Exam NCP-MCI v6.5 topic 1 question 38 discussion

The provided specification would create 2 NUMA nodes, each with 12 cores and 128GB RAM. Therefore the requested specification would put the VM on a node with insufficient CPU to place the whole VM on one NUMA node creating whats called a "wide VM". Creating 2 vNUMA nodes would therefore resolve this issue which is the recommended method of dealing with this issue. See https://portal.nutanix.com/page/documents/solutions/details?targetId=BP-2029-AHV:nonuniform-memory-access-and-guest-vms.html B is incorrect because the amount of memory needed for the VM is more than 25% of the total physical memory available. According to nutanix docs this can result in worse performance. C is incorrect because affinity policies have nothing to do with performance in this context. D is incorrect because that relates to using more memory than you have available (effectively using disk as memory)

The correct answer is A. With 2 vNUMA nodes. Explanation: Configuring the VM with 2 vNUMA nodes allows for better resource distribution across physical CPUs and memory. Given the size of the VM (16 vCPUs and 96GB of RAM), enabling two vNUMA nodes ensures that resources are allocated in a way that minimizes latency and optimizes CPU and memory access, thereby improving overall performance.

Answer appears correct: https://next.nutanix.com/how-it-works-22/enabling-vnuma-on-your-ahv-cluster-38612 and https://portal.nutanix.com/page/documents/details?targetId=AHV-Admin-Guide-v6_5:ahv-vm-enabling-vnuma-vuma-t.html