The Athlon x4 860k has not remained stable above 4.0GHz.
The Athlon x4 860k has not remained stable above 4.0GHz.
Recently, I've been attempting to overclock my CPU after installing a new cooler, hoping to boost system performance. My goal was to reach 4GHz, which proved manageable, so I tried pushing it to 4.2GHz. When I increased the multiplier to 41 and performed a stress test, the system crashed after just one second. I then increased the voltage to 1.386v at 4GHz, where it remained stable at 1.380v. Still experiencing crashes, this time due to a CLOCK_WATCHDOG_TIMEOUT. I raised the voltage further to 1.392v, but it continued to crash. At 1.398v it crashed again, then at 1.404v it stopped staying at that value during the 30-second stress test. CPU-Z indicated fluctuations between 1.392 and 1.404v, mostly staying near 1.392v. I suspect the issue might be related to the power supply, although I'm not entirely certain. Temperatures are stable, and I use AMD Overdrive for monitoring and stability tests, with a thermal margin of around 32C to 28C. This is the highest temperature I've recorded, only reaching that level during stress testing. Any advice would be appreciated.
It might depend on the power supply unit of the system not providing sufficient power.
However, the issues you encountered were more connected to the speed at which the CPU core operated.
Often linked to the cache levels of the CPU itself.
Or a mismatch in synchronization between the Cache and the CPU core frequency set.
It could also involve whether the CPU has a DMA memory access controller on its die, connecting it to the main system memory.
If this is out of sync, adjusting the speed of the main system memory in BIOS or selecting a lower HT speed might be necessary.
To maintain synchronization and prevent errors, adding voltage to the CPU core isn't the solution.
Ensuring proper synchronization between the CPU HT and memory speed is key.
It's similar to throwing a ball at a friend who is looking away at the time—because of timing, it doesn't reach him and falls or gets lost.
This results in missed data or corruption during transfers between the CPU and main memory.
Shaun o :
The issue might stem from the power supply unit not providing sufficient power. However, the errors you encountered were more connected to the CPU core's speed settings. It often relates to cache levels or synchronization problems between the cache and the CPU frequency. If the CPU has a DMA memory access controller, it might not be properly synchronized with the main system memory. In such cases, adjusting the main system memory speed in BIOS or opting for a lower HT speed could help. It’s important to maintain synchronization between the CPU's HT and memory speeds. Think of it like passing a ball: if timing is off, the ball won’t be caught. If one person is out of sync, the data gets missed or corrupted during transfers. Lowering RAM speeds might be necessary to address these sync issues. Also, note that your RAM is an 8GB module, not a dual-channel setup.
Shaun o :
It might be due to the power supply unit of the system not providing enough power.
However, the errors you were encountering were more connected to the speed at which the CPU core was operating.
Often linked to the cache levels of the CPU itself.
Or a synchronization issue between the cache and the CPU core frequency set.
It could also be related to whether the CPU has a DMA memory access controller on its die, connecting it to the main system memory, which is out of sync.
In some cases, you might need to lower the speed of the main system memory in the BIOS.
Or opt for a lower HT speed.
To prevent errors, trying to keep everything in sync is key. It’s like throwing a ball to a friend who’s looking away at the time, and because of the wrong timing, it drops or isn’t caught—meaning one person is out of sync.
The outcome is data being missed or corrupted during transfers and requests between the CPU and main system memory, and back again.
Quick Update: I used BlueScreenView to investigate the crashes and found this:
http://imgur.com/a/2dBGs
Going from bottom to top, each time I increased the voltage by 0.006v. Most issues seem linked to atikmdag.sys, which heavily interacts with my GPU. I’m still unsure what’s causing it.