Network Engineer, High Performance GPU Network Direction - San Jose

Responsibilities

About the Team ByteDance Networking brings together innovative ideas and technologies from network architecture, software-defined networking (SDN), network virtualization, switch software and hardware co-design, and high-speed networking, to create hyperscale data-center networking solutions that power several of the most popular apps of the world such as Douyin and TikTok which serve hundreds of millions of users around the globe. ByteDance Networking is responsible for designing, building, and operating the global, intelligent network infrastructure to meet the requirements of high availability, scalability, and high-performance. By joining this team, you will gain marketable software development and/or network operation experience in data center networking on a massive scale.

• Responsible for the design, validation, implementation and operation of ByteDance's global high performance computing (HPC) networks.
• Work with cross-functional teams, including but not limited to machine learning (ML), compute and storage, driving the innovation and evolution of the HPC network.
• Work closely with external vendors to explore state-of-art architecture and next-gen technology.
• Build software and tools to improve the reliability and availability of HPC network infrastructure.
• Ensuring the reliability of ByteDance global network by participating in on-call rotation.

Qualifications

Minimum Qualifications

• Bachelor's in Computer Science, Information Science, Engineering, Mathematics, or a related field, or experience equivalent to a Bachelor's degree.
• 3 years of working experience and above.
• Expertise with HPC network topologies, like RDMA over Converged Ethernet (RoCE) or InfiniBand (IB).
• Good understanding of network protocols including TCP/IP, DHCP, BGP, OSPF/IS-IS and MPLS related technologies.
• Experience with building HPC networks.
• Be self-driven, possess good communication and written skills.

Preferred Qualifications

• Candidates with experience in high-performance computing (HPC) network topologies, particularly those familiar with Remote Direct Memory Access over Converged Ethernet (RoCE) or InfiniBand (IB), are preferred.