Struggling to boost my CPU (fx-6300)?
Struggling to boost my CPU (fx-6300)?
So, according to the title, I'm attempting to push the fx-6300 overclocking. It's stable at about 3.6 ghz, but anything higher causes instability in Prime95 during a six-hour test. When I go above 3.6, it initially works well but then drops to around 24 MHz (at 4ghz). I'm using an M5A97 R2.0 motherboard. At normal speeds, the idle temperature is about 27°C (likely +5°C due to measurement error), and under load it reaches around 47°C.
Every board increases overclocking in its own way, you just need to keep experimenting until you reach the optimal setting. I'm currently at 2795 ht and 215fsb with the same 4.4ghz CPU, and the games run smoothly and fluidly. Previously, I reached up to 220fsb but experienced core drops in p95 with the default configuration before the 6-minute mark, or I would freeze during a p95 small FFT test.
The small FFT test will put stress on your CPU and verify the overclock strictly. It’s likely to fail within the first 5-6 minutes by freezing the system, but if you manage to reach 7 minutes or longer, it should be safe to continue. The blend test pushes all components, so any weakness in RAM, CPU, northbridge, clocks, or voltages will cause failure within the first 6 minutes.
I don't perform overclocking. I don't have an AMD Overdrive program installed. I'm not using any CPU cooler at the moment.
I continuously adjust the multiplier until my computer freezes or displays an error. When freezing occurs, I modify the voltage while keeping the temperature below 62°C. I repeatedly increase the voltage and multiplier until reaching temperatures near 60°C.
I have AMD Overdrive enabled but prefer using the BIOS.
I own a LC-CC-95 cooler.
P.S. I'm not attempting extreme overclocks, just aiming for around 4 GHz.
Adjusting your bclk together with the multiplier brings the processor to 4ghz with an 8320, maintaining stability even at 1.25v instead of the previous 1.425v target.
Boosting the bclk(multiplier) causes my computer to crash in 2 to 10 minutes (during prime95). Raising the voltage leads to extremely high temperatures (above 62°C) after about 2 to 4 hours of running prime95.
Honestly, hours of prime 95 feel exaggerated. I run the small fft test for about five minutes just to check for freezing issues, and the default test for four to six minutes until it passes without any core failures. My setup is at 4.4ghz with 210fsb and 2730ht link.
I had to slightly overvolt the CPU and the CPU-NB a bit to stabilize it, but unless you’re running continuous heavy tasks that need precision—like medical computing—I’d recommend pushing it until it hits 4.2–4.6ghz and then doing quick stability checks instead of long sessions.
The small fft test will stress the CPU and may cause freezing if unstable. The regular p95 test will fail under low or high voltage, or if RAM OC is incorrect, or if you skip overvolting.
Good luck! I’ve spent a lot testing my own CPU, so I hope my advice is useful. When pushing my OC beyond 4.4ghz, the p95 tests fail quickly under load, which means glitches in games—so reduce the OC if it starts to fail.
What approach should I take to check the stability of my CPU? Many users rely on hours of Prime95 to simulate stress, but I’m curious if there’s a better way to assess it without overloading the system. I’m also concerned about possible issues in the BIOS settings.
Every board increases overclocking in its own way, so keep experimenting until you reach the optimal setting. I'm currently at 2795 ht and 215fsb with a 4.4ghz CPU, and the games run smoothly and without issues. Previously, I reached 220fsb but experienced core drops in p95 under default settings before the 6-minute mark, or freezing occurred during a small FFT test.
The FFT test will put significant strain on your CPU, potentially causing it to overheat and fail within the first few minutes. If you manage to reach 7 minutes or longer, it should be safe to continue. The blend test is designed to stress every component, so any weakness in RAM, CPU, northbridge, clocks, or voltages will likely cause failure within the first six minutes.
Even if the test fails afterward, I still consider the overclock stable since normal gaming won’t push the CPU too much. However, OCCT can be a useful tool to quickly identify problems if your overclock is too high. It’s extreme stress that usually leads to issues, and you’re unlikely to subject your CPU to such conditions under regular use. I stop OCCT after just over 1 minute and 30 seconds.