🌟 NVIDIA Mellanox® MCP7Y50-N001 | 800Gb/s to 4×200Gb/s Passive Breakout Cable (DAC) | OSFP → 4×OSFP | Flat-Top | 1m | InfiniBand NDR / NDR200
Product Overview
The NVIDIA Mellanox® MCP7Y50-N001 (P/N 980-9I75E-00N001) is a high-density 1-to-4 passive splitter (breakout cable) that converts a single 800Gb/s OSFP NDR port into four independent 200Gb/s OSFP NDR200 links.
This rare OSFP-to-OSFP topology is designed for liquid-cooled servers, specialized compute blades, and thermal-sensitive accelerator modules that require OSFP cages—even at reduced 200Gb/s link rates. The Flat-Top OSFP design supports environments where RHS/Finned-Top heat dissipation is not required, such as cold-plate cooling systems.
At 1 meter, it is ideal for short-distance, ultra-low-latency breakout connectivity inside high-density AI and HPC racks.
Key Features
🚀 True NDR200 Breakout (800G → 4×200G)
The cable divides the 800G OSFP link into:
4 × 200Gb/s OSFP outputs (NDR200)
Each 200G leg uses:
2 × 100G PAM4 lanes (correct NDR200 lane assignment).🔥 OSFP on ALL Ends — Rare Topology
Most 200G servers use QSFP112.
This cable provides OSFP → OSFP, used when:
The server/accelerator uses OSFP cages for better thermal headroom
The environment is liquid-cooled or requires Flat-Top OSFP modules
Blade platforms use OSFP midplanes for uniformity
⚡ Passive Copper (0W, Zero Latency)
No active components →
• Zero added latency
• No power draw
• Higher long-term reliability
• Lower thermal load than ACC or AOC
🛡️ LSZH Jacket
Meets modern data center safety requirements.
Technical Specifications
| Specification | Details |
|---|---|
| Model | MCP7Y50-N001 |
| NVIDIA P/N | 980-9I75E-00N001 |
| Type | Passive Breakout (DAC), 1→4 |
| Upstream | OSFP, 800Gb/s NDR |
| Downstream | 4 × OSFP, 200Gb/s NDR200 |
| Lane Assignment | 2×100G PAM4 lanes per 200G leg |
| Length | 1 meter |
| Jacket | LSZH |
| Design | Flat-Top OSFP (NOT RHS) |
| Condition | Refurbished & Fully Tested |
Ideal Applications
This breakout cable is suited for:
🔹 Liquid-cooled GPU clusters
🔹 OSFP-based compute blades
🔹 Thermally constrained HPC systems
🔹 AI/ML training infrastructure
🔹 High-density NDR/NDR200 fabrics
🔹 Rack-scale fan-out (1 switch → 4 nodes)
Use case example:
1 NDR switch port → 4 OSFP compute accelerators for maximum density.
Enhanced Network Efficiency
Designed for InfiniBand NDR/NDR200
Extremely low latency for tightly coupled workloads
Reliable performance across all four breakout channels
Provides thermal advantages over QSFP112 in liquid-cooled environments
Compatibility
✅ Compatible With
• NVIDIA Quantum-2 NDR switches (OSFP)
• OSFP-based servers, blades, and accelerators (200Gb/s NDR200 mode)
• Environments requiring Flat-Top OSFP modules
❌ Not Compatible With
• Ethernet
• HDR (200Gb/s) systems – NDR200 is electrically different
• QSFP112-based servers or NICs
• Air-cooled OSFP cages requiring RHS/Finned connectors
Quality Assurance
✓ Fully signal-tested on NVIDIA NDR/NDR200 hardware
✓ Professionally refurbished and validated
✓ Cleaned, inspected, and packaged for immediate deployment
✓ 100% compatibility guarantee
Payment & Shipping
• Payment: PayPal, credit cards, bank transfer
• Worldwide Express: FedEx/TNT (8–13 days) with tracking
• Returns: Accepted within policy (buyer pays return shipping)
📦 Why Buy From T.E.S IT-Solutions?
T.E.S IT-Solutions provides enterprise-grade Mellanox/NVIDIA hardware backed by rigorous testing, expert support, and fast international delivery. Our inventory supports AI, HPC, cloud, and enterprise deployments requiring maximum reliability and performance.