OPTION - 4
(A) LOAD and STORE instructions are used to access remote memory.
✅ CORRECT
Proof:
NUMA me single global address space hota hai.
Isliye chahe memory local ho ya remote, CPU usko normal LOAD/STORE instructions se hi access karta hai.
➡️ Ye DISTINCTIVE feature hai NUMA ka.
(B) There is a single address space visible to all CPU.
✅ CORRECT
Proof:
NUMA = Non-Uniform Memory Access architecture
Par address space unified hota hai (distributed physical memory but logically shared)
➡️ Sab CPUs ek hi memory address space dekhte hain.
(C) Access to local memory is slower than access to remote memory.
❌ INCORRECT
Proof:
NUMA ka core principle hi yeh hai:
Local Memory Access Time < Remote Memory Access Time\textbf{Local Memory Access Time < Remote Memory Access Time}Local Memory Access Time < Remote Memory Access Time
Local memory = faster
Remote memory = slower
Statement ne ulta likh diya, isliye (C) NOT CORRECT ✅
(D) When the access time to remote memory is hidden, the system is called NC-NUMA.
❌ INCORRECT
Proof:
NC-NUMA ka full form hota hai:
Non-Cache Coherent NUMA\textbf{Non-Cache Coherent NUMA}Non-Cache Coherent NUMA
Iska latency hiding se koi lena-dena nahi hai.
NC-NUMA ka matlab hota hai:
Latency hiding = software/hardware optimization, NUMA type ka naam nahi.
Isliye (D) bhi NOT CORRECT ✅
(E) In CC-NUMA, Coherent caches are present.
✅ CORRECT
Proof:
CC-NUMA = Cache Coherent NUMA
Iska matlab hi hota hai: