HPE NVIDIA InfiniBand XDR 2x36-port OSFP Managed Connector to Power Airflow Switch

  • Brand: HPE
  • Category:
  • SKU: P79109-B21
  • EAN: 0190017789606
NVIDIA InfiniBand XDR 2x36-port OSFP Managed Connector to Power Airflow Switch

Trade Prices

Distributor Product SKU Stock Updated Price
EET

Nvidi Infinind Xdr

 P79109-B21 Register free see stock

Reasons To Buy

Impressive Computing and Networking Performance

The fourth generation NVIDIA SHARP offers non-blocking switching capacity of 115.2 Tb/s on Q3400-RA switch.
This scalable solution incorporates modern technologies such as RDMA, adaptive routing, and NVIDIA SHARP.

Streamlined and Data-Driven

Highest scale for AI: NVIDIA Quantum-X800 switches enable faster speeds and higher scalability for AI compute fabrics. A two-tier Quantum-X800 fat-tree topology can support over 10,000 800 Gb/s host connections.
Quantum-X800 switches support SHARP for efficient offloading of compute operations to the network, helping to boost performance. The fourth generation of SHARP adds support for FP8 precision and new collective operations, such as ReduceScatter and ScatterGather.

Enhanced Power Efficiency and Resiliency

NVIDIA’s new Quantum-X Photonics CPO switch reduces total power consumption, latency, and total cost of ownership (TCO) while enhancing resiliency and serviceability.

Enhanced Software Operations

Switches include NVIDIA NVOS operating system for comprehensive chassis management and system configuration.
NVOS supports various interfaces, including CLI, REST APIs, SNMP, and gRPC Network Management Interface (gNMI) telemetry.

Description

Do your high performance computing (HPC) and artificial intelligence (AI) applications need ever-increasing networking performance to address problems with massive datasets, complex and highly-parallelized algorithms in an extreme-scale system? NVIDIA Quantum-X800 InfiniBand Switches for HPE deliver 800 Gb/s throughput, with ultra-low latency and advanced NVIDIA In-Network Computing, which is essential for handling trillion-parameter-scale generative AI. NVIDIA Quantum-3 family of fixed-con figuration switches revolutionizes the performance, scalability, and efficiency of HPC and AI infrastructures, enabling faster and more effective AI processing and computation. These switches are available in both 4U and 2U systems; also incorporate remote direct-memory access (RDMA), the 4th gen NVIDIA Scalable Hierarchical Aggregation and Reduction Protocol (SHARP)™, adaptive routing, telemetry-based congestion control, self-healing, and Quantum-X Photonics co-packaged optics (CPO) technologies.

What is new supporting point:
Supports faster speeds and higher scalability for AI compute fabrics.
800 Gb/s throughput
144 ports at 800 Gb/s across 72 OSFP cages
Connector to power airflow

Specifications

Design
Compatible productsHPE Compute Mainstream Platforms, see QuickSpecs for more information
System requirements
Compatible productsHPE Compute Mainstream Platforms, see QuickSpecs for more information