IIC 2023 | Cadence explains the new engine of EDA development and creates the cornerstone of intelligent system design
Preface
From March 29th to 30th, the 2023 International Integrated Circuit Exhibition and Seminar (IIC Shanghai) was successfully held at the Shanghai International Convention Center. With the theme of "Innovation, Change, and Firm Forward", this conference focused on "carbon neutrality and Green Energy" electronics industry development, China's IC design achievements, EDA/IP, MCU technology and applications, efficient power management and wide bandgap semiconductor technology, radio frequency and wireless communication technology and other fields.
At the China IC Leaders Summit and EDA/IP and IC Design Forum, Cadence EDA Software Technical Sales Director Geng Xiaojie and Cadence Product Technical Sales Manager Wan Li delivered speeches on the topics of "Computational Software—The Cornerstone of Intelligent System Design" and "Big Data " respectively. Data + AI analysis, Cadence JedAI platform launches a new engine for EDA development" was a wonderful speech.
Geng Xiaojie:
“Computational Software”
It is the cornerstone of intelligent system design
At the China IC Leadership Summit at IIC Shanghai,
Geng Xiaojie
shared the complexity of intelligent system design in the era of intelligent systems, and introduced
the solutions Cadence provides for intelligent system design and even broader fields through its excellent computational software (Computational Software) technology
.
Intelligent system design is increasingly complex
Geng Xiaojie said that intelligence is changing our daily lives and work. Powerful intelligent systems have many chips, such as SoCs integrating CPU and GPU, wireless chips, memory, ISP chips, power management chips, etc.
Systems are becoming more complex, smaller in size, and software stacks are becoming more complex. Different devices have different requirements for the system layer. For example, smart cars must consider aerodynamics, aeroacoustics, thermal management, EMC/EMI, etc., and require a lot of simulations for optimization.
When an intelligent system is running, the intelligent layer on the silicon and system layers will generate a large amount of data, which requires AI/ML algorithms for data analysis and learning to provide intelligent user experience. This work can be done at the edge or in the cloud.
The complexity of designing intelligent systems depends on the application requirements, level of intelligence, technology used and level of integration. Many decisions need to be made during the silicon layer chip design phase, advanced nodes have many new challenges, and 3D-IC or system packaging will also add complexity.
The system layer needs to run a large number of simulations to optimize performance and consider the interaction of software and hardware to ensure functional implementation. All of this makes designing intelligent systems exponentially more difficult.
Intelligent system design beyond traditional EDA
Geng Xiaojie said that the Cadence Intelligent System Design ™ strategy has promoted the growth of the company's core EDA and IP businesses, which is based on providing excellent semiconductor device design through core EDA and IP businesses.
Cadence has expanded its core competencies in computing software into two new areas - system innovation and pervasive intelligence, applying AI and algorithms to core businesses and specific vertical areas.
Data explosion, cost, mechanical and silicon design complexity have increased AI computing requirements, and Moore, CPU and software performance scaling have encountered challenges. Cadence's strategy can achieve the integration of EDA, system design, and AI, and apply it across multiple system domains throughout the design.
Artificial intelligence empowers intelligent system design
Geng Xiaojie believes that AI can empower design intelligent systems , explore the design space and improve tool automation, using computer power to make the best choices faster and achieve higher quality.
Traditionally, designing a PCB to meet requirements such as SI/PI and thermal profile requires 10 design parameters and more possible values, and it takes a long time to run 10 to 10 billion simulations. Previously it could only be set up empirically and simulated, iterating until the best design was achieved or time ran out.
AI/ML algorithms can make informed decisions based on previous simulations and designs, enabling better productivity. With the iteration, the accuracy of the model is continuously verified, the model is refined, and the model converges quickly to obtain satisfactory results.
The future of system design
Geng Xiaojie explained that EDA data covers a variety of heterogeneous, structured and unstructured information, making it challenging to store and process EDA data.
The AI-driven process optimizes large amounts of EDA data in an open, enterprise-level, AI-driven, large-scale, cloud-supported data analysis environment, which can greatly increase designer production potential and allow AI to process 10 designs at the same time instead of Work on them one at a time.
He finally emphasized that chip design has entered a highly integrated, complex, and intelligent stage, with endless technical challenges emerging. The powerful capabilities of Computational Software will find applicable scenarios in more industries, greatly improving productivity and making designs more potential.
Wanli:
Cadence JedAI platform launches new engine for EDA development
At the EDA/IP and IC Design Forum, Wan Li shared data development trends and challenges, and introduced the architecture, features and advantages of the Cadence Joint Enterprise Data and AI (JedAI) platform .
Data explosion drives AI-Driven design and verification tools
Analysis shows that by 2025, data will grow to 180ZB, data interaction has increased by 5000% in the past 10 years; the data being analyzed has decreased by 2%, and unstructured data has increased by 80%. The trend is that data processing requires high-performance, low-power computing; data analysis requires information-driven decision-making; data transmission requires high-bandwidth, low-latency connections; and data storage requires high-density, cost-effective storage .
Wanli pointed out that in the large-scale SoC design and verification process, massive amounts of data are generated every day. The Cadence JedAI platform can help engineers collect useful intelligent information from large amounts of chip design and verification data , opening the door to a new generation of AI-Driven design and verification tools, thereby greatly improving productivity and achieving power, performance and area (PPA) ) for the best results.
The Cadence JedAI platform unifies big data analytics across its various AI platforms, including Verisium ™ verification, Cerebrus ™ implementation and Optimality ™ system optimization, as well as third-party silicon lifecycle management systems. The platform helps users easily manage emerging consumer, hyperscale computing, 5G communications, automotive electronics and mobile applications with increasing design complexity. When using Cadence analog/digital/PCB implementation, verification and analysis software, all big data analysis tasks can be deployed uniformly through this platform.
Cadence JedAI empowers intelligent system design
Wanli believes that AI is raising productivity to a new level . From manual design to transistor-level design, to unit-based design and IP reuse, AI-based EDA can increase productivity by 10 times at every step.
The Cadence JedAI platform includes three parts: AI-driven verification, implementation, and system analysis. This cross-domain big data EDA platform architecture has a highly scalable, distributed, secure infrastructure, and an open industry standard user interface optimized for Cadence tools. and scripting environments to help facilitate AI and analytics deployments and double productivity.
The future of smart chip design
According to Wanli, the use of the new machine ML-based tool Cadence Cerebrus can automatically expand digital chip design to achieve a revolution in productivity and power consumption, performance and area; the unique enhanced ML can provide up to 10 times productivity and 20% PPA Improve.
He finally said that Cadence's strategy is to use AI and data analysis, system design and analysis, especially core EDA and IP to promote innovation across multiple fields and empower the realization of ubiquitous intelligence, system innovation and excellent design.
About Cadence
With more than 30 years of expertise in computing systems, Cadence is a key leader in the electronic systems design industry. Based on the company's intelligent system design strategy, Cadence is committed to providing software, hardware and IP products to help electronic design concepts become reality. Cadence's customers are the most innovative companies around the world, delivering everything from chips and circuit boards to the most dynamic application markets such as hyperscale computing, 5G communications, automotive, mobile, aerospace, consumer electronics, industrial and medical. A complete system of excellence in electronics. Cadence has been ranked among Fortune magazine's 100 Best Companies to Work For for eight consecutive years. For more information, please visit the company's website at cadence.com.
© 2023 Cadence Design Systems, Inc. All rights reserved. All rights reserved worldwide. Cadence, the Cadence logo and other Cadence marks listed at www.cadence.com/go/trademarks are trademarks or registered trademarks of Cadence Design Systems, Inc. All other marks are the property of their respective owners.
京公网安备 11010802033920号