Computer Systems Architect — AI Infrastructure

About the role

We are looking for a Computer Systems Architect to help design next-generation AI computing infrastructure across node, supernode, cluster, memory, interconnect, and storage layers. You will use hardware-software co-design and workload-driven analysis to improve performance, scalability, reliability, and cost efficiency.

We welcome specialists with deep expertise in one or more areas, including cluster and supernode architecture, CPU architecture, interconnects, memory and storage systems, in-memory or near-memory computing, and emerging media such as High Bandwidth Flash (HBF).

What you will do

• Design end-to-end architectures and topologies for large-scale compute clusters, supernodes, high-performance storage systems, or CPU-based platforms.
• Architect and optimize chip-to-chip and node-to-node interconnects using technologies such as NVLink, CXL, RDMA, InfiniBand, and advanced fabrics.
• Develop hardware-software co-design strategies for heterogeneous memory and storage, including memory pooling and tiering, in-memory or near-memory computing, and HBF integration.
• Build architecture models, simulators, and workload-driven performance studies to identify bottlenecks across compute, memory, communication, and I/O.
• For CPU-focused projects, design or optimize core pipelines, cache and memory hierarchies, execution models, and ISA/microarchitecture interactions.

What we are looking for

• Master’s degree or Ph.D. in Computer Architecture, Electrical Engineering, Computer Science, High-Performance Computing, or a related field.
• Representative work in at least one primary area: cluster/supernode architecture, CPU architecture, interconnects, memory/storage architecture, in-memory computing, or emerging memory and storage media.
• Strong understanding of computer architecture and the ability to reason across hardware and system-software boundaries.
• Experience with architecture modeling or simulation, workload characterization, profiling, and
• bottleneck analysis.
• Proficiency in C/C++, Linux, and low-level system or performance engineering.
• Evidence of measurable impact in throughput, latency, bandwidth utilization, scalability, energy efficiency, reliability, or cost.

Preferred qualifications

• Experience taking an architecture from research or simulation into a prototype, product, or large-scale deployment.
• Publications, patents, standards contributions, or substantial open-source work in a relevant area.