Key facts
- Cadence Design Systems launched AuraStack, an AI agent for designing circuit boards and chip packages.
- The tool uses natural language prompts to automate design and testing processes.
- Nvidia chips will be used to accelerate AuraStack's AI capabilities.
- Cadence claims AuraStack can reduce time-to-market by up to 50% and increase task productivity by 15-fold.
- Early users include Nvidia, TSMC, and Schneider Electric.
Cadence Design Systems has introduced AuraStack, an artificial-intelligence "super agent" designed to automate the design of printed circuit boards and chip packages. The tool allows engineers to describe their objectives in plain language, with AuraStack then planning and executing the design and testing processes using Cadence's existing software. Nvidia chips are slated to accelerate the AI computations required for AuraStack's operations.
Cadence stated that AuraStack has the potential to reduce time-to-market by up to half and increase productivity on specific tasks by as much as 15-fold. This launch follows similar AI-driven offerings from Cadence earlier this year aimed at speeding up the design of integrated circuits themselves.
During a demonstration, Cadence showcased AuraStack's capabilities by having it rework a 5G smartphone's circuit board to create a more cost-effective version for a new market. The AI recommended consolidating components, projecting a 28% cost saving, and subsequently identified a more affordable power-management chip compatible with the revised board design.
Michael Jackson, Cadence's corporate vice president and general manager for system design and analysis, emphasized that the primary bottleneck in engineering is not automation but "engineering intelligence," referring to the complex reasoning across cost and performance trade-offs that AuraStack is built to manage.
Nvidia, Taiwan Semiconductor Manufacturing Co, and Schneider Electric are among the early adopters of AuraStack. Jackson also noted that customers can integrate AuraStack with their preferred AI models, including offerings from OpenAI (ChatGPT), Google (Gemini), Anthropic (Claude), or various open-source models. The pricing structure will be consumption-based, dependent on the computational load of the AI models, and will necessitate the use of Cadence's underlying design tools. AuraStack is scheduled for availability this year, with a complete rollout anticipated by September.