multicoremultiprocessor

Multicore multiprocessor is a processor that contains two or more processing cores on a single integrated circuit. Each core can execute instructions independently, enabling parallelism for multi-threaded workloads. In most designs, cores share some resources such as caches and memory controllers, while maintaining separate execution pipelines and registers.

Key design features include a shared L2 and/or L3 cache, an on-die interconnect, and a memory subsystem that serves all cores. Cache coherence protocols, such as MESI, coordinate data in private and shared caches to keep cores operating on current values. Some cores may also implement hardware support for simultaneous multithreading to improve utilization.

Performance gains come from increased instruction throughput, but depend on software parallelism and data locality. Amdahl's law places a limit on speedup from parallel execution of a given task. In practice, multicore systems scale well for parallelizable workloads, while memory bandwidth and contention can limit gains.

Multicore multiprocessors are distinct from multi-processor systems that use multiple CPUs on a motherboard. They can be complemented by multiple chips in a server environment, forming symmetric multiprocessing configurations, or deployed as single-chip solutions in consumer devices.

Common examples include Intel Core, AMD Ryzen, and ARM-based cores used in laptops, desktops, and mobile devices. They underpin general-purpose computing, virtualization, scientific computing, and high-demand applications by enabling concurrent task execution.