F5F Stay Refreshed Power Users Overclocking Problemi72600K Sandy Bridge [Ho bisogno di consiglio]!

Problemi72600K Sandy Bridge [Ho bisogno di consiglio]!

Problemi72600K Sandy Bridge [Ho bisogno di consiglio]!

F
Fluffycakes123
Senior Member
696
05-22-2021, 06:40 PM
#1
Hello everyone, since I increased my 2600k to 4.6 with offset vcore 0.005+, during the stress test my vcore kept fluctuating between 1.368 and 1.370-80. I was wondering if that voltage level is too high for my clock. P.S. My temperatures reached 80 in Cinebench.
F
Fluffycakes123
05-22-2021, 06:40 PM #1

Hello everyone, since I increased my 2600k to 4.6 with offset vcore 0.005+, during the stress test my vcore kept fluctuating between 1.368 and 1.370-80. I was wondering if that voltage level is too high for my clock. P.S. My temperatures reached 80 in Cinebench.

L
LaBreezy2544
Member
104
05-28-2021, 02:31 PM
#2
Aim to reduce vcore gradually in small steps. After each decrease, use Asus RealBench for 15 minutes. Continue lowering until the program crashes or shows a blue screen. Then increase one voltage notch and re-test. If successful, run a 4-hour loop with RealBench. If it fails again, go back one notch and retest. My Ivy-Bridge reaches 4.6GHz at 1.208v for comparison.
L
LaBreezy2544
05-28-2021, 02:31 PM #2

Aim to reduce vcore gradually in small steps. After each decrease, use Asus RealBench for 15 minutes. Continue lowering until the program crashes or shows a blue screen. Then increase one voltage notch and re-test. If successful, run a 4-hour loop with RealBench. If it fails again, go back one notch and retest. My Ivy-Bridge reaches 4.6GHz at 1.208v for comparison.

L
LoL_e_melhor
Member
96
05-29-2021, 10:40 PM
#3
Offset or Manual Vcore set?
L
LoL_e_melhor
05-29-2021, 10:40 PM #3

Offset or Manual Vcore set?

C
CutiePie45
Junior Member
27
05-30-2021, 03:55 AM
#4
Manual. If you change settings yourself, you should do it manually because auto tends to supply too much power since it doesn’t know where your CPU needs it. Each CPU is unique—some need more voltage than others, and some don’t. Auto won’t recognize the limits, so it’s up to you.
C
CutiePie45
05-30-2021, 03:55 AM #4

Manual. If you change settings yourself, you should do it manually because auto tends to supply too much power since it doesn’t know where your CPU needs it. Each CPU is unique—some need more voltage than others, and some don’t. Auto won’t recognize the limits, so it’s up to you.

C
CuzImJuli
Member
204
05-30-2021, 04:11 AM
#5
That's a fairly high vcore for this OC. I'd suggest applying a negative offset to keep the OC stable at a lower vcore. Under 1.3v usually works unless your CPU is a tough overclocker.
C
CuzImJuli
05-30-2021, 04:11 AM #5

That's a fairly high vcore for this OC. I'd suggest applying a negative offset to keep the OC stable at a lower vcore. Under 1.3v usually works unless your CPU is a tough overclocker.

C
Crystal_Topaz
Member
65
05-30-2021, 09:56 AM
#6
Sandy's often need a bit more vcore than later generation CPUs, so 1.3v at 4.6GHz isn't too bad.
When comparing VID to vcore, the gap will determine the needed offset.
For instance, if your VID is 1.3875 under load and your CPU-Z vcore under load is 1.4275, the offset will be a positive number added to the VID, resulting in +0.040 (1.3875 + 0.040 = 1.4275).
If the VID stays the same at 1.3875 but your CPU-Z vcore drops to 1.3675, the offset becomes negative: -0.020 (1.3875 - 0.020 = 1.3675).
C
Crystal_Topaz
05-30-2021, 09:56 AM #6

Sandy's often need a bit more vcore than later generation CPUs, so 1.3v at 4.6GHz isn't too bad.
When comparing VID to vcore, the gap will determine the needed offset.
For instance, if your VID is 1.3875 under load and your CPU-Z vcore under load is 1.4275, the offset will be a positive number added to the VID, resulting in +0.040 (1.3875 + 0.040 = 1.4275).
If the VID stays the same at 1.3875 but your CPU-Z vcore drops to 1.3675, the offset becomes negative: -0.020 (1.3875 - 0.020 = 1.3675).

W
Wastingman
Member
164
06-03-2021, 05:18 AM
#7
Puza Gaming, here are the suggested maximum Core voltages for various microarchitectures from 14 to 65 nanometers since 2006. For your i7-2600K, which is a second-generation 32nm chip, stay around ~1.375 Vcore. When adjusting near its top overclock setting, note that each 100 MHz boost requires roughly a 50 millivolt rise (0.050) to keep things stable. If you need 70 millivolts or more for the next stable 100 MHz jump, it indicates your processor is being overclocked beyond its safe range.

The typical core temperature limits are:
- Temperatures above 85°C should be avoided
- Temperatures below 80°C are preferred
- Core temps change with ambient conditions

• What’s your room temperature?
With strong cooling, you might hit the Vcore cap before reaching 85°C. With weaker cooling, it will hit 85°C sooner. In either case, stop once you reach that threshold.

Keep in mind the overclocking limits: a small jump like from 4.5 GHz to 4.6 GHz is only about 2.3%, which doesn’t meaningfully affect overall performance. Pushing beyond these values isn’t justified just for a marginal speed gain.
W
Wastingman
06-03-2021, 05:18 AM #7

Puza Gaming, here are the suggested maximum Core voltages for various microarchitectures from 14 to 65 nanometers since 2006. For your i7-2600K, which is a second-generation 32nm chip, stay around ~1.375 Vcore. When adjusting near its top overclock setting, note that each 100 MHz boost requires roughly a 50 millivolt rise (0.050) to keep things stable. If you need 70 millivolts or more for the next stable 100 MHz jump, it indicates your processor is being overclocked beyond its safe range.

The typical core temperature limits are:
- Temperatures above 85°C should be avoided
- Temperatures below 80°C are preferred
- Core temps change with ambient conditions

• What’s your room temperature?
With strong cooling, you might hit the Vcore cap before reaching 85°C. With weaker cooling, it will hit 85°C sooner. In either case, stop once you reach that threshold.

Keep in mind the overclocking limits: a small jump like from 4.5 GHz to 4.6 GHz is only about 2.3%, which doesn’t meaningfully affect overall performance. Pushing beyond these values isn’t justified just for a marginal speed gain.