AI SoC Networking Architect: Scale-Out & RDMA Expert

Socket.dev; Los Altos (CA); On-site; USD 180,000 - 260,000; Full time

Job summary

Majestic Labs seeks a highly technical SOC System Network Architect for AI, leading the definition and design of network engines and interconnect fabrics powering our custom AI servers.

You will work with silicon, hardware, firmware, and software teams to optimize scale-up and scale-out networks for distributed AI workloads, including RDMA, NIC engines, and advanced transport protocols.

Qualifications

• 7+ years in hands-on network architecture, SoC interconnects, or silicon engineering.
• B.Sc. or M.Sc. in Electrical Engineering, Computer Engineering, or Computer Science.
• Proven track record delivering complex network ASICs, NIC engines, or switch architectures.

Responsibilities

• Architect end-to-end SoC networking engines, NIC architectures, and switch IPs for AI accelerators.
• Define scale-up and scale-out network topologies for AI workloads.
• Design hardware acceleration and offload engines for AI collectives on the network fabric.
• Optimize transport protocols and congestion control for synchronized AI traffic flows.
• Drive performance modeling, cycle-accurate simulations, and traffic analysis.
• Collaborate across ASIC design, silicon architecture, firmware, and software teams.
• Produce network architecture specs and lead architecture reviews.

Skills

SoC networking

Network architecture

RDMA (RoCEv2)

SmartNIC/NIC

TCP/UDP/IP

Performance modeling

Hardware/software interfaces

Education

B.Sc./M.Sc. in Electrical/Computer Engineering/CS

Tools

ns-3

OMNeT++

SystemC/TLM

Verilog

Job description

Majestic Labs seeks a highly technical SOC System Network Architect for AI, leading the definition and design of network engines and interconnect fabrics powering our custom AI servers.

You will work with silicon, hardware, firmware, and software teams to optimize scale-up and scale-out networks for distributed AI workloads, including RDMA, NIC engines, and advanced transport protocols.