Generative AI

Generative AI, or GenAI, is a subset of artificial intelligence technology that creates something new from a dataset of previous examples. It typically relies on complex algorithms and neural networks to simulate human creativity and produce new output. In chip design, GenAI can help explore architectural possibilities, including fine-tuning P&R (place and route) settings to achieve better power, performance, and area (PPA). It can learn a vast spectrum of Verilog codes, and when given prompts with specifications, can produce Verilog codes automatically. This technology has the potential to revolutionize many industries by optimizing results that couldn’t have been achieved previously, as well as automating, thus increasing user productivity substantially.

What is the significance of Generative AI?

The three potential benefits of GenAI are listed below:

1. Increased efficiency: GenAI can automate processes that normally require heavy manual effort, such as place and route in chip and board designs or optimizing workflows. This can have upwards of 10X productivity improvement.
2. Improved results: GenAI can help humans generate novel ideas and solutions by exploring a wider range of possibilities than a human could reasonably consider previously.
3. Better decision-making: GenAI can help humans make better decisions by providing data-driven insights and predictions.

How does Generative AI work?

GenAI uses machine learning technology to optimize results by exploring more possibilities than traditionally humanly possible. This can be done on the fly when operation on a design, or through a large language model that has been trained on a wide variety of design data. It involves training an algorithm to make decisions based on previous data or patterns, thereby enabling the system to generate better and more accurate outputs. This approach is particularly effective in tasks with specific parameters or weights. GenAI algorithms can augment users’ work with more and better scenarios, ultimately leading to improved results and differentiated products.

Generative AI with Cadence

Cadence offers a broad range of GenAI products that enable customers to enhance or optimize their product performance and ensure increased productivity of their design workflow. The following is a list of Cadence AI products that use GenAI:

Cadence.AI Generative AI Platform is an AI-enabled big data analytics platform facilitating smarter design optimization and enhanced productivity. By accessing deep data insights that are otherwise difficult to harness, Cadence.AI enables users to unlock hidden data trends, visualize data and automatically generate design improvement strategies.

Cadence Cerebrus, which was recognized as a recipient of Fast Company's 2022 Next Big Things in Tech Awards, leverages reinforcement learning to achieve PPA objectives and streamline the entire chip design process. This not only saves time but also enhances the overall design quality. Cadence Cerebrus can expedite chip layout and produce superior outcomes, thereby improving the engineering team’s productivity.

The Cadence Verisium platform uses Generative AI to improve predictability, quality, and thought throughout the verification flow. Connected to Cadence.AI Generative AI Platform, Verisium provides advanced collaboration capabilities for simulation, formal, emulation, and prototyping platforms, accelerating coverage closure. It greatly increases total verification throughput with the most bugs found per dollar per day.

Cadence has incorporated Generative AI into its PCB design software, Allegro-X AI, to streamline the placement and routing of components on PCBs. This significantly reduces the design time by automating numerous manual processes while also enhancing design quality, enabling engineers to create more efficient PCB layouts. The automation provided by Allegro-X AI allows designers to explore a wider range of design possibilities that would be difficult to assess without this robust tool.

At the system level, the Cadence Optimality Intelligent System Explorer uses Generative AI to provide multiphysics system analysis and optimization covering the IC, package, and PCB as well as the complete electronic system. It enables designers to explore 3D electromagnetic and high-speed signal and power integrity results and zero in on the optimal design. This greatly increases design productivity while helping designers create more optimal system layouts.