Arm Moves Production-Level EDA to the Cloud with the Help of AWS
By: Rene Haas, President, IPG, Arm
- Migrating design and verification to Amazon Web Services (AWS) Graviton2 instances reduces cost, complexity and risks, while increasing throughput by up to 10x
- Cadence, Mentor, Synopsys and other leading EDA vendors are porting and optimizing their technology to run on Arm compute environments
- An Arm-based environment for designing Arm-based silicon in the cloud will accelerate innovation and choice for everyone in the ecosystem
Released in February, the Arm Cortex-M55 CPU quickly became a milestone for the Arm ecosystem because it gave hardware and software developers the power to incorporate unprecedented ML capabilities into the billions of IoT products coming to market. It also was the first time Arm leveraged Arm-based AWS Graviton2 instances to perform production-level design validation for its IP.
As AWS announced last week, Arm is moving the majority of its Electronic Design Automation (EDA) workloads to AWS, as part of our effort to reduce our global datacenter footprint by at least 45% and our on-premises compute capabilities by 80% as we complete our migration to AWS.
We have already realized a 6x improvement in performance time for EDA workflows on AWS and see the potential for increasing throughput by 10x. In addition, Arm is generating powerful engineering, business, and operational insights that help increase workflow efficiency and optimize costs and resources across the company.
Our goal ultimately is to move most of our design work to Arm-based cloud services over the next several years, and we think we will have company too. By sharing our experiences across the Arm ecosystem, we hope to encourage partners to design their Arm-based silicon on Arm-based cloud services.
EDA: An Inside Job Until Now
EDA in many ways is a natural fit for the cloud as chip designers often need simultaneous access to tens of thousands of CPUs to run this compute-intensive task. Meeting tight deadlines can mean consuming millions of core-hours per month. Building a data center to accommodate the gigantic bursts of computing power needed for tomorrow’s complex devices is well beyond the reach of most organizations.
Most EDA workloads, however, still largely get performed in on-premises data centers owned and/or managed by semiconductor companies because of the complexities and risks involved in migration. Importantly, design workflows have to be unified and semiconductor designers have to figure out a balance between workloads that are best suited for cloud, and which make more sense staying in-house.
To bridge this chasm, Arm has been working with major EDA partners such as Cadence Design Systems, Mentor Graphics’ (a Siemens Business), and Synopsys to port their EDA tools to Arm. Already we’re seeing significant results in terms of performance and cost savings for customers.
Likewise, AWS continues to expand its portfolio to accommodate these types of workloads. The recently unveiled AWS Graviton 2-based C6gn instance delivers up to 100 Gbps networking while providing a 40% price-performance advantage over conventional processors.
An Arm-on-Arm World
Arm certainly won’t be alone in looking at ways to improve the semiconductor industry through the cloud. The Fifth Wave of computing (the convergence of 5G, AI, and IoT) provides the technology community with an unprecedented opportunity to make a difference in the world. Autonomous cars will reduce accidents and emissions, while track and trace technologies will help contain COVID-19. Silicon will continue to get smarter and will be even more omnipresent in our lives.
Ubiquitous computing, however, will also mean developing products that are smaller, smarter, more distinct in their design, and less expensive than ever before. We as an industry will also have to meet these seemingly contradictory goals in less time and with fewer dedicated resources. Design, simulation, and verification in the cloud—particularly on compute instances that can deliver maximum performance with minimum energy consumption—can erode the barriers that exist between an idea and a finished product produced in volume. Just as cloud services transformed the software business, new services and new startups will follow as cloud-based silicon design takes hold.
If you would like to learn more, please check out the expert blogs on our EDA in the cloud journey in the Arm Community.
Arm technology is at the heart of a computing and data revolution that is transforming the way people live and businesses operate. Our energy-efficient processor designs and software platforms have enabled advanced computing in more than 190 billion chips and our technologies securely power products from the sensor to the smartphone and the supercomputer. Together with 1,000+ technology partners we are at the forefront of designing, securing and managing all areas of AI-enhanced connected compute from the chip to the cloud.
All information is provided "as is" and without warranty or representation. This document may be shared freely, attributed and unmodified. Arm is a registered trademark of Arm Limited (or its subsidiaries). All brands or product names are the property of their respective holders. © 1995-2021 Arm Group.