Tesla T4
Data CenterA data center GPU built on the Turing architecture with low power draw, 16 GB of GDDR6, and Tensor Cores tuned for inference.
Key Specifications
VRAM
16 GB GDDR6
Memory Bus
256-bit
FP16
65.1 TFLOPS
FP32
8.1 TFLOPS
TDP
70 W
Best Use Cases
AI Inference
Run production inference at scale with a 70 W TDP and INT8 Tensor Cores.
Video Transcoding
Hardware-accelerated video encode and decode using NVENC and NVDEC.
Virtual Desktops
Drive NVIDIA vGPU virtual desktop infrastructure for enterprise users.