How much parallelism?

Understanding hardware threads and core visibility in operating systems

This is an easy question that tests your grasp of the distinction between physical cores, logical cores, and the threads the operating system can actually see and schedule. It's foundational knowledge for anyone working on systems where CPU architecture directly impacts performance—particularly relevant for low-latency trading systems and high-frequency infrastructure.

The question requires you to reason carefully about how hyper-threading (Intel's simultaneous multithreading) relates to the number of schedulable units the OS perceives. When hyper-threading is disabled, the relationship between physical and logical cores changes in a specific, predictable way. Understanding this relationship is essential for diagnosing performance bottlenecks, tuning thread affinity, and predicting parallelism limits in production systems.

• Physical vs. logical cores
• Hyper-threading and SMT (simultaneous multithreading)
• OS thread scheduling and CPU topology