F5F Stay Refreshed Hardware Desktop Overclocking CPU becomes inconsistent when using XMP profiles.

Overclocking CPU becomes inconsistent when using XMP profiles.

Overclocking CPU becomes inconsistent when using XMP profiles.

B
Blacklisted_
Junior Member
42
07-20-2017, 09:52 PM
#1
Specs: CPU i7-7700k running at 4.20 GHz with turbo boost up to 4.5GHz (unlocked). Mobo model Asus Maximus IX Extreme GPU: Gigabyte GTX 1080 Ti Waterforce WB Xtreme Edition 11G RAM: 32GB G.Skill Trident Z RGB, DDR4 3600MHz, 32GB (8x4) - 16-16-16-36 PSU: Seasonic PRIME 850W Titanium Custom Cooling. Radiator: two units (520x30x60mm), fan array of eight Noctua (120x120x15mm). Pump: EK-XRES 140 Revo D5, overclocked to 5GHz. Configs used: CPU ratio x50, cache ratio x42, AVX offset 0, Vcore mode manual override. LLC set to 5. Stability tests: VCCIO auto, VCCSA auto. I adjusted Vcore in steps of 0.005V and ran Prime95 stress tests under XMP disabled. Results showed stability after several hours; with XMP enabled, Prime95 still failed after ~3.58 hours. All other benchmarks passed consistently.

Questions: Should I tweak VCCIO or VCCSA values? Is it best to adjust one at a time or both together? I recorded voltages during XMP testing: DRAM 1.349V, VCCIO Auto 1.270V, VCCSA Auto 1.263V. Any advice would be appreciated. Thanks!
B
Blacklisted_
07-20-2017, 09:52 PM #1

Specs: CPU i7-7700k running at 4.20 GHz with turbo boost up to 4.5GHz (unlocked). Mobo model Asus Maximus IX Extreme GPU: Gigabyte GTX 1080 Ti Waterforce WB Xtreme Edition 11G RAM: 32GB G.Skill Trident Z RGB, DDR4 3600MHz, 32GB (8x4) - 16-16-16-36 PSU: Seasonic PRIME 850W Titanium Custom Cooling. Radiator: two units (520x30x60mm), fan array of eight Noctua (120x120x15mm). Pump: EK-XRES 140 Revo D5, overclocked to 5GHz. Configs used: CPU ratio x50, cache ratio x42, AVX offset 0, Vcore mode manual override. LLC set to 5. Stability tests: VCCIO auto, VCCSA auto. I adjusted Vcore in steps of 0.005V and ran Prime95 stress tests under XMP disabled. Results showed stability after several hours; with XMP enabled, Prime95 still failed after ~3.58 hours. All other benchmarks passed consistently.

Questions: Should I tweak VCCIO or VCCSA values? Is it best to adjust one at a time or both together? I recorded voltages during XMP testing: DRAM 1.349V, VCCIO Auto 1.270V, VCCSA Auto 1.263V. Any advice would be appreciated. Thanks!

S
Summits
Junior Member
38
07-20-2017, 09:52 PM
#2
Completed a 25-hour stress test using Prime95 in blend mode, raising VCCIO by one level, from 1.2500V to 1.2625V.
S
Summits
07-20-2017, 09:52 PM #2

Completed a 25-hour stress test using Prime95 in blend mode, raising VCCIO by one level, from 1.2500V to 1.2625V.

N
NightFury009
Junior Member
5
07-20-2017, 09:52 PM
#3
In summary, I decided to leave the iGPU turned off after experiencing instability when it was active. The issue arose with Prime95 failing repeatedly, which made troubleshooting time-consuming. By adjusting settings such as Vcore and VCCSA, I eventually achieved stability. I also experimented with changing the iGPU voltage and running stress tests on both CPU and GPU simultaneously. This process helped me reach a more reliable configuration. Keep in mind these steps might not apply to everyone.
N
NightFury009
07-20-2017, 09:52 PM #3

In summary, I decided to leave the iGPU turned off after experiencing instability when it was active. The issue arose with Prime95 failing repeatedly, which made troubleshooting time-consuming. By adjusting settings such as Vcore and VCCSA, I eventually achieved stability. I also experimented with changing the iGPU voltage and running stress tests on both CPU and GPU simultaneously. This process helped me reach a more reliable configuration. Keep in mind these steps might not apply to everyone.