NVIDIA L4-2Q
資料中心An efficient, low-power data center GPU based on the Ada Lovelace architecture, optimized for AI inference and media processing.
主要規格
VRAM
2 GB GDDR6
記憶體匯流排
96-bit
FP16
67.2 TFLOPS
FP32
33.6 TFLOPS
TDP
72 W
最佳使用案例
AI Inference
Delivers high throughput for transformer-based AI models with Tensor cores optimized for FP8 and INT8 workloads in virtualized environments.
Video Encoding & Streaming
Features dedicated AV1 encoders/decoders for efficient video transcoding, making it ideal for live streaming and media delivery platforms.
Graphics Virtualization
Provides a powerful virtual GPU experience for remote desktops and cloud gaming applications with professional graphics capabilities.