{"id":12261,"date":"2026-05-14T18:22:31","date_gmt":"2026-05-14T10:22:31","guid":{"rendered":"https:\/\/ascentoptics.com\/blog\/?p=12261"},"modified":"2026-05-14T18:26:10","modified_gmt":"2026-05-14T10:26:10","slug":"osfp-nvidia-infiniband-ndr","status":"publish","type":"post","link":"https:\/\/ascentoptics.com\/blog\/osfp-nvidia-infiniband-ndr\/","title":{"rendered":"OSFP NVIDIA InfiniBand: ConnectX-7 & Quantum-2 Guide"},"content":{"rendered":"

An AI infrastructure architect at a hyperscale cloud provider faced a critical procurement decision that would impact her team\u2019s GPU cluster economics for the next three years. The project required 128 NVIDIA DGX H100 servers (1,024 H100 GPUs) interconnected via NVIDIA Quantum-2 InfiniBand NDR switches. At NVIDIA list pricing, the transceivers alone exceeded $1.5 million. By validating compatible third-party OSFP modules with ConnectX-7 NICs, her team reduced costs by $850,000 \u2014 a 55% savings \u2014 without any compromise in link quality or Bit Error Rate performance.<\/p>\n

This is the reality of procuring OSFP modules for large-scale NVIDIA InfiniBand deployments. NVIDIA\u2019s official optical portfolio is extensive and expensive. While their technical requirements are comprehensive, third-party alternatives can offer significant savings if properly validated. The success of your cluster deployment often hinges on correct form factor selection, accurate SKU matching, and appropriate cable choices.<\/p>\n

This guide explains how OSFP transceivers work in NVIDIA InfiniBand NDR networks, which SKUs are required for ConnectX-7 NICs and Quantum-2 switches, the critical differences between IHS and RHS form factors, and how to evaluate third-party options for maximum cost savings while maintaining full compatibility.<\/p>\n

Need OSFP modules tested for NVIDIA ConnectX-7?<\/strong>\u00a0Explore our OSFP catalog<\/u><\/a>\u00a0for MMA4Z00-NS compatible modules and platform validation.<\/p>\n

 <\/p>\n

 <\/p>\n

What Is OSFP for NVIDIA InfiniBand?<\/strong><\/h2>\n

NVIDIA selected the OSFP form factor for its NDR (400G InfiniBand) ecosystem because it provides the thermal headroom and signal integrity required for 100G PAM4 lane operation. Each NDR link operates at 400Gb\/s using 4 \u00d7 100G PAM4 electrical and optical lanes.<\/p>\n

On Quantum-2 switches, NVIDIA uses a unique twin-port OSFP implementation in which a single OSFP cage carries two independent 400G InfiniBand links, enabling extremely high port density within a 1U chassis.<\/p>\n

 <\/p>\n

The NVIDIA Quantum-2 NDR Platform<\/strong><\/h3>\n

The NVIDIA Quantum-2 platform includes the QM9700 and QM9790 InfiniBand switches, each providing 32 twin-port OSFP cages in a compact 1U form factor. These cages support up to 64 NDR 400G InfiniBand ports or 128 NDR200 connections through breakout configurations.<\/p>\n

The platform delivers up to 51.2Tb\/s of aggregate bidirectional switching bandwidth and is optimized for large-scale AI and HPC fabrics.<\/p>\n

 <\/p>\n

\"Quantum-2<\/p>\n

 <\/p>\n

 <\/p>\n

ConnectX-7 SmartNIC Architecture<\/strong><\/h3>\n

NVIDIA ConnectX-7 adapters are designed for NDR 400G InfiniBand and 400GbE deployments. The platform supports PCIe Gen5 x16 connectivity and is available with either OSFP or QSFP112 network interfaces.<\/p>\n

For OSFP-based configurations, ConnectX-7 adapters use flat-top RHS (Riding Heat Sink) transceivers. In this design, thermal dissipation is handled primarily by the NIC\u2019s integrated heatsink structure rather than by fins on the module itself.<\/p>\n

 <\/p>\n

Why NVIDIA Standardized on OSFP for NDR<\/strong><\/h3>\n

NVIDIA adopted OSFP for NDR InfiniBand primarily because of its superior thermal and power handling capabilities. Compared with QSFP-DD, OSFP provides additional thermal headroom, making it more suitable for sustained 100G PAM4 lane operation in dense AI fabrics.<\/p>\n

The OSFP ecosystem also aligned well with NVIDIA\u2019s long-term roadmap toward higher-density 800G and future 1.6T interconnect architectures.<\/p>\n

 <\/p>\n

 <\/p>\n

OSFP Form Factors for NVIDIA Platforms<\/strong><\/h2>\n

Thermal compatibility is essential when selecting OSFP modules for NVIDIA platforms. Using the wrong variant can result in modules that physically will not fit.<\/p>\n

 <\/p>\n

Twin-Port OSFP IHS (Integrated Heat Sink \/ Finned-Top) for Switches<\/strong><\/h3>\n

IHS (Integrated Heat Sink) OSFP modules are designed primarily for switch environments where airflow moves directly across finned-top transceivers. RHS (Riding Heat Sink) flat-top modules are optimized for NIC environments in which cooling is handled by the adapter\u2019s integrated thermal assembly.<\/p>\n

Although both variants share the same electrical signaling architecture, they are not mechanically or thermally interchangeable in most NVIDIA deployments.<\/p>\n

 <\/p>\n

Single-Port OSFP RHS (Riding Heat Sink \/ Flat-Top) for NICs<\/strong><\/h3>\n

NIC-side modules use a flat-top design optimized to work under the heatsink of ConnectX-7, ConnectX-8, and BlueField-3 cards. These single-port modules support 400G (MPO-12 APC or duplex LC) with power consumption of 8W at 400G and 5.5W at 200G.<\/p>\n

 <\/p>\n

Why You Cannot Mix Form Factors<\/strong><\/h3>\n

IHS (finned) modules are too tall for NIC cages, while flat-top RHS modules lack sufficient cooling in switch chassis designed for IHS airflow. Mixing them leads to physical fitment or thermal issues.<\/p>\n

In one real-world case, a national research lab engineer ordered finned IHS modules for both switch and DGX H100 server sides, causing physical incompatibility and a 10-day delay. Always match the thermal profile to the platform.<\/p>\n

Need help selecting the right OSFP variant?<\/strong>\u00a0Contact our optical engineers<\/u><\/a>\u00a0for platform compatibility verification before ordering.<\/p>\n

 <\/p>\n

QSFP112 as a ConnectX-7 Alternative<\/strong><\/h3>\n

ConnectX-7 NICs also support QSFP112 receptacles for 400G NDR connectivity. QSFP112 uses the same 4\u00d7100G PAM4 architecture. NVIDIA\u2019s MMA1Z00 series covers both OSFP and QSFP112 variants with identical optics and performance. The choice depends on the specific NIC port design and surrounding equipment.<\/p>\n

 <\/p>\n

\"IHS<\/p>\n

 <\/p>\n

 <\/p>\n

NVIDIA OSFP Transceiver Portfolio<\/strong><\/h2>\n

NVIDIA uses a consistent SKU naming convention that encodes form factor, fiber type, and reach.<\/p>\n

 <\/p>\n

MMA4Z00-NS Series (Twin-Port NDR OSFP)<\/strong><\/h3>\n

The MMA4Z00-NS series represents NVIDIA\u2019s twin-port OSFP transceivers for Quantum-2 InfiniBand switches.<\/p>\n

Rather than operating as a single 800G Ethernet optical link, these modules carry two independent 400G NDR InfiniBand connections within a shared OSFP form factor.<\/p>\n

Typical characteristics include:<\/p>\n