F5F Stay Refreshed Power Users Overclocking Some OCing questions (i7 4790k)

Some OCing questions (i7 4790k)

Some OCing questions (i7 4790k)

G
GamenMetLeviNL
Senior Member
638
02-05-2016, 08:01 PM
#1
When I first installed this PC, I did the overclocking in a suitable position. Now I want to revisit everything and fine-tune temperatures, voltages, and overclocks. I’m not sure how to do it all properly, so I have some questions. I’m about to experiment with all settings, so I might ask a few more after this post.
G
GamenMetLeviNL
02-05-2016, 08:01 PM #1

When I first installed this PC, I did the overclocking in a suitable position. Now I want to revisit everything and fine-tune temperatures, voltages, and overclocks. I’m not sure how to do it all properly, so I have some questions. I’m about to experiment with all settings, so I might ask a few more after this post.

T
T___________T
Member
224
02-21-2016, 10:16 AM
#2
Additionally, it's important to note that excessively high CPU core temperatures can lead to accelerated wear on the chip and even overclock instability due to heat. Recently, I discovered my CPU cache overclock was operating at a higher voltage than intended, resulting in 1.38 to 1.39V for the cache. With the cooling system in place, it seems the memory temperatures were quite high. Stability remained an issue, but adjusting the voltage to 1.28V brought long-term consistency.
T
T___________T
02-21-2016, 10:16 AM #2

Additionally, it's important to note that excessively high CPU core temperatures can lead to accelerated wear on the chip and even overclock instability due to heat. Recently, I discovered my CPU cache overclock was operating at a higher voltage than intended, resulting in 1.38 to 1.39V for the cache. With the cooling system in place, it seems the memory temperatures were quite high. Stability remained an issue, but adjusting the voltage to 1.28V brought long-term consistency.

M
muffles45
Member
189
02-21-2016, 10:23 AM
#3
You're operating at around 75-80°C during gaming, which aligns with the target you should be aiming for. Exceeding 85°C will cause more wear on the processor's reliability over time. Since you're using air cooling, consider adding more fans if you consistently hit temperatures above 85°C (or switch to a liquid cooler if you prefer). These CPUs are among the hottest Intel models, so effective cooling is essential. It might be wise to run stress tests or games after overclocking and keep an eye on the temperatures, assuming your stability holds.

Be sure to maintain voltages below 1.3V when adjusting settings. I have a unit that runs at 1.26V at 4.7GHz, but results may vary. Going above 1.3V can risk damaging or destroying the CPU.

With the Maximus VII and PWM fans, I use AI Suite to control fan speeds. Corsair Link works for me only because I own an H80i and can manage all fans through it. There could be other options, but those two are my go-to since they offer proprietary features.

These tips come from my experience with this rig (full specs in my profile).
M
muffles45
02-21-2016, 10:23 AM #3

You're operating at around 75-80°C during gaming, which aligns with the target you should be aiming for. Exceeding 85°C will cause more wear on the processor's reliability over time. Since you're using air cooling, consider adding more fans if you consistently hit temperatures above 85°C (or switch to a liquid cooler if you prefer). These CPUs are among the hottest Intel models, so effective cooling is essential. It might be wise to run stress tests or games after overclocking and keep an eye on the temperatures, assuming your stability holds.

Be sure to maintain voltages below 1.3V when adjusting settings. I have a unit that runs at 1.26V at 4.7GHz, but results may vary. Going above 1.3V can risk damaging or destroying the CPU.

With the Maximus VII and PWM fans, I use AI Suite to control fan speeds. Corsair Link works for me only because I own an H80i and can manage all fans through it. There could be other options, but those two are my go-to since they offer proprietary features.

These tips come from my experience with this rig (full specs in my profile).

C
ChayChay12345
Member
54
02-24-2016, 10:47 AM
#4
When operating the system around the clock at full capacity, maintain temperatures below 75°C. Operating outside this range might lead to reaching 80°C during stress tests, which you won’t encounter in regular use.

I’m using an i7 4790K at 4.6GHz with a voltage of 1.24V and my maximum temperature is 58°C (delidded CPU with liquid metal thermal interface on both sides). I observe around 78°C when running Prime95 28.10 with AVX extensions. That’s why I prefer not to exceed these limits, ensuring the system remains stable under continuous 24/7 heavy load.

I recommend making all BIOS adjustments permanent and hardware-based, rather than software-only changes. While software testing is possible, it doesn’t guarantee full reliability. Your BIOS should be capable of fine-tuning fan speeds, though some components may require PWM fans.
C
ChayChay12345
02-24-2016, 10:47 AM #4

When operating the system around the clock at full capacity, maintain temperatures below 75°C. Operating outside this range might lead to reaching 80°C during stress tests, which you won’t encounter in regular use.

I’m using an i7 4790K at 4.6GHz with a voltage of 1.24V and my maximum temperature is 58°C (delidded CPU with liquid metal thermal interface on both sides). I observe around 78°C when running Prime95 28.10 with AVX extensions. That’s why I prefer not to exceed these limits, ensuring the system remains stable under continuous 24/7 heavy load.

I recommend making all BIOS adjustments permanent and hardware-based, rather than software-only changes. While software testing is possible, it doesn’t guarantee full reliability. Your BIOS should be capable of fine-tuning fan speeds, though some components may require PWM fans.

C
cowcow4321
Senior Member
623
02-29-2016, 12:25 PM
#5
Do you know why the reference clock on HWMonitor or Intel Tuning Utility displays "100.174" instead of just 100?
C
cowcow4321
02-29-2016, 12:25 PM #5

Do you know why the reference clock on HWMonitor or Intel Tuning Utility displays "100.174" instead of just 100?

T
TheZenKaiZ_Yt
Member
151
02-29-2016, 07:35 PM
#6
Additionally, it's important to note that excessively high CPU core temperatures can accelerate wear on the chip and lead to overclock instability simply due to heat. Recently, I noticed my CPU cache overclock was operating at a voltage of 1.38-1.39V, which is quite high. With the cooling system in place, it seems the memory temperatures were quite elevated. Stability remained elusive, but when I manually lowered the voltage to 1.28V, things stabilized effectively.

A common belief is that heat is the primary cause of CPU damage, not voltage. While voltage can contribute to material stress and potentially accelerate degradation, the danger threshold typically exceeds 1.3V. Going beyond that increases risk significantly. However, modern processors are built to handle higher voltages safely, especially with advanced cooling solutions. My friend successfully runs a 4790k at 4.9 GHz using 1.38V with air cooling—performing well. I wouldn’t rush this setup, but it’s reliable.

It’s worth remembering that the VRM on the motherboard also plays a role in overclock stability by dividing the power supply into phases for the CPU. Increasing the number of power phases can help manage heat better, potentially extending the lifespan of both the VRM and the CPU. If you damage the VRM, it can lead to failure of the entire board, including the CPU. Although VRMs are designed to handle heat, using a board with fewer phases and pushing overclocking limits may require replacing it later.

You can fine-tune fan curves on your CPU and case fans via BIOS or Windows software, with AI Suite 3 being my preferred method.

Extra tips:
- If you're concerned about overvolting, consider purchasing an Intel Tuning Plan for your CPU. It offers a replacement warranty and costs around $20.
T
TheZenKaiZ_Yt
02-29-2016, 07:35 PM #6

Additionally, it's important to note that excessively high CPU core temperatures can accelerate wear on the chip and lead to overclock instability simply due to heat. Recently, I noticed my CPU cache overclock was operating at a voltage of 1.38-1.39V, which is quite high. With the cooling system in place, it seems the memory temperatures were quite elevated. Stability remained elusive, but when I manually lowered the voltage to 1.28V, things stabilized effectively.

A common belief is that heat is the primary cause of CPU damage, not voltage. While voltage can contribute to material stress and potentially accelerate degradation, the danger threshold typically exceeds 1.3V. Going beyond that increases risk significantly. However, modern processors are built to handle higher voltages safely, especially with advanced cooling solutions. My friend successfully runs a 4790k at 4.9 GHz using 1.38V with air cooling—performing well. I wouldn’t rush this setup, but it’s reliable.

It’s worth remembering that the VRM on the motherboard also plays a role in overclock stability by dividing the power supply into phases for the CPU. Increasing the number of power phases can help manage heat better, potentially extending the lifespan of both the VRM and the CPU. If you damage the VRM, it can lead to failure of the entire board, including the CPU. Although VRMs are designed to handle heat, using a board with fewer phases and pushing overclocking limits may require replacing it later.

You can fine-tune fan curves on your CPU and case fans via BIOS or Windows software, with AI Suite 3 being my preferred method.

Extra tips:
- If you're concerned about overvolting, consider purchasing an Intel Tuning Plan for your CPU. It offers a replacement warranty and costs around $20.