Lift-n-Shift your CAx, PLM Applications to the Cloud

PLM and CAD Applications like Siemens Teamcenter and NX, Solid Edge CAD software have traditionally been implemented On-premise with multi-tier physical servers hosted on the premises. The end-users machine needs to have a high capacity of resources and dedicated Video/GPU memory to handle heavy workloads like graphical simulation and rendering. Procuring such memory-intensive machines will need in advance planning and can get very expensive.

These issues can be addressed by migrating CAD & PLM implementations natively to Cloud and using Amazon AppStream 2.0 to stream Applications and Desktop securely and reliably to the end user’s machine in a browser window.

Native Teamcenter & CAx on AWS Cloud

High-level architecture of the Teamcenter components running on AWS Cloud along with AppStream delivering the client applications by streaming to the browser window of the user.

Teamcenter & NX

In the above architecture, the simple Teamcenter implementation consists of a License Server, App Server with Web-tier application running for Active Workspace Launcher and a Volume Server (with an EBS volume attached to the EC2 App Server)
Oracle RDS Instance for Database as a service.
The App Server and License Server are having an Elastic IP attached.
All the infrastructure is in Private Subnet with no public-facing infrastructure for maximum security.

AppStream Application Delivery

Delivery of AppStream 2.0 to end-user applications

AppStream 2.0 service by Amazon can stream the Desktop or the Application view (pixel data) of a Cloud EC2 Machine to the User in a Browser window. All the heavy-lifting and the processing of the Application and the task happens on the EC2 and only the Application view pixel data is streamed as a video consisting of frames just like YouTube and Netflix.
AppStream 2.0 supports different streaming instance types for different types of workload like memory-optimized, compute-optimized, graphics optimized instances & general-purpose instance for normal consistent workloads.
For our use case, we have a graphics optimized AppStream Instance for CAD Users and a general-purpose Instance for Teamcenter PLM Users.
The Administrator created a single base Image consisting of all necessary Client Applications required by both type of Users like Office Suite of Applications, Notepad, Mozilla or Chrome Browser and Enterprise VPN software (only if required to connect to Organization network)
The Base Image is then modified to install CAD and PLM specific applications and corresponding plugins in Teamcenter and two separate EC2 fleets and AppStream stacks are created.
The Graphics stack is associated with NX CAD Users and PLM Stack is associated with PLM Users. Both of these machines are pointing to the Cloud Teamcenter Application Server for fetching and storing data.
AppStream instances are also launched in the same private subnet as Teamcenter infrastructure. The Streaming service gateway by Amazon is responsible for securely streaming Desktop or Application view of these private instances publicly to End User.
Each User who logs in to the AppStream will have a dedicated EC2 running inside AppStream Network and the session duration and timeout can be configured. Such users can also have a Home directory configured for storing any persistent data between each AppStream sessions which in turn is stored on an S3 Bucket.

AppStream Catalog

NX 12 running on an AppStream User Instance

AppStream 2.0 developers have also created a new shared image feature as an efficient way to reuse custom image creator instances

Conclusion

Migrating CAD & PLM workloads and implementation on the cloud have never been easier and simpler. Couple that with Application Delivery to End users using AppStream 2.0 greatly reduces the CapEX and only contributes towards OpEX as the AppStream instances are on-demand based and no cost is incurred when the machine is not being utilized. It also helps the organization to scale up without procuring any Hardware infrastructure easily to any number of users across different geographical locations and creates a cohesive remote workplace environment, especially in this post-COVID-19 situation.