F5F Stay Refreshed Power Users Overclocking FX-6300 Overclocking on Gigabyte-990FXA R5 Causes Temperature Problems

FX-6300 Overclocking on Gigabyte-990FXA R5 Causes Temperature Problems

FX-6300 Overclocking on Gigabyte-990FXA R5 Causes Temperature Problems

L
Laxioss
Junior Member
5
01-29-2017, 09:06 PM
#1
Okay so I have been playing around with overclocking my FX-6300 (FIRST TIMER) for a while now and I am having some strange temp issues. My package temp while stress testing is ALWAYS higher than the socket temp...and from what I have read this is not the norm. Currently I have achieved a somewhat stable OC at 4.3 GHZ and ~ 1.39V. I have disabled Core Performance Boost and disabled power saving (C&Q, C1E, SVM, C6, APM). In that same page I left CPU unlock to enabled and HPC mode enabled as well. I did set the NB Core voltage to 1.2V and the CPU PLL to 2.65V, these setting seemed to help a bit with bringing my voltage down on my cpu. I have also set LLC to medium as with higher settings at load my voltage would spike pretty high. I am keeping my RAM at 1600 MHZ, not overclocking it at all at the moment. My clock is achieved purely by increasing the clock no FSB manipulation, as i have tried messing with this and could not achieve anything stable at high clocks. I have kept the NB freq and HT link freq at stock.
So for the temps... at prime95 for 10-20 mins at this current clock my CPU package temp in HWmonitor is getting up to 61C but the socket temp is usually 5-7 degrees lower than this at around 55-57C. I have tried pushing to 4.4 GHZ at 1.425V but the package temp got up around 64C on the package and I couldn't keep it stable and I'm not sure if that was due to temps or not enough voltage. I do know that the my case (Corsair 200R) is known for not have great cooling but I did add a 120 mm front intake and 1 140mm bottom intake and 1 140MM top exhaust right above cpu cooler to help with air flow.
My current build is:
CPU: AMD FX-6300
MOBO: Gigabyte 990FXA-UD3-R5
CPU Cooler: 212 EVO
GPU: NVIDIA 660Ti
RAM: 2X4 GB Corsair Vengeance 1600
PSU: EVGA 600W 80+ Bronze
Case: Corsair 200R
I do not currently have any screen shots or pictures of bios but I can upload some later if they would help.
Any thoughts or advice would be much appreciated.
L
Laxioss
01-29-2017, 09:06 PM #1

Okay so I have been playing around with overclocking my FX-6300 (FIRST TIMER) for a while now and I am having some strange temp issues. My package temp while stress testing is ALWAYS higher than the socket temp...and from what I have read this is not the norm. Currently I have achieved a somewhat stable OC at 4.3 GHZ and ~ 1.39V. I have disabled Core Performance Boost and disabled power saving (C&Q, C1E, SVM, C6, APM). In that same page I left CPU unlock to enabled and HPC mode enabled as well. I did set the NB Core voltage to 1.2V and the CPU PLL to 2.65V, these setting seemed to help a bit with bringing my voltage down on my cpu. I have also set LLC to medium as with higher settings at load my voltage would spike pretty high. I am keeping my RAM at 1600 MHZ, not overclocking it at all at the moment. My clock is achieved purely by increasing the clock no FSB manipulation, as i have tried messing with this and could not achieve anything stable at high clocks. I have kept the NB freq and HT link freq at stock.
So for the temps... at prime95 for 10-20 mins at this current clock my CPU package temp in HWmonitor is getting up to 61C but the socket temp is usually 5-7 degrees lower than this at around 55-57C. I have tried pushing to 4.4 GHZ at 1.425V but the package temp got up around 64C on the package and I couldn't keep it stable and I'm not sure if that was due to temps or not enough voltage. I do know that the my case (Corsair 200R) is known for not have great cooling but I did add a 120 mm front intake and 1 140mm bottom intake and 1 140MM top exhaust right above cpu cooler to help with air flow.
My current build is:
CPU: AMD FX-6300
MOBO: Gigabyte 990FXA-UD3-R5
CPU Cooler: 212 EVO
GPU: NVIDIA 660Ti
RAM: 2X4 GB Corsair Vengeance 1600
PSU: EVGA 600W 80+ Bronze
Case: Corsair 200R
I do not currently have any screen shots or pictures of bios but I can upload some later if they would help.
Any thoughts or advice would be much appreciated.

A
AFKCosmos
Member
183
01-30-2017, 04:26 AM
#2
I have a similar configuration but with FX6350BE, which cost 20 dollars more. It operates at 4.713 GHz continuously on a mix of multiplier and FSB at 205 MHz, with a voltage of 1.325v. I've tried using an Evo 212 but it didn't perform well. Now, with a Mugen3 cooler and two fans, it's not reaching over 55°C per core. Since other tools like HW Monitor give inaccurate readings, I rely on the AMD Overdrive program for monitoring. Core temperatures seem more reliable than package readings, so sticking with core temps makes sense.
A
AFKCosmos
01-30-2017, 04:26 AM #2

I have a similar configuration but with FX6350BE, which cost 20 dollars more. It operates at 4.713 GHz continuously on a mix of multiplier and FSB at 205 MHz, with a voltage of 1.325v. I've tried using an Evo 212 but it didn't perform well. Now, with a Mugen3 cooler and two fans, it's not reaching over 55°C per core. Since other tools like HW Monitor give inaccurate readings, I rely on the AMD Overdrive program for monitoring. Core temperatures seem more reliable than package readings, so sticking with core temps makes sense.

J
joeyman108
Member
56
01-30-2017, 05:39 AM
#3
What were your temperatures during the EVO? I believe I might have experienced a faulty chip, as my device couldn't reach that high performance at such low voltage.
J
joeyman108
01-30-2017, 05:39 AM #3

What were your temperatures during the EVO? I believe I might have experienced a faulty chip, as my device couldn't reach that high performance at such low voltage.

M
MasterDarky
Junior Member
20
01-30-2017, 07:15 AM
#4
It was slowing down after a few minutes under full load at 4.5 GHz, it just couldn't manage the heat.
To assess the processor's quality, check its stepping value—C0 is ideal and C3 is the worst. Black edition always uses C0.
That plus the "Silicone lottery" factor shows it's overclocking. I might not win either, since a friend has one running at 5 GHz with water cooling and around 1.505v.
M
MasterDarky
01-30-2017, 07:15 AM #4

It was slowing down after a few minutes under full load at 4.5 GHz, it just couldn't manage the heat.
To assess the processor's quality, check its stepping value—C0 is ideal and C3 is the worst. Black edition always uses C0.
That plus the "Silicone lottery" factor shows it's overclocking. I might not win either, since a friend has one running at 5 GHz with water cooling and around 1.505v.

L
LilGhostz
Junior Member
27
01-30-2017, 08:03 AM
#5
A lot of voltage might be the main reason for your heat problem. You increased the FX 6300's voltage to 1.26, but can you try a lower setting?
L
LilGhostz
01-30-2017, 08:03 AM #5

A lot of voltage might be the main reason for your heat problem. You increased the FX 6300's voltage to 1.26, but can you try a lower setting?

C
CuteFuzzyDalek
Junior Member
43
01-30-2017, 02:04 PM
#6
I don't perform well at lower voltages, the computer won't boot or failed stress tests.
C
CuteFuzzyDalek
01-30-2017, 02:04 PM #6

I don't perform well at lower voltages, the computer won't boot or failed stress tests.

G
GoonerOliver
Member
206
01-30-2017, 06:21 PM
#7
How long have you been using this fx 6300? I'm wondering if you got a processor with some poor thermal management.
G
GoonerOliver
01-30-2017, 06:21 PM #7

How long have you been using this fx 6300? I'm wondering if you got a processor with some poor thermal management.

J
JamboFTW_YT
Member
68
01-30-2017, 09:27 PM
#8
Yeah, I've been doing it for more than two months now and now I'm wishing I had started sooner...
J
JamboFTW_YT
01-30-2017, 09:27 PM #8

Yeah, I've been doing it for more than two months now and now I'm wishing I had started sooner...

P
pixcake
Member
99
01-30-2017, 11:18 PM
#9
Can you explore using another CPU on that motherboard to identify the defective one?
P
pixcake
01-30-2017, 11:18 PM #9

Can you explore using another CPU on that motherboard to identify the defective one?

T
TeenageMinerr
Member
70
01-31-2017, 06:57 AM
#10
This suggestion works well, but my previous chip isn't compatible with this socket, so I'd need to purchase a new one.
T
TeenageMinerr
01-31-2017, 06:57 AM #10

This suggestion works well, but my previous chip isn't compatible with this socket, so I'd need to purchase a new one.