HPC Structure and FAQ

This HPC system uses Slurm as the job scheduler. 

Slurm is an open source, fault-tolerant, and highly scalable cluster management and job scheduling system for large and small Linux clusters. Slurm requires no kernel modifications for its operation and is relatively self-contained. As a cluster workload manager, Slurm has three key functions. First, it allocates exclusive and/or non-exclusive access to resources (compute nodes) to users for some duration of time so they can perform work. Second, it provides a framework for starting, executing, and monitoring work (normally a parallel job) on the set of allocated nodes. Finally, it arbitrates contention for resources by managing a queue of pending work.

Detailed instruction can be found here

The High-Performance Computing (HPC) system consists of multiple compute nodes, each optimized for parallel processing, scientific simulations, and data-intensive computations. Below are the specifications of a compute node:

• Compute Cores: 128 cores

• Operating System: Ubuntu 22.04.5 LTS (x86_64)

• Host Machine: Dell PowerEdge R7525

• Kernel Version: 5.15.0-130-generic

• Shell: Bash 5.1.16

• CPU: AMD EPYC 7702 (256 threads) @ 2.0 GHz

• GPU: NVIDIA Tesla V100S PCIe 32GB

• Total Memory: 1,031,595 MiB

This node features an AMD EPYC 7702 processor with 256 logical cores (128 physical cores with SMT). It is well-suited for multithreaded applications and parallelized workloads. 

The system includes NVIDIA Tesla V100S PCIe (32GB), optimized for deep learning, CUDA-based applications, and accelerated computations. 

• Check CPU/GPU/memory usage (htop, nvidia-smi, free -h).

• Optimize your script for parallel execution if not used.

• Check your SLURM header for any over-allocation of resources.

• Contact the HPC administrators for assistance.