Senior Staff Physical Design Engineer – Contractor/Freelance (FinFET Specialist)
Le poste
About Us
Vertical Compute is an early-stage deep tech startup dedicated to pioneering next-generation memory technologies for advanced computing architecture. Our mission is to redefine the well-known trade-offs of semiconductor memory devices, ultimately enabling the future of computing.
We are welcoming passionate, experienced, and forward-thinking colleagues to join our dynamic team and disrupt the industry together.
About What You Will Do
As a Senior Staff Physical Design Engineer, you will take technical leadership in the physical implementation of next-generation high-performance SoCs targeting 16nm FinFET technologies and below.
In this role, you will:
Lead the full RTL-to-GDSII physical implementation flow, including synthesis, floorplanning, place & route, CTS, timing closure, and sign-off.
Define and execute implementation strategies optimized for FinFET technologies, addressing challenges such as secondary power grids, track patterns, and advanced DRC constraints.
Perform Multi-Mode Multi-Corner (MMMC) timing closure and power optimization to achieve the best PPA (Power, Performance, Area) targets.
Conduct power integrity analysis and ensure robust IR drop and electromigration (EM) margins.
Drive physical verification closure including DRC, LVS, ERC, and antenna checks using industry-standard sign-off tools.
Collaborate closely with RTL and DFT teams to ensure physically aware synthesis, efficient scan-chain integration, and congestion mitigation.
Interface with foundries and EDA vendors to address technology-specific implementation challenges.
Contribute to EDA flow improvements and automation through scripting (Tcl, Python, or Perl) to enhance productivity and design quality.
Act as a technical pillar and mentor within the physical design team, supporting complex debugging and advanced optimization strategies.
Vertical Compute
Vertical Compute développe une technologie de mémoire verticalement intégrée pour l’intelligence artificielle et le calcul haute performance. Son architecture empile la mémoire au-dessus de la logique de calcul sur une même tranche de silicium afin de réduire les déplacements de données.
La technologie est proposée sous forme de chiplets pouvant être associés à des CPU, GPU et accélérateurs d’intelligence artificielle. L’entreprise a réalisé le tape-out d’une première puce de test intégrant la mémoire et la logique dans un procédé de fabrication sur tranche de 300 mm.
Levée
57 000 000 € levés
Dernier round
mars 2026
Équipe
37 personnes
Ville
Louvain-la-Neuve / Grenoble