System Engineer Intern - Efficient On-Device ML Computing
Meta

Job Info

Meta Reality Labs (RL) is a pioneer in the field of Augmented Reality (AR) and Virtual Reality (VR) devices and experiences. Artificial Intelligence (AI) and on-device Machine Learning (ML) have been instrumental in driving this innovation. The Wearable System Architecture team within RL is dedicated to enhancing the power and performance of on-device ML execution through the development of custom ML accelerators, optimization of ML models, and the implementation of innovative system architectures. We are looking for skilled interns with experience in developing, profiling, and optimizing ML models for edge devices running on RTOS and AOSP. The ideal candidate will have a strong background in computing architecture, with a focus on ML accelerators and parallel computing. In this role, you will be exposed to end-to-end use case analysis and optimization, from UI to software/firmware frameworks, ML models, and underlying hardware blocks. Through detailed profiling and analysis, you will contribute to the optimization of ML models and the development of next-generation ML accelerators and Wearable system architecture.Our internships are twelve (12) to sixteen (16) weeks long.

System Engineer Intern - Efficient On-Device ML Computing Responsibilities:

• Perform in-depth power and performance profiling of ML models and ML benchmarks on ML accelerators.
• Examine the power and performance characteristics of ML accelerators in relation to various types of ML models.
• Develop an optimal mapping definition for ML models to ML accelerators.
• Identify power and performance bottlenecks and optimization opportunities in ML models, ML accelerators, and system architectures.
• Collaborate with cross-functional teams to prototype and productize optimizations.
• Conduct power and performance analysis of end-to-end AI powered use cases, identify power optimization opportunities in software, firmware and overall system architecture.
• Work alongside system architects to create a roadmap for the next generation of ML accelerators and wearable system architecture.

Minimum Qualifications:

• Currently has, or is in the process of obtaining a Master's degree in Computer Science, or Computer Engineering with a focus on ML.
• Proficient in ML frameworks such as PyTorch or TensorFlow.
• Familiarity with edge ML frameworks like TensorFlow Lite or similar technologies.
• Experienced with edge ML accelerator compiler toolchains, including ARM Vela or others.
• Experience in embedded software development using C/C++.
• Strong understanding of computer architecture.
• Must obtain work authorization in country of employment at the time of hire, and maintain ongoing work authorization during employment.

Preferred Qualifications:

• Currently holds or is pursuing a PhD in Computer Science or Computer Engineering with a focus on ML.
• Experience in developing and optimizing ML models for edge devices.
• Familiarity with ML accelerators and their internal architecture.
• Knowledge of ML model optimization techniques, such as quantization and pruning.
• Experience with profiling ML model execution on and off-device.
• Familiarity with power optimization techniques, such as DVFS, power and clock gating.
• Intent to return to degree-program after the completion of the internship.

About Meta: