F5F Stay Refreshed Power Users Overclocking Can I use 85°C fine for CPU overclocking?

Can I use 85°C fine for CPU overclocking?

Can I use 85°C fine for CPU overclocking?

Pages (3): 1 2 3 Next
X
197
03-23-2025, 02:58 PM
#1
I own an i3-8350K that's functioning well at 4.5GHz* with 1.325V on Prime95 Blend for three hours prior to my PSU failing (likely due to a faulty unit). *Technically it could reach 4.8GHz with AVX, but Prime95 Blend was using AVX which limited performance to 4.5GHz. I planned to assess stable non-AVX speeds before the failure occurred.*
The peak temperatures hit 95°C for 5-10 seconds during stress tests, with overall fluctuations between 82°C and 92°C. This is considered stress testing only; normal gaming should stay around 80-85°C or lower.
Is it safe to maintain a temperature of 85°C+ while gaming? Would this accelerate failure significantly, potentially shortening the device's lifespan to just a few years instead of its expected six years?
Additionally, could disabling a core reduce the temperature at the same voltage? Or would lowering the voltage and conducting further tests be necessary to confirm stability?
X
x_HmOoDy_KsA_x
03-23-2025, 02:58 PM #1

I own an i3-8350K that's functioning well at 4.5GHz* with 1.325V on Prime95 Blend for three hours prior to my PSU failing (likely due to a faulty unit). *Technically it could reach 4.8GHz with AVX, but Prime95 Blend was using AVX which limited performance to 4.5GHz. I planned to assess stable non-AVX speeds before the failure occurred.*
The peak temperatures hit 95°C for 5-10 seconds during stress tests, with overall fluctuations between 82°C and 92°C. This is considered stress testing only; normal gaming should stay around 80-85°C or lower.
Is it safe to maintain a temperature of 85°C+ while gaming? Would this accelerate failure significantly, potentially shortening the device's lifespan to just a few years instead of its expected six years?
Additionally, could disabling a core reduce the temperature at the same voltage? Or would lowering the voltage and conducting further tests be necessary to confirm stability?

A
anto2003roma
Member
80
03-23-2025, 02:58 PM
#2
I believe our "frog" is missing the main idea. Let me clarify things for you.
First, you're not a moderator, which means you can't establish rules or guidelines, and you definitely aren't allowed to moderate the moderation team members. If you're not happy with that, you can reach out to a member of the forum staff. Alternatively, you can simply relax and pretend you're part of the group, or leave.
Second, Prime95 version 26.6 is one of the very few programs available (or those that mimic Prime95 without Linpack, x264 encoding, or AVX instructions) that consistently runs at a steady state workload. This sets it apart from other programs you were discussing, which typically rely on different techniques.
A
anto2003roma
03-23-2025, 02:58 PM #2

I believe our "frog" is missing the main idea. Let me clarify things for you.
First, you're not a moderator, which means you can't establish rules or guidelines, and you definitely aren't allowed to moderate the moderation team members. If you're not happy with that, you can reach out to a member of the forum staff. Alternatively, you can simply relax and pretend you're part of the group, or leave.
Second, Prime95 version 26.6 is one of the very few programs available (or those that mimic Prime95 without Linpack, x264 encoding, or AVX instructions) that consistently runs at a steady state workload. This sets it apart from other programs you were discussing, which typically rely on different techniques.

A
Aleshop777
Member
153
03-23-2025, 02:58 PM
#3
My temperatures are significantly higher than what I find acceptable. For continuous daily operation, keeping the CPU below 70°C is ideal, aiming for around 65°C. During stress tests, I’m comfortable with about 80°C, but never exceeding that. Reaching 95°C during testing and tolerating a gaming temperature of 80°C isn’t something I’d permit in my own system. This is ultimately a personal preference—what suits me might not suit others.
https://www.intel.com/content/www/us/en/...8350k.html
Officially, the Tjunction rating is 100°C. (Tjunction differs from Tcase, as the junction temperature is higher.) In practice, you’re staying under the limit, though there’s a minor gap between 95°C and 100°C.
A
Aleshop777
03-23-2025, 02:58 PM #3

My temperatures are significantly higher than what I find acceptable. For continuous daily operation, keeping the CPU below 70°C is ideal, aiming for around 65°C. During stress tests, I’m comfortable with about 80°C, but never exceeding that. Reaching 95°C during testing and tolerating a gaming temperature of 80°C isn’t something I’d permit in my own system. This is ultimately a personal preference—what suits me might not suit others.
https://www.intel.com/content/www/us/en/...8350k.html
Officially, the Tjunction rating is 100°C. (Tjunction differs from Tcase, as the junction temperature is higher.) In practice, you’re staying under the limit, though there’s a minor gap between 95°C and 100°C.

A
AGLOS6
Member
184
03-23-2025, 02:58 PM
#4
It all comes down to the specific CPU model. For Intel processors, an average temperature of 85°C feels a bit too high for my needs when using a daily workload. I prefer keeping my temps under 80°C. Spikes around 80°C are acceptable, but I want to maintain lower averages. Reaching 95°C triggers thermal throttling—this is the CPU slowing down to avoid harm. Prolonged exposure to such high temperatures can significantly reduce its lifespan.

I wouldn’t recommend using Prime95 as a benchmark. There are a couple of reasons: it could potentially harm your CPU and possibly damage your power supply unit. Another reason is that no other program can replicate the stress that Prime95 applies. Therefore, Prime95 isn’t a reliable indicator of real-world performance. If you’re aiming for a stable overclock for gaming, I suggest using Intel XTU. Since it’s built for Intel chips, your CPU is optimized to run there, and you can monitor temperatures directly from applications.

XTU won’t force your CPU to work as intensely as Prime95, but neither will any game push it that hard. There are many online reports suggesting Prime95 has caused CPU damage in the past. For my experience, I’d opt for a different benchmarking tool instead.
A
AGLOS6
03-23-2025, 02:58 PM #4

It all comes down to the specific CPU model. For Intel processors, an average temperature of 85°C feels a bit too high for my needs when using a daily workload. I prefer keeping my temps under 80°C. Spikes around 80°C are acceptable, but I want to maintain lower averages. Reaching 95°C triggers thermal throttling—this is the CPU slowing down to avoid harm. Prolonged exposure to such high temperatures can significantly reduce its lifespan.

I wouldn’t recommend using Prime95 as a benchmark. There are a couple of reasons: it could potentially harm your CPU and possibly damage your power supply unit. Another reason is that no other program can replicate the stress that Prime95 applies. Therefore, Prime95 isn’t a reliable indicator of real-world performance. If you’re aiming for a stable overclock for gaming, I suggest using Intel XTU. Since it’s built for Intel chips, your CPU is optimized to run there, and you can monitor temperatures directly from applications.

XTU won’t force your CPU to work as intensely as Prime95, but neither will any game push it that hard. There are many online reports suggesting Prime95 has caused CPU damage in the past. For my experience, I’d opt for a different benchmarking tool instead.

B
Barbieee
Junior Member
17
03-23-2025, 02:58 PM
#5
@feelinfroggy777
I don't rely on Prime95 for performance checks. It's mainly for checking stability.
B
Barbieee
03-23-2025, 02:58 PM #5

@feelinfroggy777
I don't rely on Prime95 for performance checks. It's mainly for checking stability.

F
FinalHavoc
Member
59
03-23-2025, 02:58 PM
#6
GoldenSun3DS :
@feelinfroggy777
I don't rely on Prime95 for benchmarking. I use it just to check stability. The key here is that stability in Prime95 doesn't really apply to gaming. No game or software will stress your CPU like Prime95 does. So why bother? It's similar to training for a 100m dash but competing in a decathlon. For a proper stability test, try the XTU stress test or Aida64. These methods provide genuine results and won't harm your CPU, unlike Prime95.
F
FinalHavoc
03-23-2025, 02:58 PM #6

GoldenSun3DS :
@feelinfroggy777
I don't rely on Prime95 for benchmarking. I use it just to check stability. The key here is that stability in Prime95 doesn't really apply to gaming. No game or software will stress your CPU like Prime95 does. So why bother? It's similar to training for a 100m dash but competing in a decathlon. For a proper stability test, try the XTU stress test or Aida64. These methods provide genuine results and won't harm your CPU, unlike Prime95.

R
155
03-23-2025, 02:58 PM
#7
The key is that stability in prime95 means no problems in games. You push the limits to see how it performs. If it holds up then it should work consistently.
R
ReisingerJocke
03-23-2025, 02:58 PM #7

The key is that stability in prime95 means no problems in games. You push the limits to see how it performs. If it holds up then it should work consistently.

G
GC_Lewk
Member
219
03-23-2025, 02:58 PM
#8
The key is that stability in prime95 means no problems with games. You push the limits of what you can imagine. If it holds up there, it should hold up everywhere.
G
GC_Lewk
03-23-2025, 02:58 PM #8

The key is that stability in prime95 means no problems with games. You push the limits of what you can imagine. If it holds up there, it should hold up everywhere.

S
Seve_PT
Member
229
03-23-2025, 02:58 PM
#9
Is it really stable? Prime will push your CPU to full capacity. A few tests that stop at 95 or 98% aren't truly stable according to my standards. If your system remains cool and performs well while running Prime, you're good to go. Gaining a bit more performance isn't validating in my view.
S
Seve_PT
03-23-2025, 02:58 PM #9

Is it really stable? Prime will push your CPU to full capacity. A few tests that stop at 95 or 98% aren't truly stable according to my standards. If your system remains cool and performs well while running Prime, you're good to go. Gaining a bit more performance isn't validating in my view.

B
badman394
Member
221
03-23-2025, 02:58 PM
#10
Unless you're under intense pressure, for a test, you're merely failing to meet the requirements. For the OP...a steady 85°C during regular use (like gaming) is significantly higher than what I deem 'safe'.
B
badman394
03-23-2025, 02:58 PM #10

Unless you're under intense pressure, for a test, you're merely failing to meet the requirements. For the OP...a steady 85°C during regular use (like gaming) is significantly higher than what I deem 'safe'.

Pages (3): 1 2 3 Next