F5F Stay Refreshed Hardware Desktop Bad luck 9800X3D overheating issue with 420mm AIO – reaches 90°C+

Bad luck 9800X3D overheating issue with 420mm AIO – reaches 90°C+

Bad luck 9800X3D overheating issue with 420mm AIO – reaches 90°C+

Pages (3): Previous 1 2 3 Next
K
Kacper_Bored
Senior Member
389
04-13-2016, 07:20 AM
#11
The issue appears to be inconsistent performance across different models. Many 9800X3D units reach around 5.4Ghz, but some exceptions exist. The reported 170+ watts at 1.25V seem significantly higher—about 15W more than other 9800X3D models at the same voltage.
K
Kacper_Bored
04-13-2016, 07:20 AM #11

The issue appears to be inconsistent performance across different models. Many 9800X3D units reach around 5.4Ghz, but some exceptions exist. The reported 170+ watts at 1.25V seem significantly higher—about 15W more than other 9800X3D models at the same voltage.

H
hope0007
Junior Member
4
04-14-2016, 07:02 AM
#12
It seems this process requires significantly higher voltage and power, triggering excessive heat. Not all chips are suitable for overclocking. (I think it's unreasonable, I prefer stable temperatures, minimal power use, and solid performance. That means I only undervolt, not the Asus laptop. I need to verify this further—I feel it would work better with less heat and possibly a PBO undervolt.) Also, keeping temperatures around 92°C might be better than aiming for 5.400; sustained high temps usually cause issues sooner.
H
hope0007
04-14-2016, 07:02 AM #12

It seems this process requires significantly higher voltage and power, triggering excessive heat. Not all chips are suitable for overclocking. (I think it's unreasonable, I prefer stable temperatures, minimal power use, and solid performance. That means I only undervolt, not the Asus laptop. I need to verify this further—I feel it would work better with less heat and possibly a PBO undervolt.) Also, keeping temperatures around 92°C might be better than aiming for 5.400; sustained high temps usually cause issues sooner.

X
Xx_PoloMAN_xX
Junior Member
44
04-15-2016, 03:40 PM
#13
OCCT isn't detecting RAM low bin issues, maybe it's a coincidence. The 9800x3D model is rated for 5.2Ghz boost, so you're likely surpassing that. Compared to other 9800x3D chips, your CPU consumes more power at the same voltage, suggesting a less powerful die that was boosted more. Whether it has 0.1Ghz or full 1Ghz overclock, it still meets the advertised speeds. It seems 9800x3D might be using lower-quality chips than 9700x, possibly due to the need for a higher boost clock in the latter. Interestingly, 9700x requires a 5.5Ghz boost at launch: https://www.techpowerup.com/cpu-specs/ry...700x.c3651. Also, its power consumption appears lower, which could relate to the large L3 cache affecting performance. TL;DR: it's probably just bad luck. In 2025, overclocking feels like a tough challenge—modern PCs are nearly unmodifiable, except for minor tweaks or premium parts like AM5 CPUs that run hot unless you have a big cooling solution.
X
Xx_PoloMAN_xX
04-15-2016, 03:40 PM #13

OCCT isn't detecting RAM low bin issues, maybe it's a coincidence. The 9800x3D model is rated for 5.2Ghz boost, so you're likely surpassing that. Compared to other 9800x3D chips, your CPU consumes more power at the same voltage, suggesting a less powerful die that was boosted more. Whether it has 0.1Ghz or full 1Ghz overclock, it still meets the advertised speeds. It seems 9800x3D might be using lower-quality chips than 9700x, possibly due to the need for a higher boost clock in the latter. Interestingly, 9700x requires a 5.5Ghz boost at launch: https://www.techpowerup.com/cpu-specs/ry...700x.c3651. Also, its power consumption appears lower, which could relate to the large L3 cache affecting performance. TL;DR: it's probably just bad luck. In 2025, overclocking feels like a tough challenge—modern PCs are nearly unmodifiable, except for minor tweaks or premium parts like AM5 CPUs that run hot unless you have a big cooling solution.

D
DRAGON91160
Member
108
04-15-2016, 04:57 PM
#14
No, I am not setting static clocks and voltage.
D
DRAGON91160
04-15-2016, 04:57 PM #14

No, I am not setting static clocks and voltage.

U
UniPopCorn
Member
210
04-24-2016, 03:56 AM
#15
Yes
U
UniPopCorn
04-24-2016, 03:56 AM #15

Yes

I
iScoozyxx
Member
63
05-07-2016, 02:27 PM
#16
If you're unsure about performance at 5400MHz, increase the voltage. If overheating occurs, it indicates a problem needs fixing.
I
iScoozyxx
05-07-2016, 02:27 PM #16

If you're unsure about performance at 5400MHz, increase the voltage. If overheating occurs, it indicates a problem needs fixing.

E
EchoPenguin
Junior Member
13
05-09-2016, 12:33 AM
#17
The power output has dropped to around 150W under full load. The voltage remains sufficient for stability at 1.26V in BIOS. The exact BIOS adjustment is unclear, but it accounts for nearly a 30W variation.
E
EchoPenguin
05-09-2016, 12:33 AM #17

The power output has dropped to around 150W under full load. The voltage remains sufficient for stability at 1.26V in BIOS. The exact BIOS adjustment is unclear, but it accounts for nearly a 30W variation.

B
Barty__
Junior Member
3
05-22-2016, 11:18 AM
#18
For checking all temperature ranges I prefer Linpack Extreme. I test the 10GB workload. It’s the only app I’m aware of that lets my CPU reach 250W with air or 260W with an AIO. Note: With PBO, not static clocks.
B
Barty__
05-22-2016, 11:18 AM #18

For checking all temperature ranges I prefer Linpack Extreme. I test the 10GB workload. It’s the only app I’m aware of that lets my CPU reach 250W with air or 260W with an AIO. Note: With PBO, not static clocks.

_
_RGamer_
Member
53
06-07-2016, 01:45 PM
#19
Linpack Extreme evaluates CPU stability rapidly. Despite attempts, I haven't achieved consistent 5.2GHz performance, so I settled for default settings, which worked temporarily. Speed dropped slightly, but it functioned.
_
_RGamer_
06-07-2016, 01:45 PM #19

Linpack Extreme evaluates CPU stability rapidly. Despite attempts, I haven't achieved consistent 5.2GHz performance, so I settled for default settings, which worked temporarily. Speed dropped slightly, but it functioned.

J
Joaozinh
Member
124
06-16-2016, 04:15 AM
#20
These CPUs aren't designed for fixed clocks and voltages. They can manage a clock and voltage, but slight variations will always occur. You might be able to lower all core speeds to a consistent pace, but AMD's core speed ratings will complicate the process unless you have an unusual solution. To clarify, the threads on CCD1 will reach 58xx speeds, while CCD2 lags behind at 200MHz or less. Running all cores at 5800 would require significant voltage adjustments, possibly limiting performance to medium tasks like Cinebench. As observed with Linpack, the actual running speed is around 5300MHz.
J
Joaozinh
06-16-2016, 04:15 AM #20

These CPUs aren't designed for fixed clocks and voltages. They can manage a clock and voltage, but slight variations will always occur. You might be able to lower all core speeds to a consistent pace, but AMD's core speed ratings will complicate the process unless you have an unusual solution. To clarify, the threads on CCD1 will reach 58xx speeds, while CCD2 lags behind at 200MHz or less. Running all cores at 5800 would require significant voltage adjustments, possibly limiting performance to medium tasks like Cinebench. As observed with Linpack, the actual running speed is around 5300MHz.

Pages (3): Previous 1 2 3 Next