Careers at Triumph Tech

How Imerza launched Built out their Post Production Workloads on AWS


Imerza - Rendering in the Cloud

James Barlow

Thinkbox Deadline

Company Name
Case Study Title
Imerza - Rendering in the Cloud


“When a rendering job is running, the customer can select the spot job from the Deadline monitor and launch as many as a hundred instances at a time on demand. Upon completion of the rendering job, the unnecessary instances end.”

Company Info

Imerza is a provider of the highest quality bespoke experiential and simulation tech.

AWS Thinkbox Deadline Summary

AWS Thinkbox deadline is an administration and compute control software program for rendering farms, which are high-performance computer structures built to render computer-generated imagery (CGI) commonly for video games, movies and television effects. AWS Thinkbox deadline is compatible with windows, Linux, and MacOS, and it supports more than 80 exceptional content introduction applications out of the box.

Rendering On-Premise versus Cloud

With the ever-increasing move to Cloud computing for businesses and individuals, rendering can also run in the Cloud, allowing rendering teams to take advantage of the Cloud computing’s benefits.


Teams are enabled with:

  • The scalability they need to adapt to each changing project
  • Cost-efficiency through the “pay-for-what-you-use” model
  • High availability and fault tolerance to ensure work is intact and reliable
  • Security to trust their work is safe from accidental and malicious threats

Within a VPC, Triumph Tech established an architecture that allowed for flexible yet powerful rendering operations. A key component of this setup was a backend API connecting a render queue to a hub for data, a repository, and file system. The render queue traditionally contains an Application Load Balancer (ALB) to distribute across a fleet of EC2 instances. This render queue can also call upon a usage-based licensing framework which also contains an ALB pointing to several EC2 instances. Both the render queue and the usage-based licensing framework can communicate with the worker fleet, which in turn also has an ALB & respective worker instances building content. On the side of this architecture is commonly a health monitoring mechanism. Lastly, note that there is a bastion host established in order to provide the entire system a dedicated, secure method of connecting. Below is a diagram illustrating the traditional and proven setup for the aforementioned architecture.


First, Triumph Tech installed a Thinkbox Deadline monitor for Deadline. To perform rendering jobs, Triumph Tech included a VPN to allow artists on-premises to communicate privately with Thinkbox.


Also, Triumph Tech established a FS Text file system to:

  • Hold all information between the monitor and workers
  • Maintain data license integrity
  • Offer centralized communication across machines.


The monitor instance communicates with other dedicated EC2 instances with Amazon Machine Images (AMIs) designed to meet the client’s specifications. Then several main worker instances are launched. When a rendering job is running, the customer can select the spot job from the Deadline monitor and launch as many as a hundred instances at a time on demand. Upon completion of the rendering job, the unnecessary instances end.


Imerzas’ team members are able to launch their rendering jobs from the monitor easily for far lower costs and enormous scalability with the AWS Thinkbox Deadline solution. Capital and resources are more efficiently allocated and users are able to control what they produce on-demand, ensuring agile work.


Due to the AWS Thinkbox Deadline architecture arranged by Triumph Tech, customers have had the option to dispatch their work from the monitor and central hub effectively, with far lower relative costs, and with increased versatility and flexibility. From a business point of view, this opens up capital and assets to be more proficiently allocated, and from the craftsmanship and designing viewpoint clients can direct their work in an on-request way that allows for flexible work. From a creative point of view, this allows for the quality of the art and content to be at the forefront, rather than the art and content being mitigated by the aforementioned limiting factors associated with on-premise rendering.