Visual effects and animation studios can streamline their rendering processes using AWS Deadline Cloud, a fully managed service designed for efficient render management. This service simplifies the handling of compute resources for teams working on computer-generated graphics and visual effects.
One of the key features available for service-managed fleets on Deadline Cloud is Wait and Save, which offers significant discounts on CPU rendering costs by allowing flexibility in job scheduling. This article provides a guide on how to set up Wait and Save to optimize rendering expenses.
Understanding Instance Types
Deadline Cloud service-managed fleets utilize two main types of Amazon Elastic Compute Cloud (Amazon EC2) instances:
- On-Demand Instances: These provide reliable compute capacity for workloads requiring consistent processing.
- Spot Instances: These allow users to leverage unused EC2 capacity at discounted rates, potentially saving up to 90%. However, these instances may be interrupted to accommodate On-Demand requests.
How Wait and Save Works
Wait and Save capitalizes on excess CPU Spot capacity, offering lower compute rates for jobs that can be flexible with their start times. By scheduling jobs during periods of higher Spot capacity availability, studios can achieve substantial cost savings while maintaining rendering quality.
Optimizing Job Submission
To maximize the benefits of Wait and Save, consider the following:
- Submit jobs during off-peak hours.
- Select different AWS Regions to access more Wait and Save capacity.
Setting up dedicated queues for Wait and Save fleets can further streamline this process.
Setting Up Wait and Save Fleets
Before beginning the setup, ensure that Wait and Save aligns with your workload requirements and that you have the necessary AWS resources. Here’s a brief overview of the setup process:
- Create a dedicated queue for Wait and Save fleets.
- Establish a Wait and Save fleet by selecting the appropriate service-managed fleet type.
- Associate the Wait and Save queue with the fleet.
Cost Analysis Example
To illustrate the cost savings, a 200-frame turntable render was submitted using Autodesk Maya through the Deadline Cloud submitter. The results showed that:
| Market Type | Cost Savings |
|---|---|
| On-Demand | 0% |
| Spot | 22% |
| Wait and Save | 92% |
Integrating with Existing Fleets
For studios requiring higher capacity while still aiming for cost efficiency, Wait and Save can be integrated with existing Spot and On-Demand fleets. This hybrid approach ensures that jobs are processed without delays while optimizing costs.
Automating Capacity Management
Utilizing AWS CloudFormation, studios can automate the adjustment of Spot fleet capacity based on real-time Wait and Save worker availability. This setup allows for efficient scaling of compute resources without sacrificing rendering performance.
Conclusion
Wait and Save presents a valuable opportunity for studios looking to reduce CPU rendering costs significantly. By implementing dedicated queues or hybrid setups, studios can enhance their rendering capabilities while adhering to budget constraints. For further assistance, contacting an AWS representative can provide additional insights into optimizing rendering workflows.
