VPU Deployment Playbook

Overview

Start Here

Architectures

Explore

Test

Migrate

Scale

Cloud

On-Prem

Hybrid

Edge Distributed

ENCODING architecture

Cloud.

Run high-density video encoding/decoding/transcoding in the cloud using Quadra VPUs without changing ingest, playback, or orchestration workflows.

Use this architecture when:

  • Cloud-hosted or cloud-adjacent video pipelines
  • CPU or GPU encoding costs are constraining scale
  • AV1, HEVC, or high-density H.264 in production
  • Need predictable performance and linear scaling


This architecture is optimized for cost control, density, and operational predictability.

What changes

  • Video compute moves from CPU/GPU to VPUs
  • Encoding density per instance increases
  • Cost per stream decreases at scale

What doesn’t

  • Ingest sources and output destinations
  • FFmpeg/GStreamer-based workflows and codecs
  • Cloud provider, region, or account ownership

VPU access

  • Quadra VPUs sit inside the encoding layer
  • Handle encode/transcode only
  • Eliminate CPU/GPU contention
  • Scale linearly with additional instances
  • Available via: Akamai, Oracle, CDN77 and i3D.

Scaling model

  • Horizontal scale via VPU-enabled instances
  • Deterministic performance per instance
  • No GPU sharing or noisy-neighbor effects

Integration support

  • Supported cloud environment (customer-owned)
  • FFMPEG, GStreamer
  • VPU-enabled instances
  • Existing ingest and output endpoints
  • Orchestration layer (Bitstreams or existing scheduler)
  • Link to Ecosystem

Outcome

Higher encoding density. Lower cost per stream. Predictable performance using your existing infrastructure.

Supported by the VPU Ecosystem, partners operating this architecture in production today.

Let’s continue.

< Go back
Next >