Clash
of Jobs
Browse All Jobs
Job Description
Google DeepMind is seeking a Software Engineer to contribute to the next generation of ML models on TPU. The engineer will co-design models and implement components across model architecture, ML frameworks, custom kernels, and platforms. The goal is to deliver frontier models with maximum efficiency.
The Software Engineer will redefine efficient training of frontier LLMs at a massive scale. This role offers the opportunity to influence the design of frontier LLM models and drive efforts to ensure efficient training and inference.
Responsibilities:
- Being responsible for Pre-Training efficiency and optimising the performance of the latest models on Google’s fleet of hardware accelerators.
- Being responsible for guiding model design to ensure inference-efficiency.
- Improving the performance of LLM models on hardware accelerators by optimizing at all levels, including developing custom kernels when necessary.
- Collaborating with the compiler, framework, and platform teams.
- Profile models to identify performance bottlenecks and opportunities for optimization.
- Develop low-level custom kernels for maximum performance of the most critical operators.
- Collaborating with research teams by enabling new critical operators in advance of their availability in frameworks and compilers.
Requirements:
- A proven track record of critical contributions to the distributed training of LLMs at 1e25 FLOPs scale on modern GPU/TPU clusters.
- Experience in programming hardware accelerators GPU/TPUs via ML frameworks (e.g. JAX, PyTorch) and low-level programming models (e.g. CUDA, OpenCL).
- Experience in leveraging custom kernels and compiler infrastructure to improve performance on hardware.
- Experience with Python and neural network training (publications, open-source projects, relevant work experience, etc.).
The role offers:
- Opportunity to influence the design of frontier LLM models.
- Opportunity to drive an effort to ensure efficient training and inference.
