Shared-memory architecture

HSA defines a special case of memory sharing, where the MMU of the CPU and the IOMMU of the GPU have the an identical pageable virtual address space.

In computer architecture, shared memory architecture (SMA) refers to a multiprocessing design where several processors access globally shared memory.

Shared memory architectures may use:

• Uniform memory access (UMA): all the processors share the physical memory uniformly;
• Non-uniform memory access (NUMA): memory access time depends on the memory location relative to a processor;
• Cache-only memory architecture (COMA): the local memories for the processors at each node is used as cache instead of as actual main memory.

In a shared memory architecture system processors communicate by reading and writing memory locations. The two key problems in scaling an shared memory architecture system are:

• performance degradation due to "contention" when several processors try to access the same memory location;
• lack of "coherence" if memory is cached and goes out of synch with the original values as modifications take place.

In case different types of processors, like a CPU and a GPU, have a shared memory architecture, the MMU of the CPU and the IOMMU of the GPU have to share certain characteristics, like e.g. a common address space.

See also

• Shared memory (interprocess communication)
• Shared graphics memory
• Heterogeneous System Architecture