The maximum single-node Hyperion platform for AI training, molecular modelling, generative engineering and large accelerated workflows.
This brief explains where the baseline fits, the software and operating model it supports, and the evidence Axiotech should use to validate a final configuration. Indicative specifications remain subject to component availability, export compliance and formal quotation.
Baseline architecture
One 4U node; 128 AMD EPYC cores; 1.15TB ECC DDR5; four NVIDIA RTX PRO 6000 96GB GPUs providing 384GB aggregate VRAM.
The current compute node is based on the Supermicro AS-4125GS-TNRT / CSE-418G2TS 4U rack platform, integrated with matched rails, power, management, networking and rack infrastructure as required.
Best-fit workloads
Large single-node training, molecular simulation and biomolecular AI, generative engineering, multi-experiment research and workloads needing four 96GB GPUs beside 1.15TB host memory.
Use it when four high-memory GPUs, very large host memory and single-node operational simplicity are preferable to a networked cluster.
Recommended software stack
CUDA, NCCL, PyTorch Distributed, TensorFlow/JAX, BioNeMo or domain frameworks, NVIDIA Container Toolkit, MLflow, Prometheus and Nsight.
Pin host drivers and infrastructure separately from versioned application containers. Record source, image, dataset/model and hardware allocation with each benchmark or production release.
Deployment pattern
Use scheduler or container allocations even on one node when several teams share it. Pin framework/NCCL versions, place scratch/checkpoints on NVMe and promote validated images through a registry.
Define monitoring, identity, backup, change control and workload ownership at the same time as compute. Multi-user platforms require resource allocation and quotas; production services require health, overload and rollback behaviour.
Sizing boundary
Four-GPU jobs must be profiled for communication, CPU input preparation and NVMe throughput. Multi-user scheduling may still justify Slurm or Kubernetes.
Final sizing should use representative code, data, concurrency and service objectives. Aggregate core, RAM or VRAM figures do not by themselves predict application performance.
Commissioning and acceptance
Run one- and four-GPU reference jobs, profile input/collective overhead, verify simultaneous-user policy, validate checkpoint recovery and document thermal/power behaviour under sustained load.
Axiotech should retain the resulting configuration, firmware/driver baseline, environment manifest, benchmark data and recovery procedure as the system acceptance pack.
Primary technical references
References are provided for software architecture and implementation planning. Validate the versions, licences, support matrix and regulated-use requirements applicable to the final deployment.