Documentation Center

Recommended system requirements

The recommended requirements refer to the setup needed to run theLanguage Weaver Edge application and the language pairs. These requirements are proportional to the number of language pairs to be installed and the desired translation performance.

You can find below the recommended system setup to run Language Weaver Edge in two different scenarios: up to 4 language pairs, and 5 or more. If your installation includes more than 10 language pairs, please contact your RWS account representative for support on the proper sizing of your deployment.

ComponentUp to 4 language pairs5 or more language pairs
Controller
  • 8 CPU cores
  • 16 GB RAM
  • 200 GB HD
  • 8 CPU cores
  • 32 GB RAM
  • 200 GB HD
Job enginePer job engine:
  • 4 CPU cores
  • 16 GB RAM

Recommended 2 job engines

Per job engine:
  • 8 CPU cores
  • 32 GB RAM

Minimum 2 job engines

Translation enginePer engine and PU:
  • 2 CPU cores
  • 4 GB RAM
  • 4 GB HD
Per engine and PU:
  • 2 CPU cores
  • 4 GB RAM
  • 4 GB HD
Audio transcription enginePer engine and PU:
  • On GPU:
    • 2 CPU cores
    • 5 GB RAM
  • On CPU:
    • 16 CPU cores
    • 16 GB RAM
Per engine and PU:
  • On GPU:
    • 2 CPU cores
    • 5 GB RAM
  • On CPU:
    • 16 CPU cores
    • 16 GB RAM
Training engine
  • 1 GPU with 16 GB RAM
    • If CPU only: 4 CPU with 32 GB RAM
  • Minimum of 100 GB HD available per training
  • 4GB HD per adaptable LP
  • 1 GPU with 16 GB RAM
    • If CPU only: 4 CPU with 32 GB RAM
  • Minimum of 100 GB HD available per training
  • 4GB HD per adaptable LP

For Edge-Cloud deployments, we recommend dedicating one host for the controller and two for the job engine and translation engines.

Single machine deployments should be limited to a proof of concept, to low volumes, and to a reduced number of language pairs.