About the Role
You will be a member of the MTIA (Meta Training & Inference Accelerator) Software team, contributing to a comprehensive software platform strategy for training and serving deep learning models. This role involves working on core areas like PyTorch & GenAI framework components, AI compiler and runtime, or high-performance deep learning kernels to accelerate machine learning workloads on MTIA hardware. You will collaborate with AI researchers and hardware design teams to optimize models and develop performance enhancements, applying software development best practices and gaining experience in machine learning compiler frameworks and hardware-software codesign.
Responsibilities
- Develop SW stack with a focus on AI frameworks, compiler stack, high performance deep learning kernel development, runtime, or tooling for MTIA accelerators.
- Contribute to the development of PyTorch, core compilers, and GenAI OSS frameworks to support AI hardware accelerators and optimize their performance.
- Analyze deep learning models and develop/implement optimization algorithms for high performance and efficiency.
- Collaborate with AI research scientists to accelerate next-generation deep learning models like recommendation systems and Generative AI.
Requirements
- Proven C/C++ programming skills.
- Experience in AI framework development or accelerating deep learning models on hardware architectures.
- Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience.
Skills
- PyTorch
- GenAI Frameworks
- AI compiler
- Runtime development
- High performance deep learning kernels
- Tooling for device profiling, tracing, debugging and performance tuning
- CUDA programming
- OpenMP
- OpenCL
- cuBLAS
- cuDNN
- CUTLASS
- HIP
- ROCm
- MLIR
- LLVM
- IREE
- XLA
- TVM
- Halide
- TensorFlow
- ONNX
- TensorRT
- vLLM
- SGLang
- SIMD
- GPU architectures
- CPU architectures
- AI hardware accelerator architectures
Experience Level
- 7+ years of experience in AI framework development or accelerating deep learning models on hardware architectures (with Bachelor's degree)
- 4+ years of experience in AI framework development or accelerating deep learning models on hardware architectures (with Master's degree)
- 3+ years of experience in AI framework development or accelerating deep learning models on hardware architectures (with PhD)
Education Level
- Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience
- Master's degree in Computer Science, Computer Engineering, relevant technical field
- PhD in Computer Science, Computer Engineering, or relevant technical field
Salary/Compensations
- CA$160,000/year to CA$213,000/year
Benefits
- bonus
- equity
- benefits
About the Company
- MTIA (Meta Training & Inference Accelerator) Software team develops a comprehensive software platform strategy.
- The AI and Compute Foundations team focuses on specialized hardware architectures for AI.
