NVIDIA MSB7800-ES2F 100Gb/s InfiniBand Switch, 36 QSFP28 Ports, 1U

The MSB7800-ES2F is a 1U, 36-port 100Gb/s InfiniBand switch from the NVIDIA SB7800 fixed-configuration family. It delivers up to 7.2 Tb/s aggregate bidirectional throughput, ultra-low port latency, redundant hot-swappable power supplies and fans, and MLNX-OS management for HPC, AI and scale-out data center fabrics.

B2B supply and project configuration support provided by Hong Kong Starsurge Group Co., Limited. Availability, lead time and pricing are confirmed per order.

Product Overview

The NVIDIA MSB7800-ES2F is part of the SB7800 series of 100Gb/s InfiniBand fixed-configuration switches. It provides 36 QSFP28 ports in a standard 1U chassis, with an aggregated bidirectional throughput of 7.2 Tb/s and ultra-low port latency. The switch is internally managed through MLNX-OS running on a dual-core x86 CPU, and can be paired with NVIDIA UFM platforms for fabric-wide provisioning, monitoring, telemetry and preventive troubleshooting.

The platform is designed for performance, serviceability, energy savings and high availability. It uses 1+1 redundant hot-swappable power supplies and hot-swappable fans, with front-to-rear airflow in the ES2F ordering option. ATIS weighted power consumption can be as low as 122 W for a fully populated system, with further reduction when not all ports are fully utilized.

As a top-of-rack leaf connectivity solution, the MSB7800-ES2F supports small to extremely large cluster topologies. It is relevant for HPC, AI, research, government, healthcare, manufacturing, education, finance and enterprise environments that need reliable, scalable and manageable InfiniBand infrastructure.

Key Features

• 36 × 100Gb/s QSFP28 ports in a 1U non-blocking switch, with aggregate data throughput up to 7.2 Tb/s.
• Ultra-low port latency to support application performance and scalability in HPC and AI workloads.
• NVIDIA SHARP technology support for in-network aggregation and reduction, offloading collective MPI operations from the CPU to the switch network.
• MLNX-OS management on a dual-core Intel Celeron 1047UE x86 CPU, with 4GB DDR3 RAM and 16GB SSD.
• Redundant 1+1 hot-swappable power supplies, 80 Gold+ and ENERGY STAR certified, with 100–127VAC and 200–240VAC input ranges.
• Hot-swappable fans with front-to-rear airflow (P2C) in the ES2F model, plus redundant power and fan design for maximal thermal protection.
• Power-optimized design with ATIS weighted power consumption as low as 122 W for a fully populated system.
• UFM-ready fabric management when paired with NVIDIA UFM platforms for real-time telemetry, AI-powered analytics and preventive maintenance.

Technology

In-Network Acceleration with NVIDIA SHARP

High-performance computing workloads require intelligent interconnect solutions to reach the highest possible application performance. NVIDIA Scalable Hierarchical Aggregation and Reduction Protocol (SHARP) improves MPI operation performance by offloading collective operations from the CPU to the switch network. This reduces the amount of data traversing the network as aggregation nodes are reached and dramatically reduces MPI operation time, freeing CPU resources for computation rather than communication processing.

100Gb/s InfiniBand and Scale-Out Topologies

The MSB7800-ES2F provides 36 ports of 100Gb/s InfiniBand per port in a standard 1U chassis. A single switch carries an aggregated bidirectional throughput of 7.2 Tb/s with ultra-low port latency. As a top-of-rack leaf connectivity solution, it supports a variety of topologies for small to extremely large clusters. The internally managed SB7800 and externally managed SB7890 fixed-configuration switches provide flexibility across deployment models.

Management with MLNX-OS and NVIDIA UFM

The dual-core x86 CPU runs MLNX-OS software packages for complete chassis management of firmware, power supplies, fans and ports. The switch can be coupled with NVIDIA UFM platforms, which combine enhanced real-time network telemetry with AI-powered cyber intelligence and analytics. UFM solutions help operators provision, monitor, manage and preventatively troubleshoot InfiniBand data center fabric, minimizing cluster downtime.

Typical Deployments

• HPC clusters: top-of-rack leaf connectivity for compute clusters that need high bandwidth and low latency.
• AI and data-intensive workloads: fabrics that must move large datasets generated by HPC and AI technologies.
• Research and government: scalable InfiniBand infrastructure for laboratories, public sector and defense-adjacent research environments.
• Healthcare and life sciences: high-throughput networking for imaging, genomics and analytics clusters.
• Manufacturing and education: reliable fabric for simulation, design and campus research computing.
• Finance and enterprise: low-latency scale-out fabrics for risk, trading analytics and enterprise AI platforms.

Compatibility

The MSB7800-ES2F is an NVIDIA SB7800-series 100Gb/s InfiniBand switch with 36 QSFP28 ports. It is designed for InfiniBand fabrics and management through MLNX-OS, with optional NVIDIA UFM integration. Compatibility depends on the specific fabric design, cable plant, transceivers and host channel adapters in your environment.

• QSFP28 100Gb/s InfiniBand ports; exact cable and transceiver compatibility should be confirmed against the NVIDIA compatibility list for your deployment.
• Management ports: 1 × RJ45, 1 × RS232 console, 1 × USB.
• MLNX-OS internally managed software; NVIDIA UFM platforms supported for fabric management.
• Power input: 100–127VAC and 200–240VAC; confirm site power and airflow orientation before ordering.

For fabric-level compatibility, transceiver qualification and firmware matching, please confirm with our technical team before ordering.

Specification Table

Source: MSB7800 series datasheet provided by the customer. Parameters not publicly specified in the source are marked accordingly.

Selection Guide

The SB7800 family includes several 1U 36-port 100Gb/s InfiniBand models. Use the table below to match airflow direction and switch versus router configuration to your rack design and fabric role.

Selection note: confirm airflow direction against your rack hot/cold aisle design, and confirm switch versus router role against your fabric architecture before ordering.

Advantages

• High port density in 1U: 36 × 100Gb/s QSFP28 ports and 7.2 Tb/s aggregate throughput in a standard 1U chassis.
• Application performance: ultra-low port latency and SHARP in-network acceleration reduce MPI operation time and free CPU resources.
• Operational efficiency: redundant hot-swappable power and fans, 80 Gold+ and ENERGY STAR certified power supplies, and ATIS weighted power consumption as low as 122 W.
• Fabric visibility: MLNX-OS chassis management plus optional NVIDIA UFM for real-time telemetry and AI-powered analytics.
• Deployment flexibility: multiple ordering options for airflow direction and switch/router role within the SB7800 family.

Service & Support

Hong Kong Starsurge Group Co., Limited supports this product with an experienced sales and technical team. Support scope can include pre-sales configuration review, airflow and power confirmation, compatibility checks, and order follow-up.

• Pre-sales selection support for port count, airflow, switch/router role and management model.
• Compatibility confirmation for transceivers, cables and host adapters before ordering.
• Order configuration and lead-time coordination.
• Multilingual support and global delivery available through Starsurge.

FAQ

What is the MSB7800-ES2F?

It is a 1U, 36-port 100Gb/s InfiniBand switch in the NVIDIA SB7800 family, with QSFP28 ports, 7.2 Tb/s aggregate throughput, MLNX-OS management and front-to-rear (P2C) airflow.

How many ports and what throughput does it provide?

36 × 100Gb/s non-blocking ports with aggregate bidirectional throughput up to 7.2 Tb/s.

Is it internally or externally managed?

The MSB7800-ES2F is internally managed through MLNX-OS running on a dual-core x86 CPU. It can also be coupled with NVIDIA UFM platforms for fabric-level management.

What is the difference between ES2F and ES2R?

The main difference is airflow direction: ES2F uses front-to-rear (P2C) airflow, while ES2R uses rear-to-front (C2P) airflow. Confirm against your rack design before ordering.

Does it support NVIDIA SHARP?

The SB7800 platform supports NVIDIA SHARP technology for in-network aggregation and reduction, offloading collective MPI operations from the CPU to the switch network.

What power and cooling does it require?

It uses 1+1 redundant hot-swappable power supplies with 100–127VAC and 200–240VAC input ranges, and hot-swappable fans. ATIS weighted power consumption can be as low as 122 W for a fully populated system.

Can you confirm compatibility with my cables and transceivers?

Yes. Please share your planned QSFP28 optics, cables and host adapter models, and our technical team will confirm compatibility before ordering.

Precautions

• Confirm airflow direction (P2C for ES2F) against your rack hot/cold aisle design before ordering.
• Confirm whether your fabric requires a switch or router role; the MSB7880 models are the router variants.
• Verify power input range and redundancy requirements at the installation site.
• Validate QSFP28 transceiver and cable compatibility against your NVIDIA fabric plan.
• Specifications are based on the provided MSB7800 series datasheet. Parameters not publicly specified in that source are marked “Not publicly specified” or “Please confirm before ordering.”

Company Introduction

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, cables and related networking equipment.

Backed by an experienced sales and technical team, Starsurge supports industries such as government, healthcare, manufacturing, education, finance and enterprise. The company also offers IoT solutions, network management systems, custom software development, multilingual support and global delivery.

With a customer-first approach, Starsurge focuses on reliable quality, responsive service and tailored solutions that help clients build efficient, scalable and dependable network infrastructure.