AI Networking Systems Engineer - High-Perf Fabric

Tensordyne; San Jose (CA); On-site; USD 170,000 - 210,000; Full time

Job summary

Tensordyne is seeking a System Software Engineer, Networking to design and optimize ultra-high-bandwidth multi-hop fabrics for AI inference at scale. You will own networking end-to-end—from topology through control- and data-plane, kernel drivers, and firmware—to squeeze latency, bandwidth, and reliability out of large, lossless networks.

This on-site role at our Sunnyvale, CA headquarters requires hands-on work across userspace services, SDKs, and hardware interfaces.

Qualifications

• Must have strong experience designing and operating data center/HPC networks.
• Hands-on familiarity with L2/L3 forwarding, ECMP, BGP, RoCE, QoS, PFC, and telemetry.
• Demonstrated ability to work across userspace, kernel, and firmware layers.

Responsibilities

• Architect ultra-high-bandwidth multi-hop fabrics for large-scale AI inference.
• Optimize lossless Ethernet/RDMA transport with congestion-control techniques.
• Diagnose bottlenecks across userspace, kernel, and firmware layers to improve performance.
• Drive topology, control-plane, and data-plane integration end-to-end.

Skills

Data center networking

L2/L3 routing

ECMP/BGP/RoCE

C/C++

Python

Go/Rust

Kernel networking

Telemetry/observability

Troubleshooting

Education

BS in Electrical Engineering or Computer Science

MS in a related field

Tools

SONiC

Broadcom SDK

SAI

OpenConfig gNMI

Job description

Tensordyne is seeking a System Software Engineer, Networking to design and optimize ultra-high-bandwidth multi-hop fabrics for AI inference at scale. You will own networking end-to-end—from topology through control- and data-plane, kernel drivers, and firmware—to squeeze latency, bandwidth, and reliability out of large, lossless networks.

This on-site role at our Sunnyvale, CA headquarters requires hands-on work across userspace services, SDKs, and hardware interfaces.