AI Accelerator Physical Design Engineer - Full Time
Posted Updated
Description
What you'll do
- Own physical implementation of accelerator blocks and top level: synthesis, floorplanning, power planning, placement, clock tree synthesis, and routing
- Develop floorplans for repeated compute tiles, large SRAM arrays, and high-bandwidth data movement, balancing congestion, wire length, and utilization
- Close timing, power, and area (PPA) targets, including optimization of datapath-heavy and high-fanout structures
- Manage power delivery and density: power grid design, IR-drop and electromigration analysis, and thermal-aware implementation
- Implement low-power techniques such as clock gating, power gating, multi-Vt, and multi-voltage domains
- Run static timing analysis, signal-integrity, and physical verification (DRC/LVS) signoff across corners and modes
- Partition and implement hierarchical designs, including budgeting and integration of replicated blocks
- Work with architecture, RTL, DFT, and verification teams to co-optimize design for physical implementability
- Develop and maintain automation for flows, regressions, and QoR tracking
- Support silicon bring-up and correlate post-silicon results with pre-silicon predictions
What we're looking for
- Bachelor's or Master's degree in Electrical Engineering, Computer Engineering, or a related field
- 5+ years of industry experience in high performance, low power physical design
- Hands-on experience taking complex digital designs through synthesis, place-and-route, and signoff at advanced process nodes
- Strong understanding of timing closure, clocking, and static timing analysis, including OCV and multi-corner multi-mode analysis
- Experience with power integrity: power grid design, IR drop, and electromigration
- Experience with physical verification (DRC/LVS) and fixing violations
- Familiarity with accelerator or compute-array architectures, such as systolic arrays, SIMD/vector engines, and large on-chip memories
- Proficiency with industry-standard EDA tools (see Tools & Technologies)
- Scripting skills in Tcl, Python, or Perl for flow automation
- Strong debugging skills and the ability to work with RTL, architecture, and DFT teams