Driving AI efficiency
in the data center

All in on efficiency
Price / Performance References Resources

Habana partners with Supermicro

to enable customers who need AI performance, efficiency and usability in their on-premises systems. Supermicro offers server options for both Habana Gaudi and Gaudi2 Training and Habana First-generation Inference accelerators.
supermmicro logo
gaudi2 logo

Training Advanced

With a leap in deep learning training performance while preserving cost-efficiency.
Supermicro Gaudi®2 AI Training Server SYS-820GH-TNR2: The first commercial implementation of the new Gaudi2 processor.

  • Featuring eight Habana® Gaudi®2 AI Training Processors
  • Dual 3rd Gen Intel® Xeon® Scalable processors
  • Expanded networking capacity with 24x 100 Gb RoCE ports integrated onto every Gaudi2
  • 700 GB/second scale within the server and 2.4TB/second scale out
  • Ease of system build or migration with Habana® SynapseAI® Software Stack

Supermicro Gaudi®2
AI Training Server

Supermicro server
Gaudi grey logo

Training

Designed to deliver AI training efficiency, usability and scalability:
Supermicro X12 Gaudi Training Server: SYS 420 GH-TNGR

  • Featuring eight Habana Gaudi HL-205 processors
  • Each Gaudi processor contains ten 100 Gb RDMA over Converged Ethernet (RoCE) ports
  • Dual-socket 3rd Gen Intel ® Xeon® Scalable Processors
  • Two PCIe Gen 4 switches
  • Four hot swappable NVMe/SATA hybrid hard drives
  • Scale out with 24x 100 Gb RoCE ports (six QSFP-DDs)
  • 32 DIMMs; up to 8TB of DDR4-3200MHz

Supermicro X12 Gaudi AI Training Server

SuperMicro server

Inference

Supermicro 4U Server SYS-4029GP-TRT2 featuring Habana First-generation Inference accelerators delivers superior AI inference efficiency and latency

  • Up to eight Habana First-generation Inference accelerators
  • Dual-socket 2nd Gen Intel® Xeon® Scalable Processors
  • 24 DIMMs; up to 6TB 3DS ECC DDR4-2933MHz
  • Single Root System for excellent AI processor peer-to-peer communication with sufficient CPU to AI processor throughput
  • 11 total PCIe Gen 3 slots
  • Hot-swappable high RPM fans

Supermicro AI Training and Inference systems

based on Habana processors are supported by Habana SynapseAI® Software Suite, offering:
• Integration of TensorFlow and PyTorch frameworks
• Popular models for computer vision and natural language processing applications
• Habana Kernel Library and enablement of Custom Kernel Development

Pytorch
tensorflow
Habana Framework

Habana offers extensive
support for data scientists,
developers, and IT and Systems
Administrators with its:

Habana Developer Site: featuring documentation, customer-available software, “how to” content, community Forum

Habana GitHub: providing reference models, set-up and install instructions, snapshot scripts for analysis and debug, custom kernel examples,
roadmap, and “issues” sub-site tracking to dos, bugs, feature requests and more

developer site screen
Learn more about Habana Gaudi and First-generation Inference systems for on-premises implementation.
Learn More