MQM9700-NS2R 400G NDR InfiniBand Smart Switch – 64 Ports On-Board Subnet Manager

Hong Kong Starsurge Group presents the NVIDIA Quantum-2 MQM9700-NS2R — a high-performance, fully managed 400Gb/s InfiniBand switch with reverse airflow (connector-to-power). As part of the QM9700 series, this switch provides 32 OSFP ports supporting 64× 400Gb/s connections (or up to 128 ports at 200Gb/s via port-split technology). With a landmark 51.2 terabits per second bidirectional throughput and over 66.5 billion packets per second capacity, the MQM9700-NS2R includes an on-board Subnet Manager that enables simple out-of-the-box bring-up for up to 2,000 nodes, making it ideal for research institutions, finance, manufacturing, and enterprise AI clusters demanding low latency, lossless fabric, and simplified management.

• Unprecedented Port Density: 64 ports of 400Gb/s NDR InfiniBand in 1U, non-blocking architecture.
• Double-Density 200Gb/s Mode: Supports up to 128 ports of 200Gb/s using NVIDIA port-split technology, reducing TCO for leaf-spine topologies.
• Integrated Subnet Management: On-board Subnet Manager manages up to 2,000 nodes out-of-the-box; accessible via CLI, WebUI, SNMP, or JSON interfaces.
• In-Network Computing: SHARPv3 (Scalable Hierarchical Aggregation and Reduction Protocol) delivers 32x higher AI acceleration compared to prior generation.
• Advanced Fabric Services: RDMA, adaptive routing, congestion control, enhanced VL mapping, and self-healing network capabilities.
• Reverse Airflow (C2P): Connector-to-Power airflow direction ideal for data center layouts with cold aisle on the port side.
• Redundancy & Reliability: 1+1 redundant hot-swappable power supplies, hot-swappable fan units.
• Backward Compatible: Supports previous InfiniBand generations (HDR, EDR, FDR).

Built on the NVIDIA Quantum-2 ASIC, the MQM9700-NS2R leverages state-of-the-art 400Gb/s SerDes technology. It incorporates Remote Direct Memory Access (RDMA) for low CPU overhead and high throughput, adaptive routing to avoid fabric hotspots, and NVIDIA SHARPv3 for in-network reductions, dramatically accelerating collective operations in MPI and AI frameworks. The integrated Subnet Manager runs MLNX-OS software, providing full chassis management through multiple interfaces (CLI, WebUI, SNMP, JSON), reducing operational complexity and enabling rapid deployment.

With support for multiple topologies — including Fat Tree, SlimFly, DragonFly+, and multi-dimensional Torus — the MQM9700-NS2R enables architects to build cost-effective, highly resilient networks for next-generation supercomputing while maintaining simple management.

• AI & Machine Learning Clusters: High-speed GPU interconnects for large language model training and inference with integrated subnet management.
• High-Performance Computing (HPC): Research simulations, weather modeling, genomics, and quantum computing labs requiring self-healing fabrics.
• Hyperscale Data Centers: Spine-leaf architectures with massive east-west traffic and reverse airflow compatibility.
• Enterprise Cloud & Finance: Low-latency trading systems and real-time analytics where managed simplicity matters.
• Government & Education: National labs, university supercomputing centers needing deterministic performance and easy out-of-box setup.

Fully interoperable with NVIDIA InfiniBand ecosystem: ConnectX-6/7 adapters, LinkX cables, and Quantum-2 series switches. Backwards compatible with HDR, EDR, and FDR devices. Supports major Linux distributions (RHEL, Ubuntu, Rocky Linux) and Windows Server with appropriate drivers. The on-board Subnet Manager is compatible with NVIDIA UFM for enhanced telemetry and larger fabrics.

For C2P airflow (connector-to-power), cold air enters from the OSFP connector side and exhausts through the power supply side. Verify your data center cooling layout before ordering. MQM9700-NS2R is the ideal managed choice for racks requiring reverse airflow.

• Integrated Management: On-board Subnet Manager reduces external dependencies, lowers TCO for clusters up to 2,000 nodes.
• Reverse Airflow Flexibility: C2P cooling matches specific data center hot/cold aisle configurations, improving thermal efficiency.
• Highest Radix in 1U: 64x 400G ports minimizes switch tiers, lowers latency, and simplifies cabling.
• Energy Efficient: Gold+ power supplies and smart fan control reduce OPEX.
• Future-Ready Scalability: Support for SHARPv3 and adaptive routing ensures investment protection for next-gen AI frameworks.
• Easy Provisioning: WebUI, CLI, SNMP, and JSON interfaces enable automation and rapid deployment.

Hong Kong Starsurge Group provides end-to-end support including pre-sales consulting, integration services, and global logistics. Our technical team offers deployment assistance, RMA services, and extended warranty options. For volume orders, we deliver customized cabling and configuration validation. Multilingual support available for APAC, EMEA, and Americas regions.

About Hong Kong Starsurge Group Co., Limited

Founded in 2008, Starsurge is a technology-driven provider of network hardware, IT services, and system integration solutions. We serve government, healthcare, manufacturing, education, finance, and enterprise clients worldwide. With an experienced technical and sales team, Starsurge delivers tailored networking infrastructure — including IoT solutions, custom software development, and global logistics. Our customer-first approach ensures reliable quality, responsive service, and scalable network designs. As an authorized channel partner for leading brands, we bridge cutting-edge technology with real-world deployment success.

• Whitepaper: "Scaling AI Fabrics with NVIDIA Quantum-2 Managed Switches" (available on request)
• Deployment Guide: Fat Tree vs. DragonFly+ with QM9700 Series
• Airflow Best Practices: Choosing P2C vs C2P for Data Center Cooling
• Compatibility List: NVIDIA Certified OSFP Cables for NDR