NVIDIA Mellanox MQM9790-NS2F InfiniBand Switch｜Product Launch and Key Feature Overview

NVIDIA Mellanox MQM9790-NS2F InfiniBand Switch｜Product Launch and Key Feature Overview

As AI training clusters, high-performance computing (HPC), and hyperscale storage networks continue their migration to 400Gb/s fabrics, network engineers and architects face a fundamental challenge: not merely adding bandwidth, but preserving low latency and lossless transport at high port density. The NVIDIA Mellanox MQM9790-NS2F addresses this challenge directly, offering a clear upgrade path for next-generation data center networks built on NDR InfiniBand.

Background: Why the MQM9790-NS2F Matters

Legacy 200Gb/s HDR fabrics increasingly struggle to sustain large-scale GPU clusters, where collective operations and storage traffic demand deterministic performance. The MQM9790-NS2F InfiniBand switch is positioned to solve this by delivering 400Gb/s NDR throughput per port across a 64-port OSFP configuration. For teams evaluating MQM9790-NS2F specifications, the platform is designed to reduce fabric tiers, simplify cabling, and maintain the ultra-low latency that InfiniBand is known for.

Key Technical Highlights

The NVIDIA Mellanox MQM9790-NS2F is a 400Gb/s NDR 64-port OSFP switch, meaning it supports up to 64 OSFP ports running at NDR speeds. This density allows architects to build flatter, more scalable topologies while preserving non-blocking bandwidth for demanding east-west traffic. Key characteristics include:

• 400Gb/s NDR per port with 64 OSFP ports for high-radix fabric design
• Support for NVIDIA SHARPv3 and collective acceleration to offload MPI operations
• Advanced congestion control and adaptive routing for lossless, low-latency transport
• Compatibility with NVIDIA Quantum-2 based ecosystems and standard InfiniBand management tools
• Designed for AI, HPC, and storage workloads requiring deterministic performance

For those comparing MQM9790-NS2F datasheet details, the switch is part of the Quantum-2 family and is intended to interoperate with existing InfiniBand deployments. Teams seeking a MQM9790-NS2F InfiniBand switch solution should evaluate port count, power, and cooling requirements alongside fabric topology and cable reach.

Applications and Deployment Scenarios

The MQM9790-NS2F InfiniBand switch is suited to AI factories, national research labs, and enterprise HPC clusters where GPU utilization and job completion time are directly tied to network efficiency. It also fits storage-intensive environments that need a converged, low-latency fabric for parallel file systems and NVMe over Fabrics. When evaluating MQM9790-NS2F compatible infrastructure, architects should confirm that host adapters, cables, and management software align with NDR and Quantum-2 requirements.

Procurement teams often ask about MQM9790-NS2F price and MQM9790-NS2F for sale channels. Pricing varies by configuration, support level, and region, so it is best to consult authorized distributors and NVIDIA enterprise representatives for current quotes and availability. For detailed technical documentation, the MQM9790-NS2F datasheet and MQM9790-NS2F specifications should be reviewed against specific workload profiles.

Summary

As a 400Gb/s NDR 64-port OSFP InfiniBand switch, the NVIDIA Mellanox MQM9790-NS2F delivers the port density, low latency, and collective acceleration needed for AI and HPC at scale. For network engineers and IT managers planning a fabric refresh, it represents a practical, standards-aligned building block for next-generation InfiniBand networks.