Educational Labs / Computing
High Performance Computing Lab
A production-grade HPC cluster for research and postgraduate teaching: master and login nodes, multi-GPU AI compute, scheduler and cluster management software, high-speed interconnect, and full rack, cooling and UPS infrastructure.
Lab Components
What the lab is built from
Master Node
Dual-socket 40-core server at 2.6 GHz with 128 GB DDR5 (expandable to 512 GB), 8 TB SSD+HDD RAID storage, redundant PSUs, dedicated management port and Ubuntu 22.04 LTS Server.
High Performance AI Compute Server
Multi-GPU training node with 4× NVIDIA RTX A6000 (48 GB each, NVLinked), 256 GB DDR5 and 3.84 TB SSD in a 2U redundant-power chassis for deep-learning workloads.
Login Node & Job Submission
Dedicated SSH gateway where students compile, stage data and submit batch jobs to the scheduler without touching compute nodes directly.
Cluster Management & Scheduling Software
Full software stack: provisioning, job scheduler, resource quotas, monitoring dashboards, MPI/CUDA toolchains and scientific library modules.
Cluster Networking & Interconnect
Dedicated high-throughput switching plus a configured system interconnect for low-latency node-to-node communication in parallel workloads.
Rack, Cooling & Power Infrastructure
42U rack with accessories, 4× 5HP precision air-conditioning units and a 10 KVA UPS system to keep the cluster online and thermally stable.
Student Workstations
20 Core i7 PCs with 16 GB RAM, 1 TB SSD, dedicated 4 GB graphics and 21.5" displays as the cluster's client access layer.
Experiments & Practical Work
Guided experiments delivered with the lab
Cluster provisioning and node management
Batch job scheduling and resource allocation
Parallel programming with MPI and OpenMP
GPU acceleration with CUDA and cuDNN
Distributed deep-learning training
Scientific simulation and numerical methods
Storage, RAID and parallel file systems
Benchmarking, profiling and performance tuning
Learning Outcomes
What students will master
Administer a Linux HPC cluster end to end
Write and optimise parallel scientific code
Scale AI training across multiple GPUs
Design storage and interconnect for throughput
Plan power, cooling and datacentre readiness
Support faculty research with production compute
Investment Tiers
Build the configuration that fits your programme
Tailor your lab components to match your educational programme needs.
Get in touch
Ready to build the intelligent infrastructure of tomorrow?
Talk to our integration experts about deploying AI platforms, quantum solutions, robotics, or educational labs across your organization.