Principal AI Cloud Solutions Architect – GPU HPC Infra

IREN; San Francisco (CA); On-site; USD 180,000 - 260,000; Full time

Benefits offered by this job

Medical, dental, and vision insurance

401(k) plan with company match

Paid time off

Relocation assistance

Flexible work arrangements

Job summary

IREN is a vertically integrated AI Cloud provider delivering large-scale data centers and GPU clusters for AI training and inference. You will partner with client CTOs and ML engineers to design, benchmark, and optimize massively distributed workloads across GPU clusters.

You will translate complex requirements into optimized hardware/software configurations and create reference architectures that set industry benchmarks, while leading hands-on PoCs and performance evaluations.

Qualifications

• 8–10+ years as a Solution Architect or similar in enterprise cloud/HPC/AI.
• Hands-on GPU density experience with NVIDIA HGX/DGX/Blackwell platforms.
• Proficient with Kubernetes, CNI, and parallel storage systems.
• Experience leading workshops and presenting to C-suite executives.

Responsibilities

• Design, benchmark and optimize large distributed AI training/inference clusters.
• Translate customer requirements into optimized hardware/software configurations.
• Create end-to-end reference architectures for GPU clusters and multi-tenancy.
• Lead PoCs and performance benchmarking (NCCL-tests, Linpack, MLPerf, IO500).
• Collaborate with client engineers to optimize training workloads (Megatron-LM, DeepSpeed).

Skills

GPU clusters

HGX/DGX hardware

RDMA networking

Kubernetes

Networking (CNI)

Parallel file systems

Performance benchmarking

Education

MS in CS/CE or quantitative field

BS in CS/CE or quantitative field

Tools

Ceph/Lustre storage

InfiniBand/RoCE

Job description

IREN is a vertically integrated AI Cloud provider delivering large-scale data centers and GPU clusters for AI training and inference. You will partner with client CTOs and ML engineers to design, benchmark, and optimize massively distributed workloads across GPU clusters.

You will translate complex requirements into optimized hardware/software configurations and create reference architectures that set industry benchmarks, while leading hands-on PoCs and performance evaluations.