Apply on Employer Site

INSPYR Solutions · 3 months ago

HPC-Kubernetes Solutions Architect

Dallas, TX

Full-time

Onsite

Senior Level

$200K/yr - $350K/yr

INSPYR Solutions is a national expert in delivering flexible technology and talent solutions. They are seeking an HPC Kubernetes Solutions Architect who will guide customers in designing and integrating GPU-accelerated Kubernetes platforms for high-performance computing, AI/ML training, and scientific workloads.

Information TechnologyProfessional ServicesStaffing Agency

No H1B

U.S. Citizen Only

Responsibilities

Act as the primary architectural point of contact for customers adopting GPU-accelerated Kubernetes platforms for HPC and AI/ML workloads

Partner with customers to capture workload requirements, performance objectives, scaling needs, and integration constraints, translating them into reference architectures and actionable solution designs

Architect and operate Kubernetes clusters optimized for GPU workloads, leveraging NVIDIA

Integrate and tune Multi-Instance GPU (MIG), GPU sharing, and scheduler extensions (e.g., Volcano, Slurm integration, kube-scheduler plugins) to maximize efficiency in multi-tenant environments

Develop or extend custom Kubernetes operators and controllers in Go/Python to automate HPC infrastructure services

Design and recommend secure multi-tenant Kubernetes environments, implementing RBAC, OPA/Gatekeeper policies, namespace isolation, and workload quotas

Lead proof-of-concept and benchmarking engagements, using profiling tools, workload characterization, and telemetry to validate solution performance and scalability

Define and document integration strategies across compute, storage, networking, and orchestration layers, including CNI plugins (NVIDIA CNI, Multus, Cilium), storage systems (Lustre, GPFS, Ceph, VAST), and container runtimes (containerd, NVIDIA Container Toolkit)

Drive observability and monitoring solutions with Prometheus, Grafana, DCGM Exporter, and OpenTelemetry, ensuring visibility into GPU health, cluster utilization, and workload performance

Support GitOps-driven CI/CD pipelines for Kubernetes infrastructure using ArgoCD, FluxCD, Helm, and Kustomize

Collaborate with HPC, ML, and DevOps teams to validate performance and scalability in hybrid or on-premise environments

Provide architectural leadership during onboarding and deployment, ensuring successful integration of Kubernetes clusters with HPC schedulers and enterprise IT systems

Build and maintain strategic relationships with ecosystem vendors (e.g., NVIDIA, Cisco, storage partners), incorporating emerging technologies into customer environments

Share future insights with customers on GPU roadmaps, interconnect advancements (e.g., InfiniBand, RoCE, NVLink), and container orchestration trends

Represent the organization in customer design sessions, technical workshops, and industry conferences, positioning yourself as a thought leader in Kubernetes for HPC

Qualification

Kubernetes architectureNVIDIA GPU stackKubernetes internalsHigh-performance networkingGoPythonWorkload profilingCustomer engagementCollaborative mindset

Required

Extensive experience in Kubernetes architecture and operations for HPC or GPU-intensive environments

Strong technical expertise in NVIDIA GPU stack (GPU Operator, device plugins, MIG, NVML, DCGM)

Kubernetes internals (CRDs, RBAC, scheduler extensions, custom operators/controllers)

Distributed and parallel storage integration with Kubernetes for HPC workloads

High-performance networking (InfiniBand, RDMA, RoCE) in containerized environments

Proven ability to design scalable, secure, and resilient Kubernetes-based architectures for HPC and AI/ML use cases

Proficiency in Go or Python for Kubernetes operator or controller development

Experience with workload profiling, benchmarking, and performance tuning

Strong customer engagement skills, capable of translating requirements into actionable architectures and presenting solutions effectively

Collaborative mindset with experience working across engineering, product, and operations teams

Preferred

Demonstrated success in end-to-end customer solution delivery, from requirements discovery to deployment and adoption

Familiarity with containerized HPC environments (e.g., Singularity/Apptainer)

Exposure to automation and GitOps practices for Kubernetes platform management (e.g., ArgoCD, FluxCD)

Contributions to open-source projects in the Kubernetes or NVIDIA ecosystem

Experience advising on future adoption strategies, helping customers prepare for emerging GPU, interconnect, and orchestration technologies

Bachelor's or Master's degree in Computer Science, Engineering, Physics, or related technical field

Relevant Kubernetes and container certifications such as CKA, CKAD, or CKS, alongside cloud certifications like AWS Solutions Architect or Azure Solutions Architect Expert