About the job
SambaNova Systems is looking for a talented and motivated Network Architect to help us solve the most challenging and impactful problems in machine learning and artificial intelligence.
Responsibilities
Conduct research and development in state of the art networking solutions
Define and own networking architecture specifications of performant, robust and efficient communications for the next generation super-pod scale ML HW
Define and own ML networking roadmap, leading a cross functional effort of software, systems and silicon
Collaborate with executive leadership to align network initiatives against product competitiveness requirements
Stay on top of and contribute to industry standards representing SambaNova
Qualifications
Minimum
BS, MS, or PhD degree in Computer Science, Computer Engineering or Electrical Engineering
10+ years of industry experience in ML / supercomputer / public cloud / networking
Proven track record of successful full-cycle development and delivery from planning to product in networking / ML scale up processors and systems
Track record of ownership and definition of various networking and routing protocols
Deep understanding and hands-on experience defining various networking topologies for large machine clusters.
Deep understanding of congestion management, QoS, security isolation
Deep understanding of reliability, availability and serviceability from networking perspective
Capable of abstracting and analyzing network performance from real world application
Capable of breaking down networking requirements into silicon, packaging, system SW, RunTime SW and Compiler SW requirements
Strong hands-on analytical and debug skills to make progress on challenging problems in an ambiguous and highly dynamic environment
Strong coding skills with C++/Python/etc
Highly motivated, self-driven, independent, organized and strong collaboration and communication skills
Cross-functional team leadership experience
Preferred
15+ years of industry experience in ML / supercomputer / public cloud / networking industries
Good understanding of ASIC development cycles
Knowledge of machine learning architecture and/or computer architecture
Experience in building or proficiently using network simulators
Experience in leading silicon, packaging, system HW, system SW cross-functional teams
Experience in conducting perf and perf/tco trade-off analysis on different interconnect choices
Familiarity with off the shelf optical IP and optical switches
Understanding / hands-on experience in end to end optical systems
Hands-on experience in working with IP/Switch hardware vendor to enhance / customize solutions