Required Skills & Experience
- 7+ years of relevant engineering experience with a strong background in semiconductor hardware validation, post-silicon validation, chip-level validation, compute validation, accelerator validation, or closely related work.
- Hands-on experience validating and characterizing semiconductor or compute hardware at the chip, board, or system level.
- Experience benchmarking compute platforms and analyzing performance, power, memory, or other system-level characteristics.
- Ability to independently bring up, configure, test, troubleshoot, and characterize hardware platforms and evaluation boards.
- Hands-on familiarity with at least one major machine learning framework such as PyTorch or TensorFlow.
- Experience with AI model deployment or runtime environments such as ONNX, ONNX Runtime, TensorFlow Lite, or comparable tools.
- Ability to configure and troubleshoot AI software frameworks sufficiently to deploy workloads and obtain reliable hardware-performance measurements.
- Understanding of the interaction between hardware, software, firmware, memory, compute, and AI workloads.
Nice to Have Skills & Experience
- Master's degree in Electrical Engineering, Computer Engineering, Computer Science, or a related technical discipline.
- Experience specifically validating AI accelerators, NPUs, GPUs, or other specialized compute architectures.
- Experience with edge AI or power-constrained embedded computing environments.
- Exposure to emerging compute technologies such as analog or digital in-memory computing or neuromorphic architectures.
- Experience with embedded AI platforms such as NVIDIA Jetson or comparable evaluation platforms.
- Experience integrating multiple hardware evaluation boards, sensors, firmware, software, and AI workloads into proof-of-concept systems.
- Experience working directly with semiconductor vendors, compute startups, or technology partners during hardware evaluation and characterization
Job Description
We are seeking a Senior AI Accelerator Performance Validation Engineer to support a team evaluating emerging, energy-efficient compute technologies for edge AI applications. This is a hands-on hardware performance validation role focused on evaluating semiconductor and AI accelerator platforms, validating vendor performance claims, identifying system bottlenecks, and determining how effectively different compute architectures perform against real-world workloads. The ideal candidate has a strong background in semiconductor, chip-level, post-silicon, compute, or accelerator validation and is comfortable independently bringing up hardware, executing benchmarks, troubleshooting performance issues, and analyzing trade-offs across performance, power, and memory. This is not an AI research or model-development position. Candidates should have enough hands-on experience with machine learning frameworks and deployment tools to configure, run, debug, and benchmark AI workloads on different compute platforms.
Key Responsibilities
- Perform hands-on performance validation and characterization of emerging AI accelerator and compute architectures.
- Evaluate vendor-supplied semiconductor hardware and compute evaluation platforms, including working with evaluation boards and PCBA-based systems.
- Bring up hardware and software environments necessary to execute and benchmark AI workloads.
- Validate hardware specifications and vendor performance claims against independently collected results.
- Benchmark compute platforms across key performance indicators including processing performance, power consumption, memory utilization, and overall system efficiency.
- Investigate discrepancies between expected and observed performance and identify potential hardware, software, framework, memory, or system-level bottlenecks.
- Compile, configure, and run AI workloads using frameworks and deployment environments such as PyTorch, TensorFlow, ONNX, and TensorFlow Lite.
- Troubleshoot hardware/software integration issues sufficiently to ensure workloads are running correctly and performance measurements are valid.
- Evaluate and compare traditional and emerging compute architectures, which may include GPUs, NPUs, embedded AI platforms, in-memory compute architectures, and neuromorphic computing technologies.
- Support system-level experimentation involving compute hardware, sensors, firmware, software, and AI workloads.
- Work directly with compute vendors and technology partners to understand hardware capabilities, investigate performance findings, and validate technical conclusions.
- Analyze benchmark results and translate findings into clear technical recommendations.
- Document test methodologies, performance data, technical conclusions, and recommendations for engineering and program leadership.
- Contribute to technical go/no-go decisions regarding emerging compute technologies and potential future system architectures.
Pay ranges from 50-60/hr based on qualifications and years of experience.