How Is AI Infusion in EDA Fueling the Automotive Revolution?

21 Sep 2022 • 8 minute read

Artificial Intelligence (AI) is now revolutionizing every industry it touches. Its inclusion in EDA is reshaping the automotive industry. The automotive sector is witnessing a profound shift with AI infusion in electronic design automation (EDA). Experts forecast that the AI-infused global market will increase at a compound annual growth rate (CAGR) of 39.4% from 2022 to 2030, or $20.76 billion. This transformation is leading to many innovations.

One such area where AI has had a significant impact is advanced driver assistance systems (ADAS) development. Customers are not just looking for a car for transportation; they want connected, autonomous, comfortable, and safe vehicles. Cars are becoming more intelligent and autonomous with AI inclusion in the research and development of electronic design automation (EDA) tools for manufacturing. AI is also changing the semiconductor industry in a big way, from system-on-chip (SoC) design and verification to packaging.

AI extensively helps product designers and development teams tailor all future products to match consumers' expectations. Machine learning techniques built into the Cadence design flow provide better productivity for design teams with the technological advancements from chip design to functional safety FuSA) and computational fluid dynamics (CFD). With the AI/ML infusion in EDA, it is possible to take a quick and correct decision over the edge (tinyML). So, it will not be wrong to say that EDA's AI is like automotive AI. In this blog post, we will take a closer look at the role of AI in the automotive revolution.

How Is AI Revolutionizing the Automotive Industry?

The automotive industry is going through a transition with the increase in semiconductors and customer expectations. It is expected to reach $70B by 2027 with ADAS, autonomous vehicles, digital cockpits, etc. Further, with the adoption of AI and processing at the edge, self-driving cars no longer seem like a fantasy. AI with deep learning improves accuracy and helps achieve greater vehicle autonomy with ADAS. AI embedded vision with depth perception and the 360-degree view help in accident prevention, decision making, in-car assistance, etc. These advances are making our automobiles safer, more efficient, more comfortable, and more pleasurable transportation experience.

Although fully automatic passenger vehicles (L5) have not hit the road yet, the industry is closely monitoring the developing autonomous systems. However, autonomous driving technology has been successfully and safely employed in last-mile delivery (LMD). LMD vehicles run at a lower speed, so they need shorter perception distances, shorter braking distances, and fewer requirements for safety. Further, the AI infusion and autonomous vehicles help to improve productivity and reduce the overall cost of LMD.

AI in EDA

With the increasing integrated functionality in SoCs, designers are under stress to meet the tight budget constraints. The traditional EDA tools use "rule-of-thumb heuristics" and require a designer's intuition for optimal design. Such modeling and simulation technologies suffer from problems such as:

Inability to glean insight from previous designs results in limited productivity and less accurate designs

Multiple iterations lead to increased design time
HLS often takes more time to complete the synthesis
Placement and routing are dependent on the prediction/ experience of the designer and come at the cost of run time
High cost of fabrication in terms of time and resources

To ensure a design's correctness, we must perform design verification before manufacturing. Traditional random/automatic test pattern generation (ATPG) schemes fail to increase fault coverage. AI has revolutionized the EDA industry. The training and inferences used in AI improve the chip designer's productivity. It helps design chips capable of handling the compute and EDA tools with capabilities to aid them with faster time closure and verification with reduced cost and improved QoR.

How Does AI/ML Improve the Design Space?

AI/ML is a good fit for EDA and automotive, it accelerates the design, and its introduction into EDA tools is a significant relief for designers. The use of EDA tools with AI functionality can significantly alter the trajectory of the design effort and helps to deal with the above challenges. Its benefits to design teams include:

Improved accuracy and efficiency
Forward visibility
Meet aggressive power, performance, and area (PPA) goals
Better data and chip layout with less human intervention
Early design closure

How Is AI in EDA Similar to AI in Automotives?

Achieving faster results with enhanced productivity and PPA are the primary objectives in EDA and automotive industries. AI promises to revolutionize EDA and automotive sectors with various applications and innovations. Whether autonomous cars, ADAS, or EDA, AI and ML algorithms present the opportunity to enable this electronics revolution and create a new renaissance. By incorporating AI functionality into existing EDA tools to aid silicon and software designers in making the EDA design processes more efficient and productive. Adopting AI and its derivatives helps automakers to improve the overall design to deliver faster and better results with multidisciplinary analysis and optimization (MDAO) techniques. Systems' improved accurate behavioral modeling yields greater product fidelity and security.

Electronic Design Assistance System (EDAS)

Cadence Offerings

Cadence provides EDA tools with AI/ML capabilities to produce better, more predictable outcomes from manual to full automation levels, as shown below. Our tools suggest solutions to common problems that might otherwise take design teams weeks or months to evaluate. We are also pushing the leading edge of ML and deep learning research to improve the design of ICs and verification closure with a vision toward design improvement.

Cadence AI/ML Solutions/ Technologies

Verisium AI-Driven Verification Platform represents a generational shift from single-run, single-engine algorithms in EDA to algorithms that leverage big data and AI to optimize multiple runs of multiple engines across an entire SoC design and verification campaign. By deploying the Verisium platform, all verification data, including waveforms, coverage, reports, and log files, are brought together in the Cadence JedAI platform. ML models are built, and other proprietary metrics are mined from this data to enable a new class of tools that dramatically improve verification productivity
Cadence Joint Enterprise Data and AI (JedAI) Platform accelerates the AI-based chip design. It improves productivity by allowing design teams to glean actionable intelligence from the massive amount of chip design data. Engineers can seamlessly manage both structured and non-structured data. The Cadence JedAI Platform makes it easier for designers to manage design complexities associated with emerging consumers, hyperscale computing, 5G communications, automotive and mobile applications, and more.
Optimality Intelligent Chip Explorer – Accelerating the time to market is the key to staying ahead of the competition. The Optimality Explorer's multidisciplinary analysis and optimization (MDAO) helps to achieve 10X productivity gains by exploring the full design space for optimal electrical design and can be used for Level 3+ automation of vehicles.
Cadence Cerebrus Intelligent Chip Explorer is a revolutionary, ML-automated chip design flow-optimization approach. It can be used for complex, large SoCs for Level 3+ automation, enabling engineers to optimize the flow for multiple blocks concurrently, which is especially important for the large, complex SoC. Additionally, the Cadence Cerebrus full-flow reinforcement learning technology significantly improves engineering team productivity.

Xcelium ML learns iteratively over a complete simulation regression. The core engine performance enhancements accelerate verification throughput by reducing stimulation cycles with matching coverage on randomized test suites. It is a perfect fit for Level 3 SoC design.
The high-performance data processing with Lidar, Radar, automotive cameras, etc. for ADAS (L2) is supported by Cadence Tensilica processor IP
Cadence Design IP and Cadence AWR RF to mmWave solutions help implement high-performance, cost-sensitive automotive radar front-ends, and beam steering antenna array technologies.

ADAS and sensor fusion

Apart from these, Cadence automotive innovation platform supports automotive manufacturers and provides tools like Innovus ML, Allegro ML, and Virtuoso ML to design SoC, PCB, and cards for Level 2 and 3 of autonomous driving.

Summary

The inclusion of AI in ADAS applications is a crucial enabler of vehicle autonomy. AI is helping automakers reduce the cost and time to lead the market. The AI infusion is transforming both hardware and software design, helping to meet the tight PPA budgets and providing an additional layer of safety.

Applications like blind-spot detection, lane departure, and depth perception are a few features that may help us get closer to the dream of self-driven cars controlled by vision and sensors using AI. You may also read my earlier blog about the depth perception of autonomous driving.

Learn More