A SysOps administrator must provision a fault-tolerant set of Amazon EC2 instances for a high-performance computing workload that needs the lowest possible latency between nodes. Which steps should the administrator perform to satisfy these constraints? (Choose two.)

To achieve the lowest possible latency between nodes for a high-performance computing workload, launching EC2 instances into a cluster placement group is crucial. This ensures instances are physically close together within a single Availability Zone, maximizing network throughput and minimizing latency. Placing these instances in an Auto Scaling group within a single subnet provides fault tolerance by automatically replacing unhealthy instances while maintaining the low-latency benefits of the cluster placement group. Creating an Amazon EFS file system is for shared storage, not directly for low-latency instance communication. A Multi-AZ Network Load Balancer distributes traffic across multiple Availability Zones, which would increase latency between the instances themselves. A partition placement group separates instances into logical partitions to reduce correlated failures, but it doesn't guarantee the lowest latency between all instances compared to a cluster placement group.