Chipmaker Advanced Micro Devices Inc. said today it plans to shift some of its electronic design automation workloads for chip design onto Google Cloud, in order to extend the capabilities of its own data centers.
EDA is an important part of the chip design process that involves using computer-aided design software to create printed circuit boards, integrated circuits and microprocessors. Chipmakers require complex designs to pack elements onto circuit boards at extreme density, and EDA provides both standardized processes and automation to facilitate more rapid development.
Using EDA software, chipmakers can design, model, simulate, test and analyze new circuit designs to assess their performance and identify any issues before they go into production.
Given the importance of EDA, it’s easy to see why AMD would want to run some of these workloads on Google Cloud, where it will benefit from advanced networking, storage and artificial intelligence capabilities, the company said.
AMD explained further that given the critical role of scale, elasticity and efficient utilization of resources in chip design, it also intends to leverage Google’s newest compute-optimized C2D virtual machine instances that are powered by its own 3rd Gen AMD EPYC processors. Using these VMs, AMD believes, it will be able to run more designs in parallel, allowing greater flexibility in the way it manages its short-term compute demands without reducing resource allocation on longer-term projects.
Mydung Pham, AMD’s corporate vice president of silicon design engineering, said the Google Cloud C2D instances provide a new avenue of high-performance resources that allows his teams to mix and match the right compute solution to each EDA workflow.
Google stressed that AMD will enjoy benefits such as increased flexibility to run applications in the most efficient manner possible and improved design and operations enabled by its AI and machine learning tools. In addition, it will also benefit from superior transparency around costs and resource consumption, Google said.
Sachin Gupta, Google Cloud’s general manager and vice president of infrastructure, said the speed, scale and security of the cloud leads to much-needed flexibility for chip designers such as AMD.