Network Infrastructure Architect - San Jose, California, United States
$250,000 - $330,000 USD yearly
Originally posted 25 June 2026 by the employer.
About the role
The architecture directly supports platforms incorporating SambaNova's Reconfigurable Dataflow Unit (RDU) and SambaRack systems.
What you'll do
- Architect network fabrics that power AI clusters, owning fabric, interconnect, and spine-leaf design decisions across compute platforms. - Partner with hardware engineering, supply chain, and customer-facing teams to translate workload requirements into platform roadmaps. - Maintain a vendor-agnostic approach across switch, NIC, optical, and fabric ecosystems to prevent vendor lock-in. - Define architecture trade-offs and drive platforms from concept through operational deployment.
What you'll need
- 12+ years of experience designing or architecting infrastructure for hyperscale, AI, HPC, or large-scale data center environments. - Deep working knowledge of AI cluster networking, including InfiniBand, RoCEv2, Ethernet-based AI fabrics, 400G/800G interconnects, and GPU-to-GPU east-west traffic patterns. - Experience with lossless Ethernet designs, including QoS, ECN, PFC, congestion management, buffer tuning, and failure-domain isolation. - Strong understanding of spine-leaf architectures and associated technologies, including VXLAN, EVPN, BGP, ECMP, underlay/overlay design, and network automation. - Familiarity with high-speed optics and cabling for AI clusters, including 400G/800G transceivers, DAC/AOC, fiber topology, and link budgets. - Proven experience translating customer and workload requirements into product requirements, platform roadmaps, and architecture tradeoffs.
Nice to have
- Experience with AI compute platforms including GPU/accelerator systems, rack-scale compute, NVLink/NVSwitch-class fabrics, PCIe/CXL, and DPUs/IPUs. - Understanding of rack-level power, thermal, mechanical, and serviceability requirements, and how network choices impact NIC selection, cabling density, and PCIe lane allocation. - Background in large-scale fleet operations, covering lifecycle management, reliability, observability, telemetry, and field issue resolution.
Skills: hyperscale AI compute platform, AI clusters, GPU/accelerator systems, InfiniBand, Ethernet-based AI fabrics
This role has been open 119 days — well above the 51-day median for Hardware Systems Engineering roles.
Hardware Systems Engineering · AI ML Hardware
|
Open roles in category
1003
|
Median days open
51 d
|
Median salary
$155k
|
See the full market breakdown ▾Category comparison, skills in demand, and who else is hiring
| Metric | SambaNova | All employers we track in this specialty (1003 roles · 104 employers) |
|---|---|---|
| Open roles in this specialty | 4 | 1003 |
| Open roles in the wider Software, Firmware & Systems family | 26 | 6298 · 145 employers |
| Median days open | 65 d | 51 d (+14 d vs this employer) |
| Median salary (USD postings) | — | $155k |
Skills observed across this category: hyperscale AI compute platform, AI clusters, GPU/accelerator systems, InfiniBand, Ethernet-based AI fabrics
Who's hiring in this category
- Intel Corporation · 72 open roles · median 34 d
- KLA Corporation · 63 open roles · median 48 d
- Teradyne · 61 open roles · median 34 d
- NVIDIA · 53 open roles · median 72 d
- Applied Materials · 51 open roles · median 70 d
- AMD · 46 open roles · median 52 d
How we counted: 1003 open Hardware Systems Engineering (AI ML Hardware) roles from 104 employers tracked in the SemiconductorJobs index, counted 22 Oct 2026. Specialty figures count only roles carrying this exact specialty label, so an employer's related work in neighbouring specialties is not included there — it is counted in the wider Software, Firmware & Systems family row. Figures refresh nightly.