4770k Oced Stable or Not
4770k Oced Stable or Not
Hi, I've tested my 4770k at 4.3 ghz running Prime95 for two hours without any BSODs. It was only after five minutes that a BSOD appeared. I'm looking for reliable programs to use. Before starting OCCT, my computer has been heavily used for gaming and hasn't shown any BSODs. The settings I have are: CPU Base Clock: Manual, Core Ratio: 43, Turbo Boost: Off, Cpu Enhanced Halt C1E: Off, C3 State Support: Off, C6/C7: State Support Monitor: Off, CPU EIST Function: Off, VRRIN External Override: 1.8v, VCore: 1.250v, RING Voltage: 1.050v. Any suggestions?
P95 offers limited overall stress testing capabilities but performs well (v26.6 small fft) for temperature peaks. Occt, particularly during linpack execution, demands greater system resources. If passing p95 without problems, it suggests you're mainly targeting core temperatures rather than stressing RAM, VRM, or the MC. Consider increasing the RAM voltage slightly to 1.505 or 1.51 and adjusting the MC's workload. At 4.6GHz, the Vcore should stay below 1.3v. Beyond just changing multipliers, adjustments like raising the LLC to around 50%, setting current usage at 110%, turning off Eco features, disabling phasing, using c-states below c-3, modifying PLR to 1.9 or 1.7, and setting long duration limits to 255 may be necessary.
It's not a big change from the original. I went up to 4.6ghz and used prime95 as a good benchmark. It worked for me, though I got a BSOD when starting the OCCT. The main point is whether I should stick with 4.3 after a BSOD with OCCT or rely on prime95 since the OC didn't crash?
That's not a very original character. Consider increasing to 4.6ghz or more; Prime95 works well for measuring. It’s sufficient for me—I tried 4.6 with a 1.3 Vcore, but it caused a BSOD after starting the OCCT. My main question is whether I should stick with 4.3 after a BSOD with OCCT or rely on Prime95 since the OC didn’t crash? 1.3 vcore isn’t that high either. You might want to try raising it further and check temperatures and stability. I’d suggest aiming for around 1.45 vcore myself.
Depends on the Prime95 version you're using, did your tests cover memory or just the CPU?
CPU only.
Prime95 configuration:
Preset: Custom
Min FFT: 1344
Max FFT: 1344
In-place FFTs: Enabled
Min FFT: 8 Max FFT: 4096
In-place execution: Verified
ObelixThe :
That's not a big change for me. Consider going up to 4.6ghz or more; prime95 works well as a benchmark. It's sufficient for me—tested 4.6 with 1.3 Vcore, but BSOD occurred after starting the OCCT. The main point is whether I should stick with 4.3 after a BSOD with OCCT or rely on prime95 since the OC didn't crash?
1.3 vcore isn't very high either. You might want to try increasing it and check temperatures and stability. I'd aim for around 1.45 vcore myself.
Intel suggests a max Vcore range of 1.3 to 13.5.
P95 offers limited overall stress testing capabilities, yet performs well (v26.6 small fft) for temperature peaks. Occt, particularly when using linpack, demands significantly more resources across the system. If passing p95 without problems, it suggests you're mainly targeting core temperatures rather than stressing RAM, VRM, or MCs. Consider increasing the RAM voltage slightly to 1.505 or 1.51 and adjusting the MC's workload.
At 4.6GHz, the vcore should stay below 1.3v. Beyond just changing multipliers, you'll need to raise the LLC to around 50%, increase current usage to 110%, turn off Eco settings, disable phasing, set c-states below c-3, and possibly tweak PLR to 1.9 or 1.7, while adjusting long duration limits to 255. These adjustments apply across most systems, not just the Asus ROG mobo, which runs the most demanding OC bios with many uncommon or user-adjustable options.
Other models like the i5-3570k @4.3GHz 1.114v and i7-3770k @4.9GHz 1.308v have similar BIOS features, though they differ in usability and performance trade-offs.