Seeking a skilled AI Scale-Out Software Engineer to build and optimize our clients scale-out fabric (TT-fabric) for distributed inference and training infrastructure. The ideal candidate will have expertise in deep learning, distributed systems, and low-level networking.

Responsibilities

• Design, develop, and maintain TT-fabric, a low-level networking library for AI processors built on top of Ethernet protocol
• Design and implement efficient distributed training systems for large-scale deep learning models
• Optimize network communication for multi-node AI processor clusters
• Tune system performance for inference and training of key AI models
• Work in the Metalium team and integrate scale-out APIs into the Programming Model
• Work with AI model builder and researchers to improve both the scale out infrastructure and as well as model design

Experience & Qualifications

• Bachelor’s or Master’s degree in Computer Science, Electrical Engineering, or a related field.
• Proven experience in low-level software development.
• Strong proficiency in programming languages such as C / C++.
• Experience with MPI or similar distributed computing frameworks
• Experience with low-level networking libraries (e.g., libfabric, libibverbs)
• Knowledge of networking protocols, especially Ethernet and InfiniBand
• Knowledge of high-performance interconnects
• Familiarity with RDMA programming
• Familiarity with large-scale deep learning frameworks (e.g., PyTorch, TensorFlow)
• Familiarity with network offload engines and SmartNICs
• Strong communication skills and the ability to work effectively with cross-functional teams.
• Passion for technology and a commitment to pushing the boundaries of what is possible in AI.

Estimated Min Rate: $185000.00
Estimated Max Rate: $250000.00
Note: Any pay ranges displayed are estimations. Actual pay is determined by an applicant’s experience, technical expertise, and other qualifications as listed in the job description. All qualified applicants are welcome to apply.
Yoh, a Day & Zimmermann company, is an Equal Opportunity Employer. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, disability, or status as a protected veteran.
Visit https://www.yoh.com/applicants-with-disabilities [1]to contact us if you are an individual with a disability and require accommodation in the application process.
For California applicants, qualified applicants with arrest or conviction records will be considered for employment in accordance with the Los Angeles County Fair Chance Ordinance for Employers and the California Fair Chance Act. All of the material job duties described in this posting are job duties for which a criminal history may have a direct, adverse, and negative relationship potentially resulting in the withdrawal of a conditional offer of employment.

• Design, develop, and maintain TT-fabric, a low-level networking library for AI processors built on top of Ethernet protocol
• Design and implement efficient distributed training systems for large-scale deep learning models
• Optimize network communication for multi-node AI processor clusters
• Tune system performance for inference and training of key AI models
• Work in the Metalium team and integrate scale-out APIs into the Programming Model
• Work with AI model builder and researchers to improve both the scale out infrastructure and as well as model desig