Lila Sciences expanded its Series A with an additional $115 million — including investment from Nvidia’s venture arm — bringing its total haul to $350 million to fund ambitions in AI‑driven molecular design and physical‑world integration. The firm aims to connect generative models with lab automation for end‑to‑end discovery. Benchling announced an integration with Nvidia BioNeMo models and NIM microservices to provide protein structure prediction (OpenFold2) and other biomolecular modeling tools inside Benchling’s platform. The integration is intended to let scientists run structure prediction, generative modeling and property predictions without moving data across tools or managing complex infrastructure. Together these moves reflect consolidation of compute and model access in discovery workflows. Companies and academic labs will watch whether the combined investments and integrations shorten design‑test cycles and improve hit quality in molecular design.