Transforming Andresen’s Render Capabilities with AWS Cloud Integration

A Case Study on Enhancing Efficiency, Security, and Performance in Digital Production

Overview

GPL Technologies partnered with Andresen, a digital studio specializing in retouching, CGI, and VFX, to revolutionize their rendering capabilities through AWS Cloud solutions. The customer chose to work with GPL Technologies due to their proven track record and expertise in AWS best practices. GPL Technologies’ ability to deliver a secure, scalable, and high-performance solution distinguished them from other partners. By implementing a standard AWS Thinkbox Deadline render farm and enhancing it with custom AWS plug-in integrations, GPL Technologies enabled Andresen to achieve significant improvements in efficiency, security, and scalability

The customer chose to work with GPL Technologies due to their proven track record and expertise in AWS best practices. GPL Technologies’ ability to deliver a secure, scalable, and high-performance solution distinguished them from other partners. Other partners had previously tried and failed to implement a successful cloud infrastructure for the customer.

By implementing a standard AWS Thinkbox Deadline render farm and enhancing it with custom AWS plug-in integrations, GPL Technologies enabled Andresen to achieve significant improvements in efficiency, security, and scalability.

Takeaways

Challenges

  • Time-consuming, inefficient manual rendering processes using local workstations.
  • Inability to scale rendering power due to limited on-premise capacity.
  • Requirement for secure and encrypted data handling for a high-profile client.

Solution

  • Implementation of a secure, private render farm using AWS EC2 instances and AWS Thinkbox Deadline with custom GPL Tech plugins for Deadline.
  • Achieved rapid shot turnaround through enhanced data sync integration.
  • Creation and deployment of cost estimation plugins.

Results

  • Secure, scalable, and high-performance AWS cloud render solution.
  • Significant reduction in render times from hours to minutes, utilizing the right-size instances for maximum performance/value.
  • Enhanced efficiency with automated data syncing.

Challenges | Addressing Rendering Challenges: A Comprehensive AWS Solution

Andresen faced multiple failed attempts with mutiple vendors at implementing a cloud rendering solution, which underscored the limitations of their previous workflow. Their earlier workflow involved inefficient manual rendering processes, where artists had to remote into other workstations, causing significant workflow disruptions and delays. This method not only monopolized entire workstations but also required other artists to monitor the renders, exacerbating inefficiencies. The local rendering capacity was further constrained by power limitations, preventing the addition of more hardware, thereby capping their on-premise render capabilities.

The need for an enhanced rendering solution was critical. Artists could not simultaneously work and render, often having to wait for renders to finish before continuing with other tasks, or displacing other machines needed by freelancers. This “musical chairs” approach to rendering was not sustainable. The customer required render capacity beyond what was possible on-premise and sought a first-class cloud solution. Given AWS’s industry reputation for robust and comprehensive solutions, it was the clear choice.

Security was another paramount concern due to the sensitive nature of the client’s data—an AAA Tech Company—necessitating robust encryption and secure, private data flows. The solution implemented was a hybrid studio model, where all rendering is conducted fully in the cloud while artists and their workstations remain on-premise. This model not only addressed the rendering capacity limitations but also ensured seamless integration with their existing workflow, providing a secure, efficient, and scalable solution.

Solution | Implementing Efficiency: Transforming Andresen's Render Workflow with AWS

GPL Technologies implemented a comprehensive AWS-based solution to address Andresen’s rendering challenges. The initial phase involved setting up a secure, private, customer-key-controlled AWS Thinkbox Deadline render farm tailored for the AAA Tech Company, ensuring “no-public-data-flow” and “encryption-at-rest”. This included integrating AWS Site-to-Site VPN, AWS DataSync with a custom Deadline plugin, and FSx Lustre for high-performance and scalability for storage. The architecture utilized best-practice zero-trust architec- ture with zero trust network architecture within the AWS VPC.

Building on the success of the initial phase, the solution was enhanced with an upgrade for improved efficiency. This second phase included upgraded integration between Deadline and AWS via an improved data sync plugin, higher-end instances for better performance and stability, and the implementation of a PPA (power purchase agreement) for specific render capacity pricing. Additionally, a custom “Cost Estimation” plugin using the AWS Spot API was developed to provide post-render job calculations.

During implementation, challenges and discoveries required dynamic adjustments. One significant pivot involved right-sizing instances to meet render time goals, shifting from cheaper, lower-end instances to higher-end ones. This reduced render times dramatically, from an hour to 10-20 minutes. Another discovery was the customer’s unique rendering requirement of producing massive single images instead of traditional frame sequences. This necessitated the use of a “tile render” feature, adding pipeline complexity by having to break up large images into smaller tiles for individual rendering and subsequent reassembly.

These adjustments were efficiently managed due to the infrastructure-as-code approach, allowing seamless integration and minimal disruption. The tailored AWS solution not only met Andresen’s security and perfor- mance requirements but also significantly optimized their rendering workflow, enabling them to achieve their goals effectively.

Andresen - AWS Cloud Render Map

Results | Achieving Efficiency and Security: The Impact of AWS Integration on Andresen's Render Process

The AWS Cloud integration delivered significant benefits for Andresen, transforming their rendering workflow with efficiency and security. Render times were drastically reduced from hours to as little as 15-30 minutes, enabling faster project turnaround. The secure and scalable solution met the high standards required by Andresen’s clients, ensuring data was handled with the utmost security. Operational efficiency improved markedly with automated data syncing and cost tracking, freeing artists from technical tasks and allowing them to focus on their creative work.

           Key deliverables of the project included:

  • A fully functional, encrypted Thinkbox Deadline render farm in the cloud, connected to on-premises locations via AWS Site-to-Site VPN.
  • Hybrid V-Ray license server, utilizing constant usage licensing with additional “usage-based licensing” (UBL) for artist application burst scaling.
  • Locked down private networking with web proxy for traffic control, ensuring that the render farm could only communicate with specific license servers.
  • A complete code-based Linux render node software stack, able to be installed and managed via automation for streamlined updates and maintenance.

Custom plugins were a critical component of the solution:

  • A custom GPL AWS DataSync plugin facilitated seamless, on-demand data synchronization at the project level, reducing data sync management via manual intervention from 5-10 minutes per project to just 60-100 seconds with no manual intervention required.

  • The GPL AWS Cost Plugin for Deadline provided post-render cost calculations and exported detailed cost reports, although it did not account for failed jobs due to development time constraints.

Throughout the process, challenges such as integrating a PPA for specific pricing on render capacity were managed with minimal disruption due to the infrastructure-as-code approach. Ongoing support included unofficial training on render farm operation, debugging, troubleshooting, ad-hoc render wrangling, and day-to-day management. This support also covered render specific issues like “external resource links” that were only available on-premises, enabling the team to resolve rendering errors related to missing assets.

Overall, the AWS integration provided a secure, efficient, and scalable rendering solution that adhered to best practices, significantly enhancing Andresen’s operational capabilities and meeting their stringent security requirements.

Transform your workflow with a secure, scalable, and high-performance AWS cloud rendering solution by partnering with experts like GPL Technologies. Reach out to sales@gpltech.com to schedule your free consultation today.

 

Scroll to Top