F5F Stay Refreshed Hardware Desktop Engaging in dangerous actions while using IGPG

Engaging in dangerous actions while using IGPG

Engaging in dangerous actions while using IGPG

C
CombatBorn
Junior Member
13
11-26-2016, 06:33 AM
#1
specs : b40m s2h, 3200g, 16gb 2x8gb DDR4 if bios stock settings. Cinebench R23 running single core benchmark is showing Ryzen Master reports 1.4v peak voltage. Now attempting a slight overclock of the iGPU with gfx voltage at 1.16 and 1340mhz iGPU clock, the measured voltage during C23 single CPU benchmark is 1.55v! Ryzen Master confirms this figure, as does hwinfo64 (only gfx voltage and iGPU clock changed, rest remain stock), indicating a rise of about 0.15v. I’m certain the GPU overclocking is responsible since I toggled settings multiple times to verify. If I set manual Vcore to 1.2v with 3.8ghz boost, all cores would stay at max 1.2v, suggesting gfx voltage isn’t increasing Vcore either. During a 7-minute Cinebench run at 1.55v, I assumed it was automatic and it worked fine. Hope I didn’t accidentally modify the CPU. Keep in mind this wasn’t just a quick check—this was ongoing for 5-7 minutes with auto CPU frequency and boost enabled to test if 4.0ghz single core is achievable. Currently, 'Core performance boost' appears disabled when it should be, possibly due to the automatic setting.
C
CombatBorn
11-26-2016, 06:33 AM #1

specs : b40m s2h, 3200g, 16gb 2x8gb DDR4 if bios stock settings. Cinebench R23 running single core benchmark is showing Ryzen Master reports 1.4v peak voltage. Now attempting a slight overclock of the iGPU with gfx voltage at 1.16 and 1340mhz iGPU clock, the measured voltage during C23 single CPU benchmark is 1.55v! Ryzen Master confirms this figure, as does hwinfo64 (only gfx voltage and iGPU clock changed, rest remain stock), indicating a rise of about 0.15v. I’m certain the GPU overclocking is responsible since I toggled settings multiple times to verify. If I set manual Vcore to 1.2v with 3.8ghz boost, all cores would stay at max 1.2v, suggesting gfx voltage isn’t increasing Vcore either. During a 7-minute Cinebench run at 1.55v, I assumed it was automatic and it worked fine. Hope I didn’t accidentally modify the CPU. Keep in mind this wasn’t just a quick check—this was ongoing for 5-7 minutes with auto CPU frequency and boost enabled to test if 4.0ghz single core is achievable. Currently, 'Core performance boost' appears disabled when it should be, possibly due to the automatic setting.

E
EzechielDLC
Member
67
11-26-2016, 08:19 AM
#2
No need for overclocking, everything runs smoothly. I lowered the voltage on my 11-year-old CPU, and my GPU performs much better than anything on a Raspberry Pi.
E
EzechielDLC
11-26-2016, 08:19 AM #2

No need for overclocking, everything runs smoothly. I lowered the voltage on my 11-year-old CPU, and my GPU performs much better than anything on a Raspberry Pi.

G
Gotten75
Member
129
11-26-2016, 10:28 AM
#3
Undervolting is a method of overclocking that focuses on efficiency. It doesn’t involve hardware wear, so you won’t face degradation problems. Still, you can enjoy precise adjustments to improve stability.
G
Gotten75
11-26-2016, 10:28 AM #3

Undervolting is a method of overclocking that focuses on efficiency. It doesn’t involve hardware wear, so you won’t face degradation problems. Still, you can enjoy precise adjustments to improve stability.

P
Pawtex
Member
114
11-26-2016, 02:08 PM
#4
I use a dual-core processor running at 1596 MHz, not 2 GHz. This discussion isn't about me, but many others are facing issues from overclocking today.
P
Pawtex
11-26-2016, 02:08 PM #4

I use a dual-core processor running at 1596 MHz, not 2 GHz. This discussion isn't about me, but many others are facing issues from overclocking today.