Доставка по всей России
Режим работы
Работаем c 10:00 по 18:00
Написать нам
info@itg-spb.ru
Наш адрес
г. Санкт-Петербург, улица Оптиков, 4к3

Ваша корзина пуста!

Назад в каталог
Nvidia H800-80G
Nvidia H800-80G

Nvidia H800-80G

Нет отзывов
Код товара: Nvidia H800-80G GPU
2 315 503 ₽
В наличии много 1000 шт

The NVIDIA H800-80G is a high-performance GPU tailored for AI and machine learning applications in enterprise data centers. Designed by NVIDIA, this GPU offers powerful computational capabilities, catering to engineers and IT professionals seeking to optimize their AI workloads efficiently.


Nvidia H800-80G Product Highlights
  • High-performance GPU
  • AI and ML applications
  • Enterprise-focused
  • Scalable architecture
  • Reliable performance

Product Features Comparison
NVIDIA H800-80G NVIDIA A100 NVIDIA V100 AMD MI100 Intel Habana Gaudi Google TPU v4 Graphcore IPU
Performance (TFLOPS) 80 70 50 100 100 80
Memory Type HBM2 HBM2 HBM2 HBM2e HBM SRAM
Memory Capacity (GB) 80 40 32 32 32 16
PCIe Support Gen 4 Gen 4 Gen 3 Gen 4 Gen 4 Gen 3
Form Factor PCIe PCIe PCIe PCIe OCP PCIe

Product name Product Application Scenarios
  • Training deep learning models
  • Deploying AI inference services
  • Big data analytics

Optional Add-ons
Model Description
NVIDIA NVLink Bridge High-speed GPU interconnect
NVIDIA TensorRT Deep learning inference optimizer
CUDA Toolkit Parallel computing software platform

Get More Information

Discover unbeatable prices and fast shipping for the NVIDIA H800-80G. Access the detailed NVIDIA H800-80G datasheet and ensure you're getting the best deal with our reliable delivery service. Need assistance? Our free live chat support is here to help. For more information about the product and pricing, contact us via live chat or email us at sales@router-switch.com.

Характеристики

API Support DirectX 12, OpenGL 4.6, Vulkan, CUDA, OpenCL
CUDA Cores 10,240
ECC Memory Yes
Form Factor PCIe
GPU Architecture Ampere
Maximum GPU Temperature 85°C
Memory Bandwidth 2 TB/s
Memory Size 80 GB
Memory Type HBM2e
Multi-GPU Technology NVLink
NVLink Support Yes, up to 12 links
Operating System Support Windows, Linux
PCI Express Gen 4
Peak FP16 Performance 134.4 TFLOPS
Peak FP32 Performance 67.2 TFLOPS
Peak INT8 Performance 1,072 TOPS
Power Consumption 500W
Tensor Cores 320
Thermal Solution Active/Passive

Средняя оценка

1

2

3

4

5

0

На основании 0 отзывов