Day Rate: £800 per day (Outside IR35)
Duration: 3 months
Location: Fully remote / WFH
Role: We are assisting a leading global cloud provider who seek an HPC AI Architect who specialises in NVIDIA AI architecture.
This is a fully remote opportunity, ideally suited to a High-Performance Computing specialist with a deep technical understanding of NVIDIA Blackwell infrastructure, spanning hardware, software, and storage domains. The ideal candidate will be highly proficient in the following areas:
Hardware
- Advanced multi-GPU system design
- High-speed interconnects: PCIe, NVLink, InfiniBand
- Thermal and power efficiency modelling
- Expert-level knowledge of NVIDIA Blackwell infrastructure
- Optimizing CPU/GPU ratios in HPC environments
- AI/HPC workload simulation and performance analysis
- Precision optimization (FP16, BF16, INT8, etc.)
- GPU utilization profiling and tuning
- Inference workload modeling and scaling
- AI model deployment and performance optimization
- Design and operation of parallel file systems (e.g. Lustre, GPFS)
- Integration and optimization of NVMe storage tiers
- Modeling storage throughput and demand for AI/HPC workloads
