← 返回 GPU 列表

NVIDIA L4-12Q

中階

A low-power, single-slot GPU designed for edge AI inference and dense server deployments requiring high energy efficiency. It delivers solid performance for running large language models and vision tasks in compact form factors.

主要規格

VRAM 12 GB GDDR6
記憶體匯流排 192-bit
FP16 30.3 TFLOPS
FP32 30.3 TFLOPS
TDP 72 W

最佳使用案例

Edge AI Inference

Ideal for deploying computer vision and generative AI models at the edge with minimal power consumption and heat generation. It fits easily into dense rack or edge server configurations without specialized cooling.

Video Processing

Efficiently handles video transcoding and decoding tasks, making it suitable for video surveillance and streaming infrastructure. Its low TDP ensures reliable performance in always-on video processing workloads.

Digital Signage

Provides powerful graphics acceleration for high-resolution digital signage and interactive displays. Its compact size and low power draw make it perfect for fanless or passively cooled signage systems.