Distributed Systems Engineer, Multi-Chip Simulation (Mid/Senior/Principal)
Core
Design and develop advanced simulation infrastructure enabling distributed AI systems to operate across multiple interconnected chips before physical hardware is available.
Role type
Senior IC distributed systems engineer (multi-chip simulation)
Builds
Multi-chip simulation environments and tools for distributed AI inference/training
Domain
AI hardware development, distributed systems, chip architecture
Deliverable
production ML models
Required skills
Rust, C/C++, distributed systems architecture, simulation methodologies, debugging timing-dependent issues, scalable software architecture, continuous integration
Preferred skills
Multi-GPU/multi-accelerator inference, NCCL-style collective communication, hardware emulation platforms (Cadence Palladium, Synopsys ZeBu), FPGA prototyping
Technologies
Rust, C/C++, Python, NCCL, Cadence Palladium, Synopsys ZeBu
Responsibilities
Design and maintain multi-chip simulation environments covering host-to-chip and chip-to-chip communication layers; Build simulators that accurately reproduce interactions between multiple accelerators and host systems; Model complex communication fabrics including transfer timing, collective operations, and back-pressure; Calibrate communication models against real hardware measurements; Own the overall software architecture of the multi-chip simulation stack; Integrate simulators with different levels of fidelity into a unified framework; Scale simulation infrastructure from small configurations to full rack-level systems; Identify and resolve complex distributed systems issues like race conditions and deadlocks
Seniority
Mid/Senior/Principal, hands-on IC