Solution

The correct answer is (C) and (D) Only

EXPLANATION:

• (A) LOAD and STORE instructions are used to access remote memory.
  • This statement is CORRECT because in a NUMA system, a processor can access its own local memory faster than non-local memory (memory associated with a different processor), and the LOAD and STORE instructions are used for this memory access.
• (B) There is a single address space visible to all CPU.
  • This statement is CORRECT. All CPUs see a single shared address space which is broken into multiple nodes. Each node has local memory which it can access faster than memory in the other nodes.
• (C) Access to local memory is slower than access to remote memory.
  • This statement is INCORRECT. In a NUMA system, access to local memory is faster than access to remote memory (memory local to another processor or memory region).
• (D) When the access time to remote memory is hidden, the system is called NC - NUMA.
  • This statement is INCORRECT. When the high latency of remote memory access is hidden by caching or prefetching techniques, the system is called Cache-Only Memory Architecture (COMA), not NC-NUMA.
• (E) In CC-NUMA, Coherent caches are present.
  • This statement is CORRECT. CC-NUMA stands for Cache Coherent Non-Uniform Memory Access. In CC-NUMA, all nodes have a coherent view of the memory – meaning that a change of a value in one node's cache is effectively and timely propagated to the caches of the other nodes that have cached that value.