Senior RTL Engineer (Interconnect Design)
ΠΡΡΡ & Π‘ΠΎΠΏΡΠΎΠ²ΠΎΠ΄
ΠΠ»Ρ ΠΌΡΡΡΠ° Ρ ΡΡΠΎΠΉ Π²Π°ΠΊΠ°Π½ΡΠΈΠ΅ΠΉ Π½ΡΠΆΠ΅Π½ Plus
ΠΠΏΠΈΡΠ°Π½ΠΈΠ΅ Π²Π°ΠΊΠ°Π½ΡΠΈΠΈ
TL;DR
Senior RTL Engineer (Interconnect Design): Own microarchitecture and RTL implementation of scalable on-chip and off-chip interconnect components for a custom AI accelerator platform with an accent on NoC/coherent fabric design, verification closure, and performance optimization. Focus on driving design from requirements through tape-out and silicon bring-up while providing technical leadership across architecture, verification, and physical design.
Location: San Francisco, CA (hybrid: 3 days in the office per week). Relocation assistance offered.
Salary: $225Kβ$445K + offers equity
Company
builds AI research and deployment systems, including next-generation hardware for large-scale training and inference.
What you will do
- Own microarchitecture, RTL design, and delivery of major SoC interconnect components (NoC fabrics, switches/routers/bridges, protocol adapters, arbiters, traffic-management logic, and off-chip interfaces).
- Perform substantial direct microarchitecture and RTL coding work for production-quality, parameterized SystemVerilog/Verilog.
- Drive third-party engagements to develop novel networking/interface protocols and silicon IP while maintaining design integrity.
- Analyze traffic patterns and optimize interconnect behavior under realistic workloads with performance and architecture teams.
- Collaborate on verification strategy (coverage plans, assertions, stress scenarios, and debug approaches) for highly concurrent fabric behavior.
- Partner with physical design to ensure implementability at target frequency/power/area (floorplan-aware design, pipeline strategy, timing closure, congestion management) and lead design reviews/mentoring.
Requirements
- Extensive experience designing and delivering complex SoC interconnect solutions (NoC, coherent fabric, memory subsystem, cache-coherent, or chip-level integration).
- Proven track record owning major RTL blocks or SoC subsystems from microarchitecture through tape-out and silicon bring-up.
- Deep expertise in Verilog/SystemVerilog and building clean, parameterized, production-quality RTL.
- Strong interconnect fundamentals: topology/routing, arbitration, virtual channels, flow control, buffering, ordering, QoS, coherency, deadlock avoidance, congestion management, and performance monitoring.
- Experience with common on-chip/chip-to-chip protocols and interfaces such as AXI, APB, CXL, PCIe, and Ethernet.
- Experience across RTL-adjacent signoff flows (lint, CDC/RDC, synthesis, formal verification, static timing analysis, power analysis, and design-for-test).
Nice to have
- Experience designing interconnect for AI accelerators/GPUs/CPUs, HPC systems, networking silicon, or large-scale datacenter silicon.
- Experience with memory consistency, virtualization/isolation, RAS, telemetry, or security requirements in complex SoCs.
- Experience with NoC performance modeling, traffic simulation/emulation, FPGA prototyping, or post-silicon performance analysis.
- Experience leading architecture or RTL delivery for first-generation silicon programs or rapidly evolving hardware platforms.
Culture & Benefits
- Hybrid work model: 3 days in the office per week.
- Relocation assistance for new employees.
- Senior, hands-on role with broad technical ownership across the silicon lifecycle.
- Mentoring and technical leadership opportunities across multiple engineering teams and external partners.
Hiring process
- Interviews focused on interconnect/RTL ownership, verification and signoff experience, and cross-team execution.
- Technical evaluation of microarchitecture/RTL coding depth and ability to drive performance and verification closure.
ΠΡΠ΄ΡΡΠ΅ ΠΎΡΡΠΎΡΠΎΠΆΠ½Ρ: Π΅ΡΠ»ΠΈ ΡΠ°Π±ΠΎΡΠΎΠ΄Π°ΡΠ΅Π»Ρ ΠΏΡΠΎΡΠΈΡ Π²ΠΎΠΉΡΠΈ Π² ΠΈΡ ΡΠΈΡΡΠ΅ΠΌΡ, ΠΈΡΠΏΠΎΠ»ΡΠ·ΡΡ iCloud/Google, ΠΏΡΠΈΡΠ»Π°ΡΡ ΠΊΠΎΠ΄/ΠΏΠ°ΡΠΎΠ»Ρ, Π·Π°ΠΏΡΡΡΠΈΡΡ ΠΊΠΎΠ΄/ΠΠ, Π½Π΅ Π΄Π΅Π»Π°ΠΉΡΠ΅ ΡΡΠΎΠ³ΠΎ - ΡΡΠΎ ΠΌΠΎΡΠ΅Π½Π½ΠΈΠΊΠΈ. ΠΠ±ΡΠ·Π°ΡΠ΅Π»ΡΠ½ΠΎ ΠΆΠΌΠΈΡΠ΅ "ΠΠΎΠΆΠ°Π»ΠΎΠ²Π°ΡΡΡΡ" ΠΈΠ»ΠΈ ΠΏΠΈΡΠΈΡΠ΅ Π² ΠΏΠΎΠ΄Π΄Π΅ΡΠΆΠΊΡ. ΠΠΎΠ΄ΡΠΎΠ±Π½Π΅Π΅ Π² Π³Π°ΠΉΠ΄Π΅ β