View a markdown version of this page

Reference architecture - Semiconductor Design on AWS

This whitepaper is for historical reference only. Some content might be outdated and some links might not be available.

Reference architecture

The previous section provided a path to migration. This section provides an annotated reference architecture diagram that reinforces these concepts. For a deep dive into the architecture and specific AWS Services that should be used, refer to Run Semiconductor Design Workflows on AWS.

Reference architecture diagram depicting semiconductor design on AWS.

Reference architecture diagram depicting semiconductor design on AWS

Architecture diagram descriptions
1 Determine what data is needed for proof of concept or test. 6 AWS compute is flexible and robust, more than capable of running semiconductor design workflows.
2 Transfer data into AWS via AWS Snowball, AWS Direct Connect, or using several other AWS services. 7 Store tools and job data on Amazon EFS, Amazon FSx for Lustre, and local disk. Optionally, move long-term data storage to Amazon S3.
3 Transferred data is stored in Amazon S3 buckets. You can access data stored in Amazon S3 from an Amazon EC2 instance or nearly any AWS service. 8 Once your data is in AWS, you can leverage other services, such as data lakes, AI/ML, and analytics.
4 Users access their environment through a remote desktop session or command line (ssh). 9 Isolating environments leads to enhanced security and limits third parties to only the data they need.
5 All of the infrastructure needed for semiconductor design workflows is available on AWS. 10 Encryption is everywhere and can be enabled with your encryption keys.