RTL Developer - AI ASIC Accelerator

RTL Designer / ASIC Design Engineer



We are looking for a
Senior ASIC Design Engineer to own the
end-to-end design of critical AI ASIC subsystems - from PLLs and compute clusters to interconnects and multi-die orchestration. The RTL you create will become production silicon, powering real-world AI workloads faster and more efficiently. In this role, every
microsecond, watt, and millimeter directly impacts AI economics at scale




Description
Responsibilities

• Lead the end-to-end design of complex ASIC subsystems, including specification, architectural exploration, IP evaluation/selection, integration, verification planning, and post-silicon validation.


• Drive architecture and micro-architecture trade-offs across features, performance, power, and area. Define clean interfaces and deliver production-quality RTL (Verilog/SystemVerilog).


• Implement and verify designs at both block and subsystem levels (using UVM/formal as needed); set coverage goals and drive sign-off.


• Define constraints and close timing: author SDC, run STA (setup/hold, OCV/derates), perform CDC/RDC analysis, and collaborate with PD on floorplanning and CTS.


• Perform simulation and emulation (FPGA/emulator), debug waveforms/logs, and correlate pre- and post-silicon behavior.


• Define and implement DFT/DFD strategies (scan, MBIST/LBIST, boundary scan/JTAG); support ATPG and test bring-up.


• Maintain high code quality with lint, CDC/RDC, LEC, and synthesis-friendly design practices; manage ECOs where necessary.


• Build automation flows: develop Tcl/PrimeTime scripts for skew analysis, clocking, and core-to-IO interfaces (PCIe, 400GE, UCIe).


• Write clear specifications, micro-architecture docs, and user guides for downstream engineering teams.
  Collaboration
  
  You will work closely with:
  
  
  
  
  • Silicon architects shaping next-generation AI compute platforms.
  
  
  • Verification engineers ensuring design robustness.
  
  
  • Firmware and Linux driver engineers building the software stack.
  
  
  • Security teams protecting customer assets.
  
  
  • Compiler and ML framework teams optimizing workload mapping.
  
  
  • System validation engineers ensuring enterprise-grade reliability.
    Minimum Qualifications
    
    
    
    
    • 7+ years of experience in RTL design and ASIC development.
    
    
    • Strong proficiency in Verilog/SystemVerilog.
    
    
    • Hands-on experience across the entire chip development lifecycle.
    
    
    • Proven expertise in complex IP integration (e.g., multi-core CPUs, NoCs, GPUs/NPUs, or high-speed interfaces like PCIe, 100/400G Ethernet, UCIe).
    
    
    • Experience writing specifications and converting them into robust designs.
    
    
    • Deep understanding of multiple clock domains and asynchronous interfaces.
    
    
    • Familiarity with AMBA bus protocols (AXI, AHB, APB).
    
    
    • Experience in power and clock management design.
    
    
    • Proficiency with formal equivalency checking (RTL ↔ Netlist).
    
    
    • Experience with DFT implementation.
      Preferred Qualifications (Nice-to-Haves)
      
      
      
      
      • Understanding of AI accelerator architectures (systolic arrays, dataflow, tiling).
      
      
      • Experience with memory bandwidth optimization and HBM/GDDR integration.
      
      
      • Exposure to first-silicon bring-up with real AI workloads.
      
      
      • Strong scripting skills (Tcl, Python) for automation and timing/debug flows.
      
      
    
    
  
  

• Deep RTL Mastery: 7+ years of building and shipping production-quality RTL (Verilog/SystemVerilog).


• Subsystem Ownership: Has taken IP or subsystems from spec → RTL → timing closure → silicon validation.


• IP Integration Experience: Worked with complex IPs like CPUs, NoCs, GPUs/NPUs, or high-speed I/O (PCIe, 400GE, UCIe).


• Timing & Debugging Skills: Knows how to close timing, write constraints (SDC), run STA, and debug across FPGA/emulation and silicon.


• Breadth Across the Flow: Comfortable with DFT/DFD, power/clock domains, CDC/RDC, and automation scripting.


• AI Accelerator Insight (bonus): Exposure to systolic arrays, dataflow architectures, and high-bandwidth memory subsystems (HBM/GDDR).