The CPU experiences instability when overclocked using XMP settings.
The CPU experiences instability when overclocked using XMP settings.
CPU: i7-7700k @ 4.20 GHz (Turbo boost: 4.5GHz)
[delidded]
Mobo: Asus Maximus IX Extreme
GPU: Gigabyte - GTX 1080 Ti Waterforce WB Xtreme Edition 11G
RAM: 32GB - G.Skill - Trident Z RGB - DDR4 3600Mhz - 32GB (8GBx4) - 16-16-16-36
PSU: Seasonic - PRIME 850 W Titanium
Custom cooling setup:
Radiator: two units (520x30x60mm)
Radiator fans: eight (120x120x15mm)
Pump: EK-XRES 140 Revo D5
I upgraded my i7-7700k to 5GHz using this setup:
CPU ratio: 50%
Cache ratio: 42%
AVX offset: 0
Vcore mode: Manual (override)
LLC: 5 (on a scale of 1 to 8)
VCCIO: Auto
VCCSA: Auto
iGPU: Disabled
I adjusted the Vcore in steps of 0.005V and verified stability with Prime95 (XMP disabled):
Vcore XMP
Stress tests:
1.405v Disabled
Prime95 failed after 54 minutes (Blend mode)
1.410v Disabled
Prime95 failed after 11 minutes (Blend mode)
1.415v Disabled
Prime95 failed after 17 hours (Blend mode)
1.420v Disabled
Prime95 failed after 23:30 hours (Blend mode)
1.425v Disabled
Prime95 passed after 25 hours (Blend mode)
Post a 25-hour endurance test, I believed I had hit stability at 1.425v, so I turned XMP on and re-running the tests:
Vcore XMP
Stress tests
1.425v Enabled
Prime95 failed after 3:58 hours (Blend mode) around FFT 240K
1.425v Enabled
Google stressapptest -
Passed after 12 hours (stressapptest -W -M 32035 -s 43200)
1.425v Passed
After a 12-hour endurance test
1.425v After enabling XMP profile
Prime95 failed after 3:58 hours (Blend mode)
1.425v After enabling
HCI memtest -
Passed
after 9 hours (600% to 1000%)
1.425v Passed
After 4:39 hours
1.425v Passed
After 11:30 hours
1.425v Passed
After 25 hours endurance test
I thought I had reached stability at 1.425v, so I modified VCCIO and VCCSA settings. After a thorough trial, Prime95 continued to fail after about 4 hours, leading me to suspect RAM or controller instability at 5GHz. I conducted multiple RAM stress tests, and interestingly, all checks (stressapptest, HCI memtest, MemTest86) completed successfully.
I’ve read that adjusting VCCIO and VCCSA can help when XMP causes issues, but I’m unsure why only Prime95 was affected while others passed. Should I tweak one or both settings?
The voltages recorded during testing were:
DRAM: 1.349v
VCCIO: Auto - 1.270v
VCCSA: Auto - 1.263v
Any advice would be greatly appreciated!
Thank you.
Spoiler:
MemTest86
HCI memtest
stressapptest
You encounter issues with prime95 due to excessive CPU and memory usage simultaneously, unlike with tools like memtest which focus solely on RAM testing. Proceed by running a small FFT preset to check if the CPU passes—smaller FFT sizes can fit in the cache, reducing memory strain. Also, note that prime95 is a power-related bug program and real-world loads won’t reflect this.
To wrap up this discussion, I decided to leave the iGPU disabled, since turning it on caused the overclocking to become unstable (Prime95 kept failing). Once I achieved stability using the XMP profile with these configurations:
Vcore: 1.425V
CPU ratio: x50
Cache ratio: x42
AVX offset: x0
XMP: Enabled - DDR4 3600Mhz - 32GB (8GBx4) - 16-16-16-36
Vcore mode: Manual (override)
LLC: 5
(on a scale of 1 to 8)
VCCIO: 1.2625V
VCCSA: Auto
I opted to keep the iGPU off, as restarting it would be a lengthy process trying to identify what was causing the instability. Instead, I experimented by connecting a second monitor via HDMI and setting it as the main display in Windows, so the iGPU became the primary GPU rather than the GTX 1080 Ti. I then launched both Prime95 and FurMark simultaneously to stress-test the CPU and iGPU, which led to Prime95 failing after just a few minutes. I adjusted the VCCSA and iGPU voltage offset, running both tests concurrently, but Prime95 continued to fail after a short time. Eventually, I stabilized by raising both the VCCSA and Vcore voltages. I increased the Vcore from 1.425V to 1.430V and the VCCSA from Auto (1.2500V) to 1.2750V, and after about an hour of Prime95+FurMark testing without errors, I began overclocking the cache clock up to 4.6GHz. This allowed me to complete a 12-hour stress test successfully using these settings.