Cache storing recently processed vertex information.

Non-uniform memory access architecture.

Specialized hardware accelerating matrix and AI calculations.

Storage containing compiled or intermediate shader data.

Processor and memory region within a NUMA system.

Specialized hardware accelerating ray-tracing operations.

Memory structure containing commands for GPU execution.

Hardware managing processor-to-memory communication.

Hardware designed to accelerate ray-tracing calculations.

Hardware or firmware mechanism scheduling GPU workloads.