TurboBoost fails if ASUS Turbo Mode isn't active on Intel Core 10700F systems.
TurboBoost fails if ASUS Turbo Mode isn't active on Intel Core 10700F systems.
Windows allocates not all threads equally. When launching games, especially on a 6/12 CPU, turning on Wi-Fi and keeping a few background programs running can strain one core. Based on past observations, you typically experience a single core being heavily impacted—around 60-70% of cases—often confirmed by hwinfo64 reports.
I have another budget system with an i5 from 2015 priced at $400 that operates at around 0.5-2% CPU usage. My thought is it might relate to the motherboard or power configurations. When I set lower core counts and speeds, the activity of "System" and "Interrupts" drops and shifts, almost like it’s idling. Running high-end games at full settings keeps temperatures steady between 60-70°C, and everything runs smoothly. I also notice the "System" process decreasing its usage when other apps demand more resources. Update: The threads in "System" (ACPI might be responsible for the interrupts shown in System Interrupts, assuming that?)
I understand. These belong to me while waiting. Refer to the previous update for details. (Current - Min - Max - Avg)
The System and Interrupts shouldn't be consuming your CPU as much as they are. It seems a driver might not be functioning properly. If a driver becomes stuck in an infinite loop with no progress, it can force one thread to operate at full capacity while the other remains idle. This behavior is unusual for Windows 10 and usually indicates a problem. The CPU utilization numbers in Task Manager change depending on the processor speed. As MHz rise, utilization also increases. Conversely, when the speed drops, utilization decreases. Many people overlook this connection, thinking it reflects actual workload, but it doesn't. Disabling Turbo Boost causes the CPU to run slower, which naturally lowers utilization—a normal response. This metric isn't a reliable indicator of task activity. For instance, with Turbo Boost off at 3600 MHz, a 10-thread test on a 20-thread CPU yields 50% load and 50% utilization. When speed jumps to 5000 MHz, the same load shows 69% utilization due to scaling factors. At 800 MHz, the same load drops to 11%. This demonstrates that utilization values are adjusted based on CPU frequency. If Turbo Boost remains disabled at 800 MHz, utilization stays low (around 11%). The key takeaway is to identify and fix the driver issue for smoother performance.
Thank you for your feedback. The issue seems related to the Thunderbolt header or PCI Express, though I’m still trying to understand it fully. In my earlier message, I’ve attached a new thread where I explain my approach and request further clarification. It appears this type of problem isn’t too rare, with few solutions available.
I just noticed your previous discussion and concurred—it should have its own thread. Your efforts in identifying the bug are commendable. There appears to be faulty code causing it to hang, with only one thread loading fully. It could take some time before this issue gets resolved.