OCtoPus 3

The new generation of multinode CPU solution

OCtoPus 3 has been designed to further improve Data Center Energy Efficiency.

This range is specifically designed to execute complex workloads with its high sustainability.

You will get the highest possible performance for your operations.

OCtoPus 3 is our open hardware which is evolutive, efficient, and flexible to improve your IT infrastructure while providing cost reduction on CAPEX (-25%) and OPEX (-40% with immersion cooling).

Design Principles

  • • 1 OpenU 3 nodes: high level of modularity
  • • Up to 2 CPUs per node
  • • Up to 96 nodes per rack
  • • A sustainable and evolutive solution
  • • Six configurations available

Design Principles

  • • Up to 2 CPU per node (Intel ® or AMD)
  • • 1 OpenU 3 nodes Modular allows high sustainability
  • (up to 96 nodes per rack)
  • • 21’’ server form factor with drawer concept
  • • Up to 2x10GbE (RJ45) (OCtoPus 3R, 3H, 3SW, 3SP)
  • • Up to 2x10GbE (SFP+) (OCtoPus 3E)
  • • Up to 3x HDD/SSD SATA 6Gbps 3.5’’ or 2.5’’
  • • Up to 2x M.2 NVMe
  • • Server Management : IPMI 2.0 and iKVM
  • • Air Cooling, Direct Liquid Cooling or Immersion
  • configuration available
  • • No internal fan, 12V busbar power supply


  • • Big Data
  • • Cloud
  • • HPC

Key Benefits

  • • Maximized performance & density
  • • Easy maintenance
  • • Increased reliability
  • • Sustainability


Six configurations available

2CRSi rack server design for air cooling

With an all-in-one solution based on the OCP community standards and our unique server rack designs, we optimized CAPEX and OPEX as much as possible.


Say hello to the OctoRack 42SL

Rack Front View

A unique rack design


By removing the fans from the servers and by mutualizing them at the back of the rack, we were able to reduce power consumption and include a control manager that enables to control the fan speed according to the temperature of the servers.

Optimized air cooling increases the air flow and heat dissipation inside the servers, so space is optimized for component management.

The failure rate and intervention time are reduced thanks to the centralized power supply of the servers by busbars and shared ventilation.

Rack Rear View