AMD enables exceptional flexibility and performance with high-performing Azure VMs, Azure NetApp Files

Take it from a company at the forefront of the industry—it’s an incredible time to be in the semiconductor business. Every day, engineers at AMD are hard at work inventing and improving the chipsets, data processors, and software applications that help form the computational backbone of a digital information economy measured in trillions of dollars’ worth of goods and systems. The leadership solutions from AMD impact the way people live, learn, and play. This level of customer-focused execution requires high-performance computing capacity for electronic design automation (EDA).

Moreover, there is an inexorable customer appetite for computation that requires AMD to deliver products quickly and cost-effectively, with energy efficiency to help customers advance their sustainability goals. As semiconductor workloads become significantly more complex with higher integration and advanced technology nodes, AMD has worked closely with Microsoft Azure to deliver these outcomes for AMD engineers. By using Azure Virtual Machines and Azure NetApp Files, AMD drives continuous innovation in addressing the compute and storage demands for its scale-out workloads.

“Working with Microsoft, our engineers are innovating at a remarkable pace, and all of that begins with the radical new way that we approach electronic design automation on the Azure cloud platform.”

Mark Papermaster, Chief Technology Officer and Executive Vice President of Technology and Engineering, AMD

Running EDA on Azure

AMD, a leading multinational company, has been delivering some of the world’s most advanced processors for the world’s most advanced ideas. As part of a continued focus on digital transformation, AMD continuously evaluates its EDA workloads to deliver breakthrough innovations at a fast pace. The cloud is a critical ingredient in this journey. “This is a pivotal time for transitioning semiconductor workloads to the cloud, especially as we advance our goals and commitments to our employees and the communities we serve,” notes Mark Papermaster, Chief Technology Officer and Executive Vice President of Technology and Engineering at AMD.

Expanding on its existing relationship with Microsoft, AMD is providing its engineers exceptional performance for key EDA workloads by using Azure HBv3-series virtual machines (VMs) featuring 3rd Gen AMD EPYC™ CPUs with AMD 3D V-Cache™ technology. This is complemented by high-performance storage in the cloud with Azure NetApp Files. “The world’s most ambitious innovations rely on high-performance computing to bring them to reality, and supporting that is our mission at AMD,” says Papermaster. “Working with Microsoft, our engineers are innovating at a remarkable pace, and all of that begins with the radical new way that we approach electronic design automation on the Azure cloud platform.”

A complete solution stack for chip design in the cloud

High performance is built in from the ground up with new Azure solutions powered by AMD. The company’s EDA workloads run on Microsoft Azure high-performance computing (HPC) resources on its HBv3 VMs, which are optimally configured with high node-to-node interconnectivity and advanced AMD processors. AMD uses Azure CycleCloud to spin up VMs on demand. This allows AMD to easily respond to peak capacity needs for high-performance computing and, most importantly, provide engineers with excellent performance for EDA scenarios and simulations.

AMD leverages Azure NetApp Files as a key high-performance storage solution for EDA. It helps ensure that the cloud compute instances used by AMD are provisioned with a level of read/write performance that matches the scale of their computational needs. “Azure NetApp Files is essential to achieve our performance and storage requirements,” Papermaster says.

Together, these Azure solutions give AMD a compelling solution stack for chip design in the cloud. One of the primary benefits of the cloud is scalable computing capacity. This has enabled AMD to speed up chip design and verification, while helping ensure accuracy and precision of results. Another key benefit of the cloud is to optimize computing capacity when the demand profiles change. After the designs are complete and handed off to manufacturing, AMD engineers have many strategies for the computational capacity on cloud. They can pause, reuse, or take on new initiatives to accelerate the pace of innovation, all the while helping the company meet strong demand for consumer products and technological advancements.

HPC that benefits everyone

Based on its engineers’ experiences to date and exceptional customer feedback, AMD trusts Azure HPC resources, including its HBv3 VMs and Azure NetApp Files, to serve as the destination for its most complex and critical workloads. Typical engineering projects require thousands of engineers and millions of compute hours, and AMD has used Azure to make its capacity manageable and scalable. This strategy leads directly to cost optimization and continuous innovation, while benefiting customers with compelling compute performance.

With new generations of high-performance VMs perfectly suited for semiconductor development and design in the pipeline, AMD is excited for what’s ahead in its ever-evolving cloud journey. “Our breakthrough success at AMD is proof that what we’re building on Azure is world class and will continue to break records moving forward,” says Papermaster.

Find out more about AMD on YouTube, Twitter, Facebook, and LinkedIn.