Nvidia Tesla H100 NVL 94GB

NVIDIA Tesla H100 Tensor Core GPU

The NVIDIA Tesla H100 is a cutting-edge GPU designed to accelerate the most demanding AI, machine learning, and high-performance computing workloads. Built on the Hopper architecture, it delivers exceptional performance for training large AI models, deep learning inference, and complex simulations.

Optimized for next-gen AI applications, the H100 enables faster model development and more efficient scaling across multi-GPU environments. Perfect for data centers and high-performance computing, the Tesla H100 provides the power and efficiency needed for breakthrough innovations in AI, healthcare, and scientific research.

Specifications:

GPU Architecture: NVIDIA Hopper
CUDA Cores: 14,592
GPU Memory: 94 GB
Memory Type: HBM3e
Memory Bandwidth: 3.9 TB/s
FP64 Tensor Core : 60 TFLOPS
TF32 Tensor Core²: 835 TFLOPS
BFLOAT16 Tensor Core²: 1,671 TFLOPS
FP16 Tensor Core²: 1,671 TFLOPS
FP8 Tensor Core²: 3,341 TFLOPS
INT8 Tensor Core²: 3,341 TFLOPS
NVLink:2- or 4-way NVIDIA NVLink bridge: 900GB/s per GPU
PCIe Gen5: 128GB/s
Form Factor: PCIe 5.0 Dual-slot air-cooled
Thermal Design Power (TDP): 350-400W(configurable)
NVIDIA AI Enterprise: Included
Interconnect: NVIDIA NVLink, PCIe Gen 5
Supported Software: NVIDIA CUDA, cuDNN, TensorRT, Deep Learning SDK, HPC Libraries
Target Use Cases: AI Model Training, Deep Learning, Data Analytics, Scientific Computing, High-Performance Computing (HPC)