IT teams that need fast, predictable cluster networking with clear control over growth use Nvidia InfiniBand Switches to keep AI, HPC and shared research environments responsive. The range covers managed and externally managed options, so teams can choose whether switch control sits locally or in central fabric tools as the estate scales.
In data centres and shared compute platforms, these switches support low-latency, high-bandwidth traffic between nodes, helping reduce bottlenecks, keep jobs moving and standardise operations across storage, leaf and spine layers without adding unnecessary management overhead.
Comparing multiple platforms? Our experts are available to help.
No commitment needed, no hard sells. Just straightforward technical guidance tailored to your infrastructure.
Full technical specifications are available on each product page.
| Model | Popularity | Deployment | Target Organisation Size | Primary Use Case | Access Ports | Port Configuration | Maximum Port Speed | PoE Budget (W) | Maximum Switching Capacity (Gbps) | Stackable | Operating System / Management Platform | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
Mellanox QM9700 Quantum-2 40-Port NDR 400Gb/s InfiniBand Switch
|
★ ★ ★ | Data Centre | Service Provider | Spine | 64 | 64×QSFP56 | 200G | 0, None | 51200 | Supported | MLNX-OS | View |
Mellanox QM9790 Quantum-2 40-Port NDR 400Gb/s InfiniBand Switch
|
★ ★ ★ | Data Centre | Service Provider | Spine | 64 | 64×QSFP56 | 200G | 0, None | 51200 | Supported | MLNX-OS | View |
Mellanox QM8700 Quantum 40-Port HDR 200Gb/s InfiniBand Switch
|
★ ★ ★ | Data Centre | Service Provider | Spine | 40 | 40×QSFP56 | 200G | 0, None | 16000 | Supported | MLNX-OS | View |
Mellanox QM8790 Quantum 40-Port HDR 200Gb/s InfiniBand Switch
|
★ ★ ★ | Data Centre | Service Provider | Spine | 40 | 40×QSFP56 | 200G | 0, None | 16000 | Supported | MLNX-OS | View |
Data centre teams need low-latency InfiniBand fabrics for AI, HPC and storage clusters, but not every estate is ready for the cost and scale of NDR. This category supports compact HDR deployments for fast cluster performance, or denser NDR fabrics where radix, hop count and cable reduction matter.
Studio and service-provider teams need predictable node-to-node traffic for render, simulation and generative-media jobs. These switches help keep shared farms responsive, with HDR options for busy clusters and NDR options where larger scene data and faster interconnects need more headroom.
Research, genomics and imaging teams need tightly coupled cluster performance for parallel analysis, often within tighter power and budget limits. This category provides low-latency fabrics that support controlled HDR deployments, or centralised NDR fabrics for larger research estates with stronger governance needs.
Quant, risk and AI modelling teams need deterministic interconnect behaviour so distributed runs finish without network delays becoming a constraint. InfiniBand switches in this category support fast cluster fabrics with either local management or central control, depending on how tightly the environment is governed.
AI and HPC development teams need realistic cluster networking for training, benchmarking and distributed code validation before workloads move to production. These switches give shared labs and platform teams the fabric performance to test at scale, with management options that fit either smaller labs or centrally run estates.
Nvidia AI Enterprise, NIM, and AI Workbench give IT and data science teams validated, enterprise-ready platforms to deploy, serve, and manage AI workloads at scale. We help organisations implement and support these software environments so AI infrastructure remains performant, current, and straightforward to operate.
We provide end-to-end services to design, deploy, and optimise GPU-accelerated AI environments. From initial architecture through to integration and scaling, we help infrastructure teams build Nvidia environments that are aligned to workload demands and positioned to grow alongside the business.
As an authorised Nvidia partner, we help organisations maximise uptime and performance across their Nvidia estate. From technical support and software updates through to lifecycle planning, we help IT teams keep AI infrastructure reliable, current, and aligned to long-term operational requirements.
We help organisations get more from their Nvidia investments — from initial architecture through to ongoing optimisation and support. Contact our Nvidia specialists for guidance today.
Backed by decades of expertise in the IT sector, our specialists support every stage of your deployment — from initial selection through to long-term lifecycle management.
Standard Ethernet works well for general business traffic, but AI and HPC workloads create far heavier communication between systems during training, simulation and parallel processing. As clusters grow, latency and congestion can start affecting overall workload efficiency.
Nvidia InfiniBand Switches are designed for high-performance cluster fabrics, helping maintain predictable throughput across AI training environments, HPC platforms and high-performance storage deployments where the network directly affects compute performance.
Not every clustered environment needs 400Gb/s fabric capacity immediately. Departmental AI clusters, research environments and mid-sized HPC deployments often still need low-latency performance without the scale and infrastructure overhead associated with NDR.
The QM8700 and QM8790 provide 40 HDR 200Gb/s ports in a compact 1U design, making them well suited to clustered workloads that need high-performance InfiniBand without moving into larger NDR fabrics.
Internally managed switches can work well in smaller or isolated environments. The challenge appears when multiple AI, HPC or research clusters need consistent operational control across a wider infrastructure estate.
The QM8790 is externally managed for use with centralised fabric management platforms, while the QM8700 includes onboard management. Both provide the same 40 HDR 200Gb/s connectivity, so the decision is mainly operational rather than performance-based.
HDR fabrics remain effective while cluster sizes and traffic demands stay moderate. As environments scale, larger datasets and distributed workloads increase the need for higher bandwidth and flatter fabric designs.
The QM9700 provides 64 NDR 400Gb/s ports in 1U, helping support large AI training clusters, simulation platforms and research computing environments with fewer switching layers and lower infrastructure complexity.
Onboard switch management can suit smaller deployments, but larger clustered environments often need centralised operational control across the wider fabric.
The QM9790 delivers the same Quantum-2 NDR switching class as the QM9700, but is designed for externally managed environments using Nvidia Unified Fabric Manager tools for provisioning, monitoring and fabric maintenance across shared AI and HPC estates.
Smaller AI and HPC environments may continue operating effectively on HDR infrastructure while cluster traffic remains manageable. The move to NDR becomes more relevant as workloads, datasets and inter-node communication demands increase.
The QM9700 and QM9790 provide 64 NDR 400Gb/s ports in 1U, supporting large AI training environments, simulation platforms and research computing estates where consistent node-to-node performance matters at scale.
If these options aren’t the right fit for your environment, we provide a wide portfolio of product series and solutions that may better suit your infrastructure. Explore below, or speak to our team and we’ll help you find the right match.
Whether you know exactly what you need or you’re still evaluating options, our team is available for a no-obligation conversation.
