100G QSFP28 VR1 and 200G QSFP56 VR2 Optical Modules

With the widespread adoption of 100G PAM4 modulation technology, current 400G/800G optical modules extensively utilize the 4×100G PAM4 and 8×100G PAM4 architectures. Against this backdrop of technological evolution, traditional 4×25G NRZ schemes (e.g., 100G QSFP28 SR4) or 4×50G PAM4 schemes (e.g., 200G QSFP56 SR4) face three critical challenges: firstly, rate mechanism incompatibility causing link negotiation failures; secondly, reduced port resource utilization leading to bandwidth waste; and thirdly, exponentially increasing hardware renovation costs. To address these pain points in the industry, TARLUZ has introduced the 100G QSFP28 VR1 and 200G QSFP56 VR2 optical modules with natively compatible designs.

Application Introduction

When using traditional 200G QSFP56 SR4 optical modules plugged to NVIDIA ConnectX-7 (dual-port 200Gb/s) network interface cards with 400G QSFP112 switches, there is a risk of link disconnection, and the switch may trigger an «Unsupported speed» error. This issue can be resolved by deploying the 200G QSFP56 VR2/SR2 optical module, which employs dual-lane PAM4 modulation technology. This technology allows the 400G QSFP112 port to self-adapt and split into two independent 200G channels, creating a seamless connection solution with 200G QSFP56 Adapters without requiring any physical modifications to the existing hardware infrastructure. This technical solution also offers significant scalability: a 400G QSFP112 switch can be automatically configured into a 4×100G link mode by deploying four 100G QSFP28 VR1 optical modules.