Cadence Integrates NVIDIA NIM Microservices into Its Design Process

From semiconductors to pharmaceuticals, incorporating tailored AI applications into design has become crucial for accelerating the discovery process. However, using AI requires vast datasets for training, which necessitates safeguarding sensitive information and valuable IP. With our deep appreciation of how important security is to the AI-driven design process, Cadence now integrates NVIDIA NeMo and NIM microservices, part of the NVIDIA AI Enterprise software platform, into all our generative AI applications.

Flexible and Secure Solutions

Figure 1. Cadence’s AI analytics and optimization flow focuses on flexibility for design flows and security of intellectual property.

Through our collaboration with NVIDIA, we are steering innovation in semiconductor design, automotive, and robotics by optimizing software and hardware. This effort works to reduce the carbon footprint of AI data centers and accelerate advancements in digital biology and drug discovery. We're also working with NVIDIA to deploy solutions onsite or in the cloud to give users the flexibility and security they need.

Cadence delivers our generative AI applications using NeMo Retriever and NIM microservices for retrieval-augmented generation (RAG). Using RAG, models can fetch facts from external resources, providing authoritative responses that cite sources and build trust. This step is crucial for enhancing the accuracy and dependability of the content created by generative AI models. RAG also aids models in clearing up ambiguities in user queries and reduces the possibility of incorrect assumptions. It can be implemented with only a few lines of code, making it faster and less costly than retraining models. RAG also allows users to hot-swap new sources as needed.

RAG provides a secure method to enhance the accuracy of AI-generated outputs without exposing sensitive IP—such as email, documentation, and design information—to the outside world or expending the resources needed to fine-tune an existing large language model (LLM). Using NVIDIA NeMo Retriever microservices throughout the design process enhances generative AI applications with large scale ingestion and RAG capabilities which can be connected to proprietary data wherever it resides. The NeMo Retriever collection of microservices enables world-class information retrieval with high accuracy and data privacy, generating real-time business insights.

The Future of AI-Driven Design

The future of AI in design will be shaped by those who can leverage its power while adhering to strict security standards. Cadence, through collaborations with companies like NVIDIA, is pushing technological boundaries, helping keep our customers at the forefront of advancement.

Learn more about the NVIDIA and Cadence partnership to accelerate AI-driven design.