300-640 Dumps with Practice Exam Questions Answers

backend network

frontend network

storage network

management network

Large AI environments normally separate traffic into purpose-built fabrics because GPU collective communication, storage I/O, application access, and infrastructure management have fundamentally different performance and availability characteristics. Cisco identifies the backend network as the east-west fabric dedicated to inter-GPU communication. It carries high-bandwidth RoCEv2/RDMA traffic associated with distributed training and must provide exceptionally low latency and lossless behavior.

The frontend network provides north-south connectivity between AI nodes and the wider data center, supporting application access, inference traffic, general data movement, logging, and other in-band services. The storage network connects compute nodes to shared datasets, model checkpoints, and high-performance parallel or distributed storage systems. Cisco AI POD designs may use dedicated storage-leaf connectivity to prevent storage traffic from interfering with backend GPU collective communication. The management network serves out-of-band management interfaces for servers, switches, storage platforms, and other infrastructure components. Cisco AI reference architectures explicitly distinguish backend, frontend/storage, and management connectivity for these purposes.

Maintaining these traffic domains independently improves deterministic performance, fault isolation, congestion control, security, and operational manageability.

Study Guide Reference: AI Infrastructure Components and Architecture — network fabrics, bandwidth, latency, scalability, and AI infrastructure connectivity.

===============

Priority Flow Control is a foundational mechanism used to provide lossless Ethernet behavior for RoCEv2. PFC operates on individual Ethernet priorities and allows a congested receiver or switch to pause transmission for the affected priority while permitting traffic in other classes to continue. Cisco's AI networking guidance explicitly states that PFC is enabled for RoCEv2 traffic to prevent packet drops caused by congestion.

This characteristic is critical for RDMA workloads because packet loss can trigger retransmission behavior and significantly degrade GPU collective communication performance. In an AI backend fabric, RoCEv2 is normally placed in a no-drop QoS class with PFC configured for that traffic class. ECN is commonly used alongside PFC to provide proactive congestion signaling, but PFC is the mechanism that directly prevents loss when buffers approach critical utilization.

Traffic shaping in option A may moderate burst rates but does not establish lossless transport. A high DSCP marking in option C provides classification and forwarding preference, not loss prevention. Weighted fair queuing in option D allocates scheduling bandwidth but also does not guarantee that packets will not be discarded during congestion.

Study Guide Reference: 3.1.a Congestion control mechanisms — PFC, ECN, ETS; 3.1.b RoCE/RoCEv2; 3.1.c QoS.

===============

The server contains six 3000-W supplies and must preserve N+2 redundancy . A 4,500-W workload requires two actively supplying PSUs because a single 3000-W supply cannot support the current load. Maintaining N+2 redundancy requires two additional PSUs to remain available beyond those required for the workload. The minimum operational complement is therefore four supplies: two supporting the power demand plus two retained for redundancy.

With six PSUs physically installed, the remaining two PSUs can be placed into standby , making option C correct.

The calculation is:

Required active capacity = ceiling(4500 W ÷ 3000 W) = 2 PSUs

Required redundancy = 2 PSUs

Minimum available PSUs = 2 + 2 = 4

Installed PSUs = 6

Eligible for standby = 6 − 4 = 2 PSUs

Power Save Mode is intended to reduce conversion losses when installed PSU capacity exceeds current demand. Cisco uses the same operating principle in UCS power management: additional supplies that are not needed to satisfy both the requested power budget and configured redundancy can be placed into power-saving operation.

Placing three supplies into standby would leave only three available supplies, violating the N+2 requirement for a workload requiring two PSUs.

Study Guide Reference: AI Infrastructure Operations and Troubleshooting — power efficiency, PSU redundancy, Power Save Mode, and UCS AI server power management.

===============

Workload scheduling determines where and when computational jobs execute by assigning workloads to available processing resources according to resource requirements, capacity, policies, priorities, and performance objectives. In AI environments, this commonly means allocating training, fine-tuning, or inference jobs to CPUs, GPUs, GPU memory, or groups of accelerator nodes so expensive computational resources remain efficiently utilized.

Cisco's DCAI blueprint explicitly includes orchestration as an AI-environment component. Cisco material covering NVIDIA Run:ai also identifies scheduling and GPU orchestration as core AI workload-orchestration capabilities and describes infrastructure pooling and resource management designed to maximize compute efficiency. Therefore, assigning tasks to compute elements for optimized performance precisely describes workload scheduling.

Option B concerns software licensing and compliance management. Option C is monitoring or telemetry aggregation, which can inform scheduling decisions but is not scheduling itself. Option D is a backup and recovery operation.

Effective AI scheduling is particularly important because GPU resources are expensive and workloads vary significantly in duration and resource demand. A capable scheduler therefore improves accelerator utilization, reduces queueing, and coordinates workload placement across the available compute infrastructure.

Study Guide Reference: 1.0 AI Fundamentals and Applications — 1.5.d Orchestration.

===============

RoCEv2 transports the InfiniBand transport protocol over UDP/IP, allowing RDMA traffic to traverse Layer 3 networks. The RDMA transport layer preserves packet-level sequencing, reliability, and ordering semantics that are inappropriate for arbitrary IP fragmentation. Consequently, RoCEv2 traffic is designed to traverse the network without IP fragmentation, making option D the intended technical answer. NVIDIA documentation confirms that RoCEv2 places the InfiniBand transport protocol behind UDP and IP rather than using the native InfiniBand network-layer header.

A consistent end-to-end MTU is therefore essential in GPU fabrics. AI clusters commonly use jumbo frames, and every switch interface, port channel, Layer 3 interface, and server NIC along the path must support the configured packet size. Unexpected fragmentation is incompatible with the deterministic behavior required by high-performance RDMA workloads.

Option A identifies a general IPv4 consequence of setting DF—Path MTU Discovery can use DF—but it is not the fundamental RoCE transport requirement tested here. Option B is incorrect because DF does not classify packets as RDMA. Option C is also incorrect because ECMP path selection is normally determined by packet-header hashing; the DF flag does not force packets onto a single path.

Study Guide Reference: AI Infrastructure Deployment and Data Management — RoCEv2, MTU design, lossless transport, and high-performance Ethernet.

===============