NVIDIA T4 16GB GDDR6 inference GPU on Turing TU104 architecture. 70W TDP, 320 GB/s bandwidth, 65 TFLOPS FP16, 130 TOPS INT8 and 260 TOPS INT4. The workhorse GPU for scale-out AI inference, virtualization, and video transcoding.
T4 16GB
Turing TU104
16 GB GDDR6
320 GB/s
65 TFLOPS
8.1 TFLOPS
130 TOPS
260 TOPS
PCIe 3.0 x16
70W
Low-profile, single-slot
900-2G183-0000-000
NVIDIA T4 is the industry-standard inference GPU — deployed across every major cloud and enterprise for real-time AI serving, VDI, and video transcoding. Its 70W TDP and single-slot low-profile form factor enable high-density multi-GPU nodes, making it the most cost-efficient path to scale-out inference.
Need NVIDIA T4?
Enterprise hardware — configured, tested, deployed. Contact us for pricing and availability.
Request Quote