PNY
NVIDIA Tesla T4 - GPU Compute Processor - Tesla T4 - 16 GB GDDR6 - PCIe 3.0 x16 - fanless - retail
Estimated shipping in 30 business days
EAN: 3536403367510
PNY NVIDIA Tesla T4 16GB GDDR6
Compact computing power to accelerate AI inference, virtualization, video, and professional workloads, with a low-profile, single-slot form factor and passive cooling.
Designed for professional environments and servers with adequate airflow, the Tesla T4 delivers high performance per watt, high computing density, and simple integration into infrastructures with limited space.
More performance per watt
Accelerates inference, video, and professional computing workloads with high energy efficiency, helping to increase performance density without requiring larger form factors.
Ready for AI and inference
Combines CUDA Cores and Tensor Cores to handle modern frameworks, model acceleration, and workflows that require low latency and consistency.
Ideal for limited space
The single-slot, low-profile form factor facilitates integration into compact chassis and infrastructures where space and controlled ventilation make all the difference.
Transforms installed capacity into real acceleration
Instead of occupying more space with larger and hotter hardware, the Tesla T4 allows you to increase processing capacity with a compact, efficient, and ready-for-demanding-workloads GPU.
Limited infrastructure
- Reduced chassis space
- Sensitive energy consumption
- Slow processing in AI and video
- Difficulty scaling without complicating the environment
More acceleration, less compromise
- Compact, single-slot GPU
- Greater efficiency per watt
- Better response in inference and virtualization
- Professional integration with a reduced footprint
Technical specifications
Complete product characteristics table.
| Brand | PNY |
| Model | NVIDIA Tesla T4 |
| EAN | 03536403367510 |
| GPU Architecture | NVIDIA Turing |
| Tensor Cores | 320 |
| CUDA Cores | 2.560 |
| FP32 Performance | 8.1 TFLOPS |
| Mixed Precision (FP16/FP32) | 65 TFLOPS |
| INT8 | 130 TOPS |
| INT4 | 260 TOPS |
| Memory | 16 GB GDDR6 |
| Memory Bandwidth | 300 GB/s |
| ECC | Yes |
| Interconnect Bandwidth | 32 GB/s |
| System Interface | PCIe Gen 3.0 x16 |
| Form Factor | Low-profile, single-slot |
| Cooling | Passive, fanless |
| Max Power Consumption | 70 W |
| Computing APIs | CUDA, NVIDIA TensorRT, ONNX |
| Recommended Applications | AI inference, virtualization, video, accelerated computing and professional workloads in a server environment |
-10% additional
Subscribe to our newsletter and get an additional 10% discount on your order and access to new promotions, products and publications