the i7 4770k remains capped at 4.3ghz even with increased voltages
the i7 4770k remains capped at 4.3ghz even with increased voltages
So here are my system details:
Memory: Gigabyte Z97x-Gaming 3 rev 1.1 BIOS: F8d
RAM: Kingston HyperX Savage 2400mhz 2x8gb custom timings 10-12-13-30-1T
Processor: i7 4770k 4,3ghz 1,3v stable with AIDA64 8h. (replaced, used liquid pro)
GPU: Sapphire Fury Nitro-X unlocked from 56/64 to 60/64, 1100/550mhz
Power Supply: XFX XXX PRO 850w bronze 80+
Storage: 1tb Caviar Blue + PNY XLR8 120gb SSD (OS)
Operating System: Windows 10 Pro 64-bit
CPU Cooler: Custom 360x120mm (no name China), rad, with ID-cooling block and red car fule line tubing. Performed slightly better than H100i at full load. Fans are Corsair RGB 120mm with 1000rpm fixed.
USB devices: mouse, keyboard, HHD hot swap, USB web adapter, and USB soundcard for Creative Wrath headphones.
The story is this: I purchased the CPU with a prebuilt system that had B series MB, so I was confident it wasn’t overclocked. I removed the CPU and installed it in my current setup. After some adjustments, after about a year, I only achieved stability at 4,3ghz with 1,3V, indicating my CPU is one of the least responsive to overclocking.
Based on Extreme tips, I understand the limits for benchmarking (not continuous). I tried up to 1,45V core voltage to see if 4,4ghz would work. No success. I also increased input voltage to 2V, raised ringbus voltage to 1.25V. I lowered memory to stock speeds and kept voltage at 1.6V. Disabled all power-saving features and enabled K overclocking.
With the custom cooler and 1.45V, the CPU reaches 88°C with the hottest core, averaging around 72.5°C. The room temperature is about 25°C.
It seems the CPU simply stops exceeding 4,3ghz regardless of adjustments—could it be the motherboard giving up? A heat scanner shows the motherboard temperatures under 80°C and VRM heatsinks below 70°C, so power input appears safe.
I’ve also noticed that the BIOS freezes if I try to raise the base clock by even 1mhz, even after increasing voltages and lowering frequencies. The same issue occurs when attempting to adjust the gear ratio. It feels odd that the CPU gives up at certain multipliers, which should be fixable by raising the base clock, but that didn’t work either.
I’m expecting problems on the motherboard, so I’m seeking your advice.
I know the i7 4770k is one of Intel’s least favorable CPUs for overclocking, but at 1.45V I should still see at least 4,4ghz.
So something unusual is occurring and I managed to drop my core voltage to 1.35v while the BIOS voltage is actually 1.356v. The LLC boost raises it to 1.368v, and after 44 hours of continuous testing in AIDA64 at 1h intervals, I’ve noticed a pattern. Each stress test started at 1.42v with a 0.05V drop per hour. When the input voltage is increased to exactly 2v (which reads as 1.98v in BIOS), and the Input LLC is set to normal, the system seems to stabilize. It appears that maintaining the correct voltage on the input side is key for stability, while lower voltages cause issues. This discovery is interesting, and I’m not doubting the motherboard anymore since this behavior is happening now.
Seems like the input voltage and core ratio have been limiting progress.
I achieved stable operation at 4,4 GHz with a PIV of 1,98V, VID between 1,340–1,356 Vcore, and a RingBus voltage of 1,25V at a 37 ratio.
Now I'm exploring whether 4,5 GHz or even 4,6 GHz is feasible with a 2,2V PIV and a 1,45V Vcore.