Hardware
Optimizing Agent Reliability and Tool Calling Efficiency for Chip Design Using NVIDIA NeMo Gym
NVIDIA NeMo Gym is an open‑source evaluation environment that diagnoses failure modes and tracks error propagation across multistage chip‑design workflows. In Siemens’ Fuse EDA case, the Gym was used
NVIDIA NeMo Gym is an open‑source evaluation environment that diagnoses failure modes and tracks error propagation across multistage chip‑design workflows. In Siemens’ Fuse EDA case, the Gym was used to refine a Liberty‑file generation agent, raising its verification score from ~55 % to ~95 %, while reducing agent turns from 155 to 53 and output tokens from ~81 k to 29 k through skill‑set and tool‑description adjustments guided by verifier diagnostics. A domain‑specific verifier assesses coverage, validity, accuracy, and integrity of the final Liberty artifacts—imposing a zero reward for any gate failure—and this same evaluation methodology can be applied to engineering agents in EDA, CAE, and other domains.
Related
- How to Run an Autoresearch Workflow with RL Agent Skills and NVIDIA NeMo
- Evaluating AI Agent Skill Performance with NVIDIA SkillEvaluator
- NVIDIA-Verified Agent Skills Provide Capability Governance for AI Agents
- How to Evaluate AI Agents From Tool Calls to Task Completion
Source: NVIDIA Developer | 2026-10-14