Biren · Module / Kit
BIREN 166L
The BIREN 壁砺 166L is a train-and-inference converged liquid-cooled product emphasizing energy saving, high compute efficiency, high interconnect and high generality. It is a cold-plate liquid-cooled OAM module with peak power 600W, compatible with standard OAM V1.X GPU servers, used for AI training and inference including LLMs, multimodal AIGC, image/speech recognition, NLP and recommendation systems.
Key Highlights
- Energy-saving, high efficiency, high interconnect, high generality
- cold-plate liquid-cooled OAM module, peak 600W
- compatible with standard OAM V1.X GPU servers.
Technical Specifications
| Form Factor | 冷板式液冷 OAM 模组 / Cold-plate liquid-cooled OAM module |
|---|---|
| Peak Board Power | 600W |
| Server Compatibility | 标准 OAM V1.X 版本 GPU 服务器 / standard OAM V1.X GPU servers |
| 显存容量 / GPU Memory: | |
| 精度算力（FP16/FP8/INT8, TFLOPS/TOPS）/ Precision Compute: | |
| Interconnect | OAM高速互连（带宽/协议官网未公开）/ OAM interconnect |
| details not disclosed | |
| Positioning | 训练与推理一体（训推一体）/ train & inference converged |
Typical Use Cases
LLM training & inferencemultimodal AIGCimage recognitionspeech recognitionNLPrecommendation systems
img1: main; alt=官网白底产品渲染主图(render); 官网无英文版，英文列为翻译稿; 官网未公开显存容量、FP16/FP8/INT8精度算力(TFLOPS/TOPS)、显存带宽、PCIe代数与互连带宽；官网无硬件Datasheet/PDF；本站未披露整机/集群产品页。
Related Products
Biren
BIREN 106B
FHFL Dual-Wide PCIe AI Training & Inference Accelerator Card View details →
Biren
BIREN 106M
Air-Cooled OAM Module for AI Training View details →
Biren
BIREN 166C
Datacenter High-Performance Inference PCIe AI Accelerator View details →
Biren