Tenstorrent
AI/ML Physical Design Flow Engineer
$100K–$500K
Austin, Texas, United States +3 · Full-time · Posted Apr 3, 2026
- Location
- Austin, Texas, United States
Fort Collins, Colorado, United States
Santa Clara, California, United States
Austin, Texas, United States - Experience
- 5+ years; candidates at various experience levels are welcome and assessed for an appropriate level.
- Education
- BS in Electrical or Computer Engineering (or equivalent experience).
- Visa sponsorship
- Employment is contingent on eligibility to access U.S. export-controlled technology; citizenship or permanent-residency status or prior license approval may be required for certain country nationals.
About the job
Develops and deploys physical design methodologies and AI/ML-driven tools to improve PPA, runtime, and engineering efficiency across advanced-node RTL-to-GDS flows.
What you'll do
- Lead cross-functional physical design efforts across IPs, projects, and advanced technology nodes.
- Develop RTL-to-GDS methodologies covering floorplanning, synthesis, P&R, STA, signoff, and assembly.
- Deploy AI/ML solutions and optimize EDA tools and custom CAD flows using data-driven techniques.
What they ask for
- BS in Electrical or Computer Engineering (or equivalent experience).
- 5+ years in Physical Design CAD methodology at advanced nodes.
- Track record improving PPA and/or runtime on high-performance, low-power taped-out designs.
- Hands-on experience with industry-standard EDA tools across synthesis, P&R, STA, signoff, and hierarchical flows; strong Python/Tcl and data skills.
Nice to have
- Interest or experience in ML frameworks, including PyTorch or TensorFlow.
Pay
$100k - $500k (total pay, including base and variable compensation targets)
Tools
Fusion Compiler, Python, Tcl, PyTorch, TensorFlow
AI focus
Architects and deploys AI/ML-driven solutions in production physical design flows and uses data-driven and ML-based methods to improve efficiency, turnaround time, QoR, PPA, and runtime.
Model-written summary · the original posting is authoritative