F5F Stay Refreshed Power Users Overclocking i need assistance with overclocking and want to learn more, even though i'm very new to it.

i need assistance with overclocking and want to learn more, even though i'm very new to it.

i need assistance with overclocking and want to learn more, even though i'm very new to it.

Pages (2): Previous 1 2
M
71
03-27-2017, 05:16 AM
#11
5Ghz is quite significant for everyday driving. I wouldn't anticipate it lasting long at such a high operating frequency. I'd suggest lowering it to around 4.8Ghz and adjusting the voltage if feasible. Running Realbench for eight hours is essential to confirm stability, which is crucial if you intend to keep it there. Without this step, instability and micro-errors could cause corruption, especially if the voltage isn't sufficient or the chip can't maintain stability at that speed.
M
miningmario101
03-27-2017, 05:16 AM #11

5Ghz is quite significant for everyday driving. I wouldn't anticipate it lasting long at such a high operating frequency. I'd suggest lowering it to around 4.8Ghz and adjusting the voltage if feasible. Running Realbench for eight hours is essential to confirm stability, which is crucial if you intend to keep it there. Without this step, instability and micro-errors could cause corruption, especially if the voltage isn't sufficient or the chip can't maintain stability at that speed.

L
Leader77
Junior Member
23
03-27-2017, 05:23 AM
#12
Reduced it to 1.328 and I'm still running realbench right now.
L
Leader77
03-27-2017, 05:23 AM #12

Reduced it to 1.328 and I'm still running realbench right now.

L
LuminousPeter
Member
139
03-28-2017, 10:53 PM
#13
What voltage would suit those frequencies, and what should be the ideal one for those? My temperatures are stable, not exceeding 80°C, I'm using an air cooler, and the ambient is cool around 58°F. I wasn't planning to run at 5.0GHz, but thought about 4.8 for summer OC until water cooling arrives. Would it be better to stick with 4.8?
L
LuminousPeter
03-28-2017, 10:53 PM #13

What voltage would suit those frequencies, and what should be the ideal one for those? My temperatures are stable, not exceeding 80°C, I'm using an air cooler, and the ambient is cool around 58°F. I wasn't planning to run at 5.0GHz, but thought about 4.8 for summer OC until water cooling arrives. Would it be better to stick with 4.8?

T
Tiggy123
Member
63
03-29-2017, 06:46 AM
#14
You need to test voltage carefully. Since each chip behaves differently, achieving 5 is possible in less than 1.37, but it's still a lottery. I'm currently running 2600k with 4.6 at 1.360, which won't get me past the 4.6 event close to 1.4, but it doesn't seem worth it for me.
T
Tiggy123
03-29-2017, 06:46 AM #14

You need to test voltage carefully. Since each chip behaves differently, achieving 5 is possible in less than 1.37, but it's still a lottery. I'm currently running 2600k with 4.6 at 1.360, which won't get me past the 4.6 event close to 1.4, but it doesn't seem worth it for me.

F
FlameSquid32
Senior Member
501
03-30-2017, 02:43 PM
#15
You need to test with voltage. Since each chip behaves differently, someone can reach 5 in less than 1.37, but it's still a lottery. I'm running 2600k with 4.6 at 1.360, and I can't get past the 4.6 event close to 1.4, but it doesn't matter to me.
F
FlameSquid32
03-30-2017, 02:43 PM #15

You need to test with voltage. Since each chip behaves differently, someone can reach 5 in less than 1.37, but it's still a lottery. I'm running 2600k with 4.6 at 1.360, and I can't get past the 4.6 event close to 1.4, but it doesn't matter to me.

H
HatterHologram
Junior Member
45
03-30-2017, 03:10 PM
#16
I was at 5 at 1:34 but not on water; the temps rose into the high 80s, so I lowered it to 4:8 at 1:32.
H
HatterHologram
03-30-2017, 03:10 PM #16

I was at 5 at 1:34 but not on water; the temps rose into the high 80s, so I lowered it to 4:8 at 1:32.

L
Leify9
Member
55
04-02-2017, 04:44 PM
#17
He is right. It is different on every configuration. Each specific CPU, plus the number of memory modules AND their configuration, plus what motherboard you have, plus what cooler you have, plus what kind of case cooling arrangement you have, .........................................and so on.
All things have a direct relationship with stability and thermal compliance, so there is no X works with Y voltage. It is "every single system" is different. If we both had EXACTLY identical hardware, it would still potentially be different and you would need to see what works for your system through testing.
If you cannot pass Realbench for 8 hours, and do not cut corners here or you will be risking corruption, then you need to either reduce the frequency or increase the voltage. Also, keep in mind that it's NOT just core voltage. Often you can run at a lower vcore if you increase LLC. BUT, increasing LLC also increases temps. There are checks and balances. It's a multi faceted process, not just a plug in your number here type operation.
Memory voltage affects overall thermals and stability too, which is why your memory should be at the default configuration, NOT at the XMP or custom profile settings, until AFTER the CPU overclock is validated as being stable and thermally compliant.
Thermal compliance is achieved by running Prime95 version 26.6, and ONLY version 26.6 unless you edit the local.txt file on newer versions of Prime to disable the use of AVX instructions, for 15 minutes. Then 8 hours of Realbench to verify stability.
L
Leify9
04-02-2017, 04:44 PM #17

He is right. It is different on every configuration. Each specific CPU, plus the number of memory modules AND their configuration, plus what motherboard you have, plus what cooler you have, plus what kind of case cooling arrangement you have, .........................................and so on.
All things have a direct relationship with stability and thermal compliance, so there is no X works with Y voltage. It is "every single system" is different. If we both had EXACTLY identical hardware, it would still potentially be different and you would need to see what works for your system through testing.
If you cannot pass Realbench for 8 hours, and do not cut corners here or you will be risking corruption, then you need to either reduce the frequency or increase the voltage. Also, keep in mind that it's NOT just core voltage. Often you can run at a lower vcore if you increase LLC. BUT, increasing LLC also increases temps. There are checks and balances. It's a multi faceted process, not just a plug in your number here type operation.
Memory voltage affects overall thermals and stability too, which is why your memory should be at the default configuration, NOT at the XMP or custom profile settings, until AFTER the CPU overclock is validated as being stable and thermally compliant.
Thermal compliance is achieved by running Prime95 version 26.6, and ONLY version 26.6 unless you edit the local.txt file on newer versions of Prime to disable the use of AVX instructions, for 15 minutes. Then 8 hours of Realbench to verify stability.

Pages (2): Previous 1 2