F5F Stay Refreshed Power Users Overclocking Broadwell-E thermal limit at 6850k

Broadwell-E thermal limit at 6850k

Broadwell-E thermal limit at 6850k

R
ReborntoKill
Posting Freak
821
03-21-2016, 12:01 AM
#1
I searched online using the title you provided, but didn't find detailed information about your setup.
I've customized my 6850k to 4.3 @ 1.35v with dram at 2400, running 10 hours P95 stable. The HWmonitor reports a maximum temperature of 62°C.
With the 3200 MHz RAM, I considered applying the XMP profile while keeping the CPU overclocked. P95 failed after 7 minutes at 4.3 @ 1.35v, so I increased the voltage to 1.38. Since then, P95 has remained stable for about an hour and ten minutes, though the peak CPU temperatures reached 86°C.
I'm wondering if this is pushing the CPU beyond its safe operating range—whether it's approaching a throttle limit or entering a "danger zone" for CPUs around 90°C, 100°C, or 110°C.
For reference, my current rig includes: I7 6850k @ 4.3 1.38v, MSI X99-A SLI plus motherboard, 32GB Corsair Vengeance DDR4 3200 (16x2), Corsair H110i GTX cooler, 512GB Samsung 850 EVO SSD, and a Cougar 1200w 80+ bronze PSU. I'm using an XFX 390x (two units) crossfire.
Appreciate any advice you can offer.
R
ReborntoKill
03-21-2016, 12:01 AM #1

I searched online using the title you provided, but didn't find detailed information about your setup.
I've customized my 6850k to 4.3 @ 1.35v with dram at 2400, running 10 hours P95 stable. The HWmonitor reports a maximum temperature of 62°C.
With the 3200 MHz RAM, I considered applying the XMP profile while keeping the CPU overclocked. P95 failed after 7 minutes at 4.3 @ 1.35v, so I increased the voltage to 1.38. Since then, P95 has remained stable for about an hour and ten minutes, though the peak CPU temperatures reached 86°C.
I'm wondering if this is pushing the CPU beyond its safe operating range—whether it's approaching a throttle limit or entering a "danger zone" for CPUs around 90°C, 100°C, or 110°C.
For reference, my current rig includes: I7 6850k @ 4.3 1.38v, MSI X99-A SLI plus motherboard, 32GB Corsair Vengeance DDR4 3200 (16x2), Corsair H110i GTX cooler, 512GB Samsung 850 EVO SSD, and a Cougar 1200w 80+ bronze PSU. I'm using an XFX 390x (two units) crossfire.
Appreciate any advice you can offer.

R
ReD_T1000
Member
168
03-21-2016, 05:03 AM
#2
I wouldn't exceed the 86 degs you're recording. Consider gradually reducing the core voltage in small steps (0.05V increments) and perform additional stability checks. You might also experiment with adjusting the CPU SA and CPU VCCIO voltages to approximately 1.15, while slightly lowering the v-core. See if this improves the situation.
R
ReD_T1000
03-21-2016, 05:03 AM #2

I wouldn't exceed the 86 degs you're recording. Consider gradually reducing the core voltage in small steps (0.05V increments) and perform additional stability checks. You might also experiment with adjusting the CPU SA and CPU VCCIO voltages to approximately 1.15, while slightly lowering the v-core. See if this improves the situation.

R
ronny2003
Junior Member
49
03-21-2016, 06:59 PM
#3
Intel suggests using its server chips nonstop at 103-107°C, which aligns with my memory. Most users, however, favor keeping the chip below 80°C for continuous operation, and I concur. Surges to 80°C or higher, or under heavy stress (Prime95 is risky and doesn't reflect real conditions), are acceptable. The short version is your call.
R
ronny2003
03-21-2016, 06:59 PM #3

Intel suggests using its server chips nonstop at 103-107°C, which aligns with my memory. Most users, however, favor keeping the chip below 80°C for continuous operation, and I concur. Surges to 80°C or higher, or under heavy stress (Prime95 is risky and doesn't reflect real conditions), are acceptable. The short version is your call.

A
axxerr77
Member
120
03-21-2016, 07:55 PM
#4
I wouldn't exceed the 86 degs you're recording. Consider gradually reducing the core voltage in small steps (0.05V increments) and perform additional stability checks. You might also experiment with adjusting the CPU SA and CPU VCCIO voltages to approximately 1.15, while slightly lowering the v-core. See if this improves the situation.
A
axxerr77
03-21-2016, 07:55 PM #4

I wouldn't exceed the 86 degs you're recording. Consider gradually reducing the core voltage in small steps (0.05V increments) and perform additional stability checks. You might also experiment with adjusting the CPU SA and CPU VCCIO voltages to approximately 1.15, while slightly lowering the v-core. See if this improves the situation.