Principal Architect, Performance Analysis and Modeling
Core
Build analytical performance models and simulation tools to project behavior of emerging ML workloads on current and future AI inference accelerator hardware.
Role type
Principal Software Engineer, Performance Analysis and Modeling
Builds
Analytical performance models and architecture simulators for d-Matrix silicon
Domain
AI hardware acceleration, computer architecture, HW/SW co-design
Deliverable
production ML models
Required skills
Computer architecture, HW/SW co-design, performance modeling, DNN fundamentals, C/C++ or Python programming, building analytical performance models, architecture simulators
Preferred skills
Optimizing AI/ML workloads on accelerators, research in AI/ML architecture/microarchitecture
Technologies
C, C++, Python, DNNs
Responsibilities
Analyze emerging ML workloads (multi-modal LLMs, CoT reasoning, video/audio generation) to identify performance-relevant properties; Build and maintain analytical performance models for current and future hardware generations; Develop and extend architecture simulators for proposed HW/SW features; Partner with hardware design, compiler, and kernel teams to validate modeling assumptions; Track ML architecture research and incorporate findings into modeling; Propose targeted HW/SW feature improvements based on modeling results; Document modeling methodology and findings.
Seniority
Principal, hands-on IC
