Principal Software Architect- High Performance Computing
Core
Design and implement robust, scalable HPC infrastructure combining CPUs, GPUs, and FPGAs for low-latency image processing and deep-learning workloads in chip manufacturing equipment.
Role type
Principal Software Architect (HPC & Heterogeneous Computing)
Builds
High-performance computing solutions for semiconductor process control equipment
Domain
Semiconductor manufacturing / High-Performance Computing
Deliverable
production ML models | infrastructure
Required skills
HPC system design, heterogeneous computing (CPU/GPU/FPGA), CUDA programming, parallel programming (MPI, OpenMP, OpenACC), distributed computing, low-level code optimization, performance profiling, capacity planning
Preferred skills
GPU inference server experience (Nvidia Triton), microservices architecture, containerization (Docker/Singularity), RDMA/RoCE/InfiniBand, job-scheduling software (SLURM, Torque, LSF), DAG execution frameworks
Technologies
CUDA, MPI, UCX, OpenMP, OpenACC, OpenCL, C/C++, Docker, Singularity, NVIDIA DCGM, vtune, Oprofiler, perf, NVIDIA Nsight, RDMA, RoCE, InfiniBand, SLURM, Torque, LSF, Intel TBB
Responsibilities
Analyze and partition workloads to appropriate compute units; code and develop prototypes to validate designs; profile clusters and nodes to identify bottlenecks; conduct performance tuning and capacity planning; evaluate and recommend technologies; lead design of complex software components; ensure system scalability, reliability, and cost-effectiveness
Seniority
Principal, hands-on IC with mentorship
