← Back to GPU List

NVIDIA L4-2Q

Data Center

A low-power, high-efficiency data center GPU built on the Ada Lovelace architecture, purpose-built for AI inference and media processing workloads.

Key Specifications

VRAM 2 GB GDDR6
Memory Bus 96-bit
FP16 67.2 TFLOPS
FP32 33.6 TFLOPS
TDP 72 W

Best Use Cases

AI Inference

Achieves strong throughput for transformer-centric AI models with Tensor cores tuned for FP8 and INT8 operations in virtualized settings.

Video Encoding & Streaming

Comes equipped with dedicated AV1 encoders and decoders for streamlined video transcoding, suited for live streaming and content delivery networks.

Graphics Virtualization

Offers a robust virtual GPU solution for remote desktops and cloud gaming platforms, complete with professional-grade graphics features.