HPC System Architecture - United States

Originally posted 6 October 2026 by the employer.

About the role

This involves shaping silicon requirements and defining system-level architecture for large-scale, scale-out clusters.

What you'll do

• Define scale-out interconnect architecture, covering fabric selection, network topology, bisection bandwidth, blocking ratios, collective and communication performance, and resilience.
• Partner with CPU and GPU teams to define product requirements for future HPC systems, providing system-level and workload evidence.
• Build performance models to project delivered application throughput at scale on unreleased silicon and drive configuration and roadmap decisions.
• Lead architecture co-design and benchmark definition with HPC customers, including acceptance criteria for application suites.
• Drive HPC requirements into Intel's CPU, GPU, platform, and software roadmaps.
• Produce reference architectures and system specifications for platform teams, OEM, ODM partners, and system integrators.

What you'll need

• Bachelor's degree in Computer Engineering, Computer Science, Electrical Engineering, Physics, or a related technical field.
• 10+ years of experience in HPC, data center, or server system architecture, design, or engineering.
• 5+ years of experience in a customer-facing or externally-facing technical role.
• Demonstrated experience defining or deploying large-scale compute clusters.
• Deep expertise in HPC system architecture at cluster scale, spanning CPU and GPU configuration, memory subsystem design, scale-out interconnect, and parallel storage.
• Demonstrated experience defining systems around unreleased silicon and influencing CPU or accelerator requirements from a system and workload perspective.
• Fluency with HPC workload characteristics and performance analysis, including quantitative reasoning about scaling behavior, communication patterns, and bottlenecks at scale.
• Working knowledge of the HPC software stack, including MPI, OpenMP, GPU programming models, schedulers, and math and communication libraries.
• Strong understanding of the competitive HPC landscape, including alternative CPU and accelerator architectures, fabric options, and system vendor positioning.

Nice to have

• Master's degree or PhD in a relevant technical discipline.
• Experience supporting DOE national laboratory or comparable large-scale public-sector HPC procurements.
• Domain experience in a definitional HPC vertical, such as financial services, oil and gas, seismic processing, weather and climate, computational chemistry or biology, or manufacturing and CAE.
• Hands-on experience with scale-out fabrics including InfiniBand, Ultra Ethernet, or other high-performance interconnect.
• Experience with system-level performance modeling and workload characterization at scale.
• Experience with converged HPC and AI system design.

Skills: HPC system architecture, compute node design, scale-out fabric, CPU and GPU configuration, performance models