Most Patch Cables (including select MTP®/MPO cables) Ship in 3–6 Business Days · Buy any 6, Get 5% Off ·

SpeedyFiber  |  SKU: OSFP-1600G-DR8-500M

1.6T 2xDR4/DR8 OSFP IHS/Closed Finned Top 8 x 200G PAM4 1310nm 500m Dual MPO-12/APC SMF Optical Transceiver Module


  • 8x 212G PAM4 Retimed 1.6TAUI-8 Electrical Interface
  • Designed to IEEE P802.3dj Draft for 200G-per-Lane Ethernet
  • Full Parameter Tests (BER, TDECQ, Eye Diagram, Temperature Cycling) for Superior Performance, Quality, and Reliability
  • Hot Pluggable, OSFP MSA and CMIS 5.2 Compliant
  • 200G-per-Lane DFB Laser Chip to Ensure Stable Production and Supply

1.6T 2xDR4/DR8 OSFP IHS/Closed Finned Top 8 x 200G PAM4 1310nm 500m Dual MPO-12/APC SMF Optical Transceiver Module has a longer lead time. Please feel free to contact us via chat before placing your order.

Tax included Shipping calculated at checkout.
  • 🎉 Buy any 6, Get 5% Off.
  • 🚚 Typically ships in 5 weeks
  • ✈️ $199 USD+ Free Shipping — US, Canada, UK, EU

📦 For bulk orders, please contact us for a customized quote

Contact us

1.6T 2xDR4/DR8 Twin-port OSFP PAM4 Optical Transceiver Module (SMF, 1310nm, 500m, Dual MTP/MPO-12 APC, IHS/Closed Finned Top)

The 1.6T twin-port OSFP transceiver delivers 1.6Tb/s aggregate throughput over OS2 single-mode fiber (SMF) at 1310nm, reaching up to 500m. It integrates eight optical lanes of 200Gb/s PAM4, split across two independent MTP/MPO-12 APC ports of 800Gb/s each (4 lanes / 8 fibers per port), enabling a native 1.6T-to-2×800G DR4 breakout with no external breakout cable. The host side uses an 8×200Gb/s PAM4 (200GAUI-1 C2M) electrical interface.

The module is designed to the IEEE P802.3dj draft and the OSFP MSA, and is managed over CMIS, with built-in digital diagnostics monitoring (DDM/DOM) for real-time access to temperature, supply voltage, bias current and optical Tx/Rx power. The IHS (Integrated Heat Spreader) closed finned-top housing is intended for air-cooled Ethernet and InfiniBand switches. With low latency, low power and high reliability, it links upward in spine-leaf architectures for switch-to-switch applications, and downward from top-of-rack switches to 800G network adapters and DPUs in compute and storage nodes — an ideal fit for AI/ML back-end fabrics, HPC and cloud data centers.