Instances are essentially a virtual environment or a virtual computer. An instance is created when you create a Cluster or Production Module in VVCR. You can also create multiple instances within the same cluster for redundancy.
VVCR connects to your AWS EC2 account and acts as a UI to create and work within your virtual computer. You can create your instance in different sizes and in different regions around the world.
AWS limits the number of running Instances per account and region. The limit is determined by the number of Virtual Central Processing Units (vCPUs) the active instance uses.
There are two types of instances, CPU Instances and GPU Instances. CPU Instances always utilize software transcoding while GPU Instances always utilize hardware encoding.
If you need to transcode any VVCR streams or destinations within a CPU instance, remember that two vCPUs are required per transcoded feed.
Here’s a look at some of our most commonly used CPU Instance Types.
| Instance Type | vCPU | Transcoded Feeds |
| t3.medium | 2 | 1 |
| t3.large | 2 | 1 |
| t3.xlarge | 4 | 2 |
| t3.2xlarge | 8 | 4 |
| c5a.large | 2 | 1 |
| c5a.xlarge | 4 | 2 |
| c5a.2xlarge | 8 | 4 |
| c5a.4xlarge | 16 | 8 |
| c5a.8xlarge | 32 | 16 |
GPU Instances utilize hardware encoding and are more efficient than CPU Instances. They can handle more transcoded streams and more graphic and NDI-heavy workflows.
However, the limiting factor of GPU transcoding is bitrate. Each instance can handle up to 4K 60 FPS resolution with up to 150 Mbps.
For example, you can transcode:
- 3 streams at 4k 60 FPS 50 Mbps
- 6 streams at 4k 30 FPS 50 Mbps
G4dn instances use the same type of GPU, so the GPU transcoding capabilities are the same.
Click here to learn more about Instance Type Capacities.
Click here to learn more about vCPUs.