Mellanox

Showing 241–243 of 243 results

• Inquiry

  Mellanox MQM8790-HS2R Quantum HDR InfiniBand Switch – 40 Port – QSFP56 – C2P Airflow – Rail Kit

  $Quote1 Piece (MOQ) Minimum Order Quantity Inquiry

  **Mellanox MQM8790-HS2R Quantum HDR InfiniBand Switch – 40 Port – QSFP56 – C2P Airflow – Rail Kit**
  A high-performance 40-port InfiniBand switch delivering 200Gb/s per port via QSFP56, optimized for ultra-low latency and massive data throughput in HPC, AI, and cloud environments. Features C2P airflow for efficient cooling, adaptive routing for load balancing, and seamless scalability. Includes a rail kit for easy rack installation. Ideal for demanding workloads requiring unmatched speed, reliability, and energy efficiency in data-intensive applications.

  Learn more
• Inquiry

  Mellanox MQM9790-NS2R Quantum 2 NDR InfiniBand Switch

  $Quote1 Piece (MOQ) Minimum Order Quantity Inquiry

  The Mellanox MQM9790-NS2R Quantum 2 NDR InfiniBand Switch delivers ultra-high performance with 400Gb/s per port, 64-port density, and ultra-low latency for AI, HPC, and cloud workloads. Key features include seamless scalability (14,000+ nodes in a single subnet), adaptive routing, advanced congestion control, and end-to-end encryption for robust security. Its energy-efficient design reduces operational costs, while backward compatibility ensures integration with existing infrastructure. USP: Unmatched speed, scalability, and reliability for data-intensive applications, positioning it as the premier choice for next-gen data centers.

  Learn more
• Inquiry

  Mellanox MQM9700-NS2R NVIDIA Quantum-2 QM9700 switch 32 ports smart rack-mountable

  $Quote1 Piece (MOQ) Minimum Order Quantity Inquiry

  **NVIDIA Quantum-2 QM9700 Switch (MQM9700-NS2R)**
  A high-performance, 32-port 400GbE InfiniBand switch designed for AI, HPC, and hyperscale data centers. Delivers ultra-low latency, 12.8 Tbps throughput, and scalability for demanding workloads. Features NVIDIA’s Quantum-2 platform with in-network computing (SHARP) to accelerate AI training and reduce data movement. Optimized for GPU-driven environments with GPUDirect support, enhancing efficiency for distributed systems. Rack-mountable, energy-efficient design with advanced security and cloud-native orchestration. Uniquely integrates with NVIDIA’s full stack (GPUs, DPUs, software) for end-to-end AI infrastructure optimization. Ideal for accelerating large-scale AI clusters and data-intensive applications.

  Learn more