The NVIDIA Quantum MQM8790‑HS2F is an unmanaged 200G InfiniBand smart switch designed for large‑scale data center deployments where centralized fabric management is preferred. Part of the NVIDIA Quantum QM8700 series, this switch delivers forty 200Gb/s ports in a compact 1U form factor with 16 Tb/s aggregate non‑blocking throughput and sub‑130ns cut‑through latency. Unlike its managed counterparts, the MQM8790‑HS2F is optimized for external management via NVIDIA Unified Fabric Manager (UFM™), enabling data center operators to efficiently provision, monitor, troubleshoot, and maintain modern data center fabrics at scale.

Each 200Gb/s QSFP56 port can be split into two independent 100Gb/s ports, providing up to 80 ports of 100Gb/s connectivity — ideal for double‑density top‑of‑rack deployments. The MQM8790‑HS2F features P2C (port‑to‑power) airflow, dual redundant power supplies, and full support for NVIDIA SHARP™ in‑network computing acceleration.

• 200Gb/s InfiniBand per port – forty QSFP56 ports supporting 200G or 100G split modes, non‑blocking architecture.
• Unmanaged SKU for External Control – No on‑board Subnet Manager; designed for centralized management via NVIDIA UFM™ platform.
• In‑Network Computing Acceleration – NVIDIA SHARP™ technology enables in‑switch data aggregation, reducing MPI, NCCL, and SHMEM communication time by orders of magnitude.
• High Radix & Split Capability – Convert 40x 200G ports into 80x 100G ports for double‑density topologies without extra switches.
• Advanced Congestion Management – Adaptive routing, static routing, and quality of service (QoS) to eliminate hot spots and maximize effective fabric bandwidth.
• Redundant & Hot‑Swappable PSU – 1+1 redundant power, 80 Plus Gold certified, ENERGY STAR compliant, with power optimization on partial port usage.
• UFM Ready – Seamless integration with NVIDIA Unified Fabric Manager for advanced telemetry, predictive analytics, and automated fabric orchestration.
• Backward Compatible – Seamless interoperability with previous InfiniBand generations (EDR, FDR).

NVIDIA Quantum switches embed scalable hierarchical aggregation and reduction protocol (SHARP) engines directly in the silicon. Data traversing the switch can be processed — aggregated, reduced, or broadcast — without multiple round‑trips to server endpoints. This dramatically accelerates collective operations like all‑reduce, barrier, and broadcast, which are critical for deep learning frameworks (TensorFlow, PyTorch via NCCL) and MPI‑based HPC simulations. The result is up to 10x performance gains for communication‑intensive workloads and reduced CPU overhead, freeing compute resources for actual application processing.

The MQM8790‑HS2F also supports adaptive routing and congestion control algorithms that automatically balance traffic across multiple paths, delivering near‑line‑rate throughput even under high contention.

• Large‑Scale AI & ML Clusters – GPU‑based systems requiring centralized fabric management and telemetry across hundreds or thousands of nodes.
• High‑Performance Computing (HPC) – Research labs, national labs, and universities where external fabric managers provide enhanced monitoring and automation.
• Hyperscale Data Centers – Fat‑tree, DragonFly+, and multi‑dimensional torus topologies managed through UFM for maximum efficiency.
• Enterprise & Cloud Service Providers – Environments requiring unified control plane across multiple switch fabrics.
• Top‑of‑Rack (ToR) with Centralized Management – Double‑density 100Gb/s per server connectivity with fabric‑wide visibility.

The MQM8790‑HS2F works seamlessly with NVIDIA ConnectX‑6, ConnectX‑7, and BlueField DPU adapters, supporting both InfiniBand and mixed fabrics. It is backward compatible with previous InfiniBand speeds (EDR 100Gb/s, FDR 56Gb/s). Fully interoperable with existing NVIDIA Quantum fabric switches. For management, the switch is designed to be controlled via NVIDIA Unified Fabric Manager (UFM) platform, which provides comprehensive fabric provisioning, monitoring, and predictive troubleshooting. Operating system support includes major Linux distributions (RHEL, Ubuntu, Rocky Linux) and NVIDIA certified GPU servers.

For environments requiring centralized fabric management across hundreds or thousands of switches, the MQM8790‑HS2F offers a cost‑effective, unmanaged building block optimized for NVIDIA UFM orchestration.

• Centralized Fabric Management – Leverage NVIDIA UFM for unified visibility, automation, and predictive analytics across the entire fabric.
• Superior ROI – Reduce capital expenditure with double‑density 100G port capacity and lower switch count for large fabrics.
• Energy Efficient – Dynamic power scaling based on port utilization, lowering operational costs.
• SHARP™ Acceleration – Up to 10x faster collective communications without consuming host CPU cycles.
• Scalable Topologies – Native support for Fat Tree, DragonFly+, and Torus to future‑proof data center growth.
• Proven Ecosystem – Backed by NVIDIA cumulative software stack and 24/7 partner support.

Starsurge Group provides end‑to‑end lifecycle services for NVIDIA Quantum switches, including pre‑sales architecture consulting, proof‑of‑concept testing, and global logistics. Our experienced technical team offers remote troubleshooting, firmware upgrades, and RMA coordination. For UFM deployments, we offer professional services for platform configuration and integration. Warranty extension options and 24x7 priority support available upon request. Multilingual support for EMEA, Americas, and APAC regions ensures rapid response for mission‑critical deployments.

Hong Kong Starsurge Group Co., Limited is a technology‑driven provider of network hardware, IT services, and system integration solutions. Founded in 2008, the company serves customers worldwide with products including network switches, NICs, wireless access points, controllers, cabling, and infrastructure equipment. Backed by an experienced sales and technical team, Starsurge supports industries such as government, healthcare, manufacturing, education, finance, and enterprise.

With a customer‑first approach, Starsurge focuses on reliable quality, responsive service, and tailored solutions. As an authorized partner for leading networking brands, we deliver global logistics, custom software development, and multilingual support — helping clients build efficient, scalable, and dependable network infrastructure.

• NVIDIA Unified Fabric Manager (UFM) Platform Licenses
• NVIDIA Quantum QM9700 Series (NDR 400G InfiniBand)
• NVIDIA ConnectX‑6 VPI Adapter Cards (100Gb/s Dual‑port)
• NVIDIA BlueField‑3 DPU for infrastructure acceleration
• Starsurge Custom Rack Integration Kits & QSFP56 Cables (passive/active)