Analog Devices (ADI) is seeking an AI Compute System Engineer to benchmark and validate novel analog and in-memory compute primitives for Edge AI, aiming to integrate these innovative compute IPs into ADI's future SoCs and drive advancements in digitized factories, mobility, and digital healthcare.
Requirements
- Good understanding of SoC architecture, IP integration, and co-simulation with mixed-signal blocks.
- Familiarity with analog compute, in-memory compute, or
- Experience with building test benches, performance measurement, and system validation.
- Hands-on skills in scripting and modeling (e.g., Python, MATLAB, or SystemC).
- Strong analytical skills to evaluate performance, power, and reliability metrics.
- Experience in SoC system engineering, integration and validation, compute benchmarking.
- Big Advantage – Experience in AIML development, Neuromorphic architectures.
Responsibilities
- Benchmark and validate analog and in-memory compute primitives for Edge AI applications.
- Lead the evaluation and validation of new compute approaches in collaboration with startups, internal teams, and technology partners.
- Analyze and benchmark third-party analog and in-memory compute IPs to assess suitability for real-world AI inference tasks.
- Evaluate performance, power efficiency, and reliability to inform architectural and product decisions.
- Collaborate with the SoC System Architect to validate system integration requirements and constraints.
- Build and test early-stage proof-of-concept (PoC) systems combining ADI IP with novel compute technologies.
- Build testbed setup, driver bring-up, and co-simulation with analog and digital subsystems.
Other
- Master's degree in electrical engineering, Computer Engineering, Computer Science, or a related field.
- 5-7 years of experience in SoC system engineering, integration and validation, compute benchmarking.
- Ability to collaborate cross-functionally with architects, hardware teams, and external partners.
- Clear technical communication and documentation skills.
- Required Travel: Yes, 10% of the time
