lost due to thermal throttling?
Roland Of Gilead :
alonbl :
Roland Of Gilead :
Hmm, just double-check your final configuration. 80-83c feels quite high for me, but it won’t damage your processor. Just monitor the temperatures and if you notice instability or crashes, turn off the overclock. Remember, your CPU can still boost to 4.9 anyway. An all-core overclock only helps in certain situations.
Yes, the jump from 4.8 to 4.9 is minimal when it comes to actual game performance gains. I expected around 1.3v for a 4.8ghz chip with 84c on two cores—does that sound typical?
Is it really important to run at 5ghz with 1.31v and 87c versus 4.8ghz at 1.3v and 84c? It seems like a 3-core boost at 5ghz could be a better option.
I think you might have missed the key point. If your goal is to hit the highest clock speeds, go ahead and fine-tune whatever setting feels right for you.
For a 100-200mhz increase and a 4-5°C temperature rise plus higher voltage, it doesn’t seem worth it from my perspective.
The main idea is to test both settings and see if there’s any real difference in performance. You might see slight improvements in some benchmarks, but overall it’s unlikely. So, is this worth risking a potentially unstable system where gains could be wiped out by overheating?
I cleaned the dust filter too, and it turns out I only set the fan curve to 100% on all fans when the CPU reached 66°C. Now I keep it at 100% at 60°C.
In AC Odyssey and BF5, temperatures usually stay between 62-66°C with peaks up to 71-72°C. Here’s a screenshot from HWiNFO64:
https://imgur.com/a/IAYQuy3
alonbl :
Roland Of Gilead :
alonbl :
Roland Of Gilead :
Hmm, just double-check your final configuration settings. 80-83c feels quite high for me, though it won’t damage your processor. Just monitor the temperatures and if you notice instability or crashes, turn off the overclock. Keep in mind your CPU can still boost to 4.9 anyway—having a full-core overclock only helps in certain situations.
The gap between 4.8 and 4.9 is barely noticeable when it comes to actual game performance gains. I was expecting around 1.3v for a 4.8ghz chip running at 84c on two cores. Is that typical?
It seems odd that I need 1.31v for a 4.8ghz with 84c on two cores. Could it be normal?
I wondered if running 5ghz at 1.31v versus 4.8ghz at 1.3v with 84c would be better. It looks like pushing the frequency by three cycles could be a smarter choice.
I think you might have missed the key point. If your goal is to hit the highest clock speeds, go ahead and fine-tune whatever setting feels right for you.
For a 100-200mhz increase and a 4-5°C rise in temperature plus higher voltage, it doesn’t seem worth it. The main idea is to test both settings and see if there’s really a performance difference. You might see slight improvements in some benchmarks, but that’s about it.
So, is this worth risking a potentially unstable system where any gains could be lost due to thermal throttling?
I cleaned the dust filter and found out I’d set the fan curve at 100% for all fans only when the CPU hit 66°C. Now I’m keeping it at 100% at 60°C.
In AC Odyssey and BF5, temperatures usually sit between 62-66°C with peaks around 71-72°C. Here’s a screenshot from HWiNFO64:
https://imgur.com/a/IAYQuy3
That’s not too bad—those temps are close to what I expect, just on the edge of being too high.
Generally, this range is acceptable (though it depends on ambient conditions).
Idle: 25-low 30°C
Gaming: 45-65, depending on the game. There are a few exceptions where temperatures can spike to around 70°C.
Stress testing/max load: 70-80°C.
Looking at some discussions, mid-80s for max load seems reasonable. Unless you notice a big drop in performance, stick with 4.8/4.9 at lower voltage and mid-80s temps with Prime and AVX optimizations. Don’t exceed 85°C—see how it goes. Often, just small adjustments can unlock more MHz from a processor.
Roland Of Gilead :
alonbl :
Roland Of Gilead :
alonbl :
Roland Of Gilead :
Hmm, just double-check with your final configurations. 80-83c feels extremely high for me, though it won’t harm your processor. Just monitor the temperatures and if you notice instability or crashes, turn off the overclock. Remember, your CPU can still boost to 4.9 anyway—having a full-core overclock only helps in certain situations.
The gap between 4.8 and 4.9 is barely noticeable for actual FPS gains in games. I thought I’d need 1.3v at 4.8ghz with 84c on two cores, but is that typical?
It seems normal to run 5ghz at 1.31v versus 4.8ghz at 1.3v with 84c. A 3-core boost of 5ghz might not be worth it.
I believe you’re missing the key point. If your goal is maximum clock speeds, go ahead and fine-tune whatever setting feels right for you.
For a 100-200mhz increase and a 4-5°C temperature rise plus higher voltage, it’s probably not worth it. Just test both speeds—you’ll likely see almost no performance difference. A few extra points in a synthetic test might help, but that’s about it.
So, is this worth risking a potentially unstable system where any gains could be lost due to thermal throttling?
I cleaned the dust filter and found out I’d set the fan curve at 100% for all fans only when the CPU hit 66°C. Now I’m keeping it at 100% at 60°C.
In AC Odyssey and BF5, temperatures usually sit between 62-66°C with peaks up to 71-72°C. Here’s a screenshot from HWiNFO64:
https://imgur.com/a/IAYQuy3
That’s not too bad—those temps are close to what I expect, though still near the limit.
Typically, this range is acceptable (though it depends on ambient conditions).
Idle: 25-low 30°C
Gaming: 45-65, depending on the game. A few exceptions exist; some games can push temps high enough to reach 70°C.
Stress testing/max load: 70-80°C.
Looking at other discussions, mid-80s for max load/stress seems reasonable. Unless you notice a significant drop in performance, stick with 4.8/4.9 at lower voltage and mid-80s temps with Prime and AVX optimizations. Don’t exceed 85°C—see how it goes. Often, small adjustments can unlock the last few MHz.
Thanks for your time and effort—it’s appreciated!
alonbl :
Roland Of Gilead :
alonbl :
Roland Of Gilead :
alonbl :
Roland Of Gilead :
Hmm, just double-check your final configuration settings. 80-83c feels quite high for me, though it won’t harm your processor. Just monitor the temperatures and if you notice instability or crashes, turn off the overclock. Remember, your CPU can still boost to 4.9 anyway—having a full-core overclock only helps in specific situations.
The gap between 4.8 and 4.9 is barely noticeable for actual game performance gains. I expected around 1.3v at 4.8ghz with 84c on two cores, but is that typical?
It seems normal to aim for 1.31v at 87c versus 4.8ghz and 1.3v at 84c. A 5ghz setting for 3 cores might be a better choice in some cases.
I think you might have missed the mark. If your goal is top clock speeds, go ahead with whatever setting feels right. For a 100-200mhz increase and a 4-5°C temperature rise plus higher voltage, it’s probably not worth it. Just test both speeds—you’ll likely see almost no real difference in gameplay. A few extra points in a benchmark might be nice, but it’s risky.
You might get slightly better results in some synthetic tests, but that’s about it. The main concern is stability: if you risk instability for marginal gains, it could backfire.
I cleaned the dust filter and found I’d set the fan curve to 100% only when the CPU hit 66°C. Now I keep it at 100% at 60°C.
In AC Odyssey and BF5, temperatures usually sit between 62-66°C with peaks around 71-72°C. Here’s a screenshot from HWiNFO64:
https://imgur.com/a/IAYQuy3
That’s close enough to what I expect, though still near the limit.
Generally, these temps are within acceptable ranges, depending on other factors like room temperature. Idle: 25-low 30°C; Gaming: 45-65°C, depending on the game. Stress tests usually reach 70-80°C.
Looking at discussions, mid-80s is often considered safe for stress scenarios. Unless you notice a big drop in performance, stick with 4.8/4.9 and lower voltage, aiming for mid-80s with Prime and AVX optimizations. Don’t exceed 85°C. See how it goes—sometimes small tweaks can unlock more performance.
Thanks for your time and effort. It means a lot!
No worries. Feel free to reach out if you have more questions.