What Youll Be Doing:
Investigating emerging technologies and methodologies in ML and AI to discern their interactions with network infrastructure.
Executing workloads on AI systems, conducting profiling, and analyzing bottlenecks and possible enhancements.
Conducting research and implementing optimizations for communication libraries like NCCL and UCX.
Spearheading the conceptualization of next-generation networking products tailored to support and accelerate state-of-the-art ML workloads.
Develop models for simulations, analyze simulation results, and develop optimization algorithms.
Collaborate with multi-functional teams, including other architecture teams, logic design, system software, firmware, and ML research teams, to ensure the successful execution of the project.
What We Need To See:
M.Sc, or Ph. D degree in Computer Science, Computer Engineering, or Electrical Engineering.
At least 2+ years of industry or research experience in computer networks.
Extensive expertise in ML/AI workloads, particularly in distributed training.
Excellent understanding of large-scale network behavior and the effect of distributed computing workloads on the network.
Experience in the development of simulation environments.
Great problem-solving and critical-thinking skills.
Ability to thrive in a fast-paced and dynamic environment is necessary.
Ability to work concurrently with multiple groups in the organization.
Ways To Stand Out Of The Crowd:
Knowledge of communication libraries such as NCCL, UCX, and UCC.
Good knowledge of network protocols – such as InfiniBand, IP, TCP, RoCE, and network topologies.
Experience with Python, C++, and dockers.
Expertise in system engineering, operations research, and intricate hardware-software integrated systems.
Demonstrated experience in DLRM, LLM or other generative AI.
