F5F Stay Refreshed Power Users Overclocking Utilisation de Prime95 pour analyser les températures inégales du cœur

Utilisation de Prime95 pour analyser les températures inégales du cœur

Utilisation de Prime95 pour analyser les températures inégales du cœur

Pages (2): 1 2 Next
S
Supermannen00
Member
115
09-18-2018, 05:26 AM
#1
I was testing cinebench and achieved fairly consistent temperatures across all cores, with only a 6°C variation between the lowest and hottest cores.
However, with Prime95 there were fluctuations of up to 17°C on two cores, which was completely unexpected.
This setup runs a 9700k at 5GHz, reaching a maximum of 80°C in Prime95 but just 69°C in cinebench.
S
Supermannen00
09-18-2018, 05:26 AM #1

I was testing cinebench and achieved fairly consistent temperatures across all cores, with only a 6°C variation between the lowest and hottest cores.
However, with Prime95 there were fluctuations of up to 17°C on two cores, which was completely unexpected.
This setup runs a 9700k at 5GHz, reaching a maximum of 80°C in Prime95 but just 69°C in cinebench.

X
XCOOLGUY5000
Member
51
10-02-2018, 04:19 AM
#2
ivandanko :
jojesa :
Try
Prime95 version 26.6.
That works fine but the temps are fairly low compared to newer prime95?
Just as it should be, and as expected.
ivandanko
,
Sorry you missed the memo. Do
not
use Prime95 versions
later than
26.6. Here's why:
Not all loads are created equal. “Stress” tests vary widely and can be characterized into two categories;
stability
tests which are
fluctuating
workloads, and
thermal
tests which are
steady
workloads. Intel tests their processors at a
steady
100% TDP workload to validate Thermal Specifications.
Prime95 version 26.6 Small FFT's is ideal for CPU thermal testing
, because it's a
steady
100% workload with...
X
XCOOLGUY5000
10-02-2018, 04:19 AM #2

ivandanko :
jojesa :
Try
Prime95 version 26.6.
That works fine but the temps are fairly low compared to newer prime95?
Just as it should be, and as expected.
ivandanko
,
Sorry you missed the memo. Do
not
use Prime95 versions
later than
26.6. Here's why:
Not all loads are created equal. “Stress” tests vary widely and can be characterized into two categories;
stability
tests which are
fluctuating
workloads, and
thermal
tests which are
steady
workloads. Intel tests their processors at a
steady
100% TDP workload to validate Thermal Specifications.
Prime95 version 26.6 Small FFT's is ideal for CPU thermal testing
, because it's a
steady
100% workload with...

M
MilkIsAwesome
Member
143
10-03-2018, 08:55 AM
#3
Conducting additional trials shows mid high 70s across several cores, with four in the low 40s.
This analysis employs Prime 95s with small FFTs for testing.
M
MilkIsAwesome
10-03-2018, 08:55 AM #3

Conducting additional trials shows mid high 70s across several cores, with four in the low 40s.
This analysis employs Prime 95s with small FFTs for testing.

K
khaledkb_
Senior Member
724
10-03-2018, 12:11 PM
#4
It seems the Intel guidelines might not match the current Intel specs. The Prime95 version might not be fully optimized for the 9th generation yet. Consider using Prime95 26.6 instead. Additionally, Intel allows a 10° C variation in both the highest and lowest core temperatures due to factors like thermal solution application between the IHS and die, or Digital Thermal Sensor inaccuracies. If you notice a 20° C difference, it may indicate an issue requiring a return for replacement.
K
khaledkb_
10-03-2018, 12:11 PM #4

It seems the Intel guidelines might not match the current Intel specs. The Prime95 version might not be fully optimized for the 9th generation yet. Consider using Prime95 26.6 instead. Additionally, Intel allows a 10° C variation in both the highest and lowest core temperatures due to factors like thermal solution application between the IHS and die, or Digital Thermal Sensor inaccuracies. If you notice a 20° C difference, it may indicate an issue requiring a return for replacement.

W
Weider
Junior Member
18
10-03-2018, 12:52 PM
#5
Played a few games and benchmarks, but temperatures stayed normal. It seems like it's just prime performance, though none of the tests required as much CPU as prime. If something went wrong, would it affect other aspects?
W
Weider
10-03-2018, 12:52 PM #5

Played a few games and benchmarks, but temperatures stayed normal. It seems like it's just prime performance, though none of the tests required as much CPU as prime. If something went wrong, would it affect other aspects?

C
coolervanqed
Member
66
10-03-2018, 09:41 PM
#6
Which
version
of Prime95?
C
coolervanqed
10-03-2018, 09:41 PM #6

Which
version
of Prime95?

X
220
10-04-2018, 11:20 AM
#7
CompuTronix:
Which version of Prime95?
29.4b8
I haven't found any other good free CPU tests I've tried, like cinebench, aida64 extreme, and 3dmark demo time spy. All the temperatures are fine.
X
xXStrikeBackXx
10-04-2018, 11:20 AM #7

CompuTronix:
Which version of Prime95?
29.4b8
I haven't found any other good free CPU tests I've tried, like cinebench, aida64 extreme, and 3dmark demo time spy. All the temperatures are fine.

G
Gradually
Junior Member
2
10-04-2018, 12:07 PM
#8
I returned to the standard stock speed (4.8ghz on my motherboard due to the MCE).
It's showing normal temperatures across all cores.
That must have been the overclock that caused it...
G
Gradually
10-04-2018, 12:07 PM #8

I returned to the standard stock speed (4.8ghz on my motherboard due to the MCE).
It's showing normal temperatures across all cores.
That must have been the overclock that caused it...

D
DerChriis
Junior Member
9
10-05-2018, 02:25 AM
#9
Try
Prime95 version 26.6.
D
DerChriis
10-05-2018, 02:25 AM #9

Try
Prime95 version 26.6.

L
LoiZer
Junior Member
44
10-08-2018, 03:44 PM
#10
Try Prime95 version 26.6. It functions properly, though the temperatures are relatively low when compared to more recent versions.
L
LoiZer
10-08-2018, 03:44 PM #10

Try Prime95 version 26.6. It functions properly, though the temperatures are relatively low when compared to more recent versions.

Pages (2): 1 2 Next