F5F Stay Refreshed Power Users Overclocking Question frequency increases with my OC 5950X - from 5100 to 3600.

Question frequency increases with my OC 5950X - from 5100 to 3600.

Question frequency increases with my OC 5950X - from 5100 to 3600.

Pages (2): 1 2 Next
D
DarkBoy__YT
Posting Freak
898
05-29-2021, 12:36 PM
#1
Thread moved to the Overclocking Forum.
Frequency fluctuations observed with my OC 5950X, rising from 5100 to 3600.
Currently testing on Windows, but intend to revert to Linux, so I’d like to clarify everything beforehand.
Recently acquired a new PC and have limited overclocking experience, yet still made some adjustments. Don’t worry if I’m a bit new to this.
Spec list:
- 5950X
- X570 TUF Gaming
- Dark Rock Pro 4
- CL14 Trident Neo 3600 MHz
- seasonic gx-850 gold
- gigabyte 6800xt 16gb OC.
-samsing 9100 pro 2 tb heatsink. only windows 11
-another m2 wd sn570
-3 old hdds
-8 artic p12 max fans
I activated maximum performance mode in Windows by editing the powercfg.cpl file:
powercfg -duplicatescheme SCHEME_MIN
powercfg -duplicatescheme e9a42b02-d5df-448d-aa00-03f14749eb61
These are my BIOS settings:
- **CPPC** → Enabled
- **CPPC Preferred Cores** → Disabled (similar issue with enabled setting)
pbo fmax enchancer = disabled
vddcr soc power phase control = extreme
vddcr cpu power phase control = extreme
D
DarkBoy__YT
05-29-2021, 12:36 PM #1

Thread moved to the Overclocking Forum.
Frequency fluctuations observed with my OC 5950X, rising from 5100 to 3600.
Currently testing on Windows, but intend to revert to Linux, so I’d like to clarify everything beforehand.
Recently acquired a new PC and have limited overclocking experience, yet still made some adjustments. Don’t worry if I’m a bit new to this.
Spec list:
- 5950X
- X570 TUF Gaming
- Dark Rock Pro 4
- CL14 Trident Neo 3600 MHz
- seasonic gx-850 gold
- gigabyte 6800xt 16gb OC.
-samsing 9100 pro 2 tb heatsink. only windows 11
-another m2 wd sn570
-3 old hdds
-8 artic p12 max fans
I activated maximum performance mode in Windows by editing the powercfg.cpl file:
powercfg -duplicatescheme SCHEME_MIN
powercfg -duplicatescheme e9a42b02-d5df-448d-aa00-03f14749eb61
These are my BIOS settings:
- **CPPC** → Enabled
- **CPPC Preferred Cores** → Disabled (similar issue with enabled setting)
pbo fmax enchancer = disabled
vddcr soc power phase control = extreme
vddcr cpu power phase control = extreme

A
AndyTEAM
Member
143
05-29-2021, 09:32 PM
#2
Frequency variations observed with my OC 5950X, shifting from 5100 to 3600 MHz. Currently, I'm working on Windows but intend to revert to my Linux environment, so I want to resolve this beforehand.

Recently, I purchased a new PC and have limited overclocking experience, yet I still made some adjustments. Don't worry if I'm a bit new to this.

Spec list:
- 5950X
- X570 TUF Gaming
- Dark Rock Pro 4
- CL14 Trident Neo 3600 MHz

I activated maximum performance mode in Windows by using:
(ofc set it in the powercfg.cpl)
powercfg -duplicatescheme SCHEME_MIN
powercfg -duplicatescheme e9a42b02-d5df-448d-aa00-03f14749eb61

And these are my BIOS settings:
- **CPPC** → Enabled
- **CPPC Preferred Cores** → Disabled (similar issue with enabled)
pbo fmax enchancer = disabled
vddcr soc power phase control = extreme
vddcr cpu power phase control = extreme
A
AndyTEAM
05-29-2021, 09:32 PM #2

Frequency variations observed with my OC 5950X, shifting from 5100 to 3600 MHz. Currently, I'm working on Windows but intend to revert to my Linux environment, so I want to resolve this beforehand.

Recently, I purchased a new PC and have limited overclocking experience, yet I still made some adjustments. Don't worry if I'm a bit new to this.

Spec list:
- 5950X
- X570 TUF Gaming
- Dark Rock Pro 4
- CL14 Trident Neo 3600 MHz

I activated maximum performance mode in Windows by using:
(ofc set it in the powercfg.cpl)
powercfg -duplicatescheme SCHEME_MIN
powercfg -duplicatescheme e9a42b02-d5df-448d-aa00-03f14749eb61

And these are my BIOS settings:
- **CPPC** → Enabled
- **CPPC Preferred Cores** → Disabled (similar issue with enabled)
pbo fmax enchancer = disabled
vddcr soc power phase control = extreme
vddcr cpu power phase control = extreme

C
chamaballz
Member
127
05-29-2021, 11:02 PM
#3
your voltage readings are quite high, be cautious since these components are already performing well under standard conditions.
my performance metrics varied significantly because the system was consistently operating at 1.5v or higher, leading to overheating or instability.
C
chamaballz
05-29-2021, 11:02 PM #3

your voltage readings are quite high, be cautious since these components are already performing well under standard conditions.
my performance metrics varied significantly because the system was consistently operating at 1.5v or higher, leading to overheating or instability.

E
EisTeeKlaus
Senior Member
490
05-31-2021, 09:59 PM
#4
Thanks for the reply!
Approx 1.475V MAX / 1.3 avg (cinebench)
I haven't adjusted voltage and completed 20-30 cinebench tests yet
😳
Believed 1.475 for short bursts is typical with the negative curve already optimized
Temperatures are fine, which is surprising so I'm proceeding
E
EisTeeKlaus
05-31-2021, 09:59 PM #4

Thanks for the reply!
Approx 1.475V MAX / 1.3 avg (cinebench)
I haven't adjusted voltage and completed 20-30 cinebench tests yet
😳
Believed 1.475 for short bursts is typical with the negative curve already optimized
Temperatures are fine, which is surprising so I'm proceeding

V
ViiRaL_Hyper
Member
167
06-11-2021, 09:52 PM
#5
negative curve reduces voltage which is why it doesn’t rise above 1.475. that’s beneficial since exceeding 1.5v can harm the CPU. for better longevity, 1.35v is preferable for extended use. 1.4v is pushing performance and anything beyond to 1.4 to 1.5v is only suitable for short periods, not long-term. temperatures at 1.3v are acceptable, but sustained highs in 1.4v or higher can stress air cooling. I’ve had this experience with the 5900x.
V
ViiRaL_Hyper
06-11-2021, 09:52 PM #5

negative curve reduces voltage which is why it doesn’t rise above 1.475. that’s beneficial since exceeding 1.5v can harm the CPU. for better longevity, 1.35v is preferable for extended use. 1.4v is pushing performance and anything beyond to 1.4 to 1.5v is only suitable for short periods, not long-term. temperatures at 1.3v are acceptable, but sustained highs in 1.4v or higher can stress air cooling. I’ve had this experience with the 5900x.

X
xCHADWICK
Junior Member
8
06-19-2021, 01:57 PM
#6
Currently, my top cores are:
- Core 02 → –8
- Core 06 → –10
- Core 09 → –8
- Core 11 → –10
🔥 the hottest options are:
- Core 01 → –3
- Core 05 → –3
Other:
core 15,14,13,12,10 = -15
for the rest -12
should I opt for -10 instead of -15 in the "other" group to lower the voltage? Could be related to the clock spikes too?
And if not, perhaps I should raise the boost override to at least +50 and set ppt to 200?
My temperatures are quite low – around 89 hotspot during heavy benchmarks. Sometimes it reaches 92 because of cooler cooling delays.
X
xCHADWICK
06-19-2021, 01:57 PM #6

Currently, my top cores are:
- Core 02 → –8
- Core 06 → –10
- Core 09 → –8
- Core 11 → –10
🔥 the hottest options are:
- Core 01 → –3
- Core 05 → –3
Other:
core 15,14,13,12,10 = -15
for the rest -12
should I opt for -10 instead of -15 in the "other" group to lower the voltage? Could be related to the clock spikes too?
And if not, perhaps I should raise the boost override to at least +50 and set ppt to 200?
My temperatures are quite low – around 89 hotspot during heavy benchmarks. Sometimes it reaches 92 because of cooler cooling delays.

P
paulkadots
Member
210
06-19-2021, 09:32 PM
#7
what tools are you employing to check temperatures? it shows both ccds on the hw monitor. the cores are divided between two ccds—one might read 89 while the other is higher. 89 is probably anything above 92 or 93, which could trigger the chip to shut down. negative 10 would increase the temperature; you’d need to go further into the negative range to get a lower voltage.
P
paulkadots
06-19-2021, 09:32 PM #7

what tools are you employing to check temperatures? it shows both ccds on the hw monitor. the cores are divided between two ccds—one might read 89 while the other is higher. 89 is probably anything above 92 or 93, which could trigger the chip to shut down. negative 10 would increase the temperature; you’d need to go further into the negative range to get a lower voltage.

G
Gagsu
Member
195
06-20-2021, 06:28 AM
#8
Yeah, I'm checking hwinfo. 92 was the highest temperature for 1-2 tests on CB. On average, I'm below 90 on CCD. The crystal stays under 85. I've got really noisy P12 max coolers and set a very quiet curve for them.

Edit:
Thanks a lot for the voltage tips!
Adjusted voltage under load: max 1.475, avg 1.25.
Temps stay under 88 hotspot, but idle is usually 1.4-1.5 which seems normal for RYZEN?
By tweaking:
200 PPT override boost to +125
TDC: 200
EDC: 150
Curve Optimizer:
Top 4 cores -14, Hot cores -3, rest -12. Spikes still an issue (especially with HWINFO, riva tuner, and afterburner). Performance improves 5-6% at lower temps or with multicore.
-5% singlecore.
G
Gagsu
06-20-2021, 06:28 AM #8

Yeah, I'm checking hwinfo. 92 was the highest temperature for 1-2 tests on CB. On average, I'm below 90 on CCD. The crystal stays under 85. I've got really noisy P12 max coolers and set a very quiet curve for them.

Edit:
Thanks a lot for the voltage tips!
Adjusted voltage under load: max 1.475, avg 1.25.
Temps stay under 88 hotspot, but idle is usually 1.4-1.5 which seems normal for RYZEN?
By tweaking:
200 PPT override boost to +125
TDC: 200
EDC: 150
Curve Optimizer:
Top 4 cores -14, Hot cores -3, rest -12. Spikes still an issue (especially with HWINFO, riva tuner, and afterburner). Performance improves 5-6% at lower temps or with multicore.
-5% singlecore.

H
hayhaytaylor
Member
192
06-22-2021, 03:00 AM
#9
Update your post with the required system details and specifications. Add PSU information, disk drive specs, and a list of peripherals. Clarify whether the network is wired or wireless. Also, specify the quantity and type of overclocking and expected outcomes.
H
hayhaytaylor
06-22-2021, 03:00 AM #9

Update your post with the required system details and specifications. Add PSU information, disk drive specs, and a list of peripherals. Clarify whether the network is wired or wireless. Also, specify the quantity and type of overclocking and expected outcomes.

B
bobbuilder2222
Junior Member
23
06-22-2021, 03:07 AM
#10
Hey! Thanks for the update.
seasonic gx-850 gold. Used in previous builds - psi.
gigabyte 6800xt 16gb oc.
ryzen 9 5950x.
asus TUF gaming x570 plus wifi.
32 gb cl14 b-die trident z neo f4-3600c14d-32gtzn
samsing 9100 pro 2 tb heatsink. only windows 11
another m2 wd sn570
3 old hdds
8 artic p12 max fans
Just wanted to try some OC tweaks for a bit.
And since I often work with virtualization, I thought it might help boost performance.
(because my system is usually stable)
I just wanted to check if my voltage is safe (around 1.47v) and how to reduce those annoying frequency spikes in games.
B
bobbuilder2222
06-22-2021, 03:07 AM #10

Hey! Thanks for the update.
seasonic gx-850 gold. Used in previous builds - psi.
gigabyte 6800xt 16gb oc.
ryzen 9 5950x.
asus TUF gaming x570 plus wifi.
32 gb cl14 b-die trident z neo f4-3600c14d-32gtzn
samsing 9100 pro 2 tb heatsink. only windows 11
another m2 wd sn570
3 old hdds
8 artic p12 max fans
Just wanted to try some OC tweaks for a bit.
And since I often work with virtualization, I thought it might help boost performance.
(because my system is usually stable)
I just wanted to check if my voltage is safe (around 1.47v) and how to reduce those annoying frequency spikes in games.

Pages (2): 1 2 Next