Check serviceability
at your Pincode
AI Across Industries With DGX
The NVIDIA DGX A100 is revolutionizing various industries by offering unparalleled AI performance and versatility.
From accelerating medical research with advanced machine learning models to enhancing financial forecasting with deep learning algorithms, the DGX A100 provides the computational power needed to drive significant advancements.
Its 8 NVIDIA A100 Tensor Core GPUs, extensive memory, and high-speed networking capabilities enable businesses across healthcare, finance, manufacturing, and more to deploy sophisticated AI solutions, streamline operations, and achieve better outcomes faster and more efficiently.
From accelerating medical research with advanced machine learning models to enhancing financial forecasting with deep learning algorithms, the DGX A100 provides the computational power needed to drive significant advancements.
Its 8 NVIDIA A100 Tensor Core GPUs, extensive memory, and high-speed networking capabilities enable businesses across healthcare, finance, manufacturing, and more to deploy sophisticated AI solutions, streamline operations, and achieve better outcomes faster and more efficiently.
Component
Specification
CPU
1x AMD EPYC 7742, 64 cores, 2.25 GHz (base) to 3.4 GHz (max boost)
GPU
4x NVIDIA A100 Tensor Core GPUs, 40 GB each (160 GB total), with MIG support
System Memory
512 GB DDR4 RAM (8x 64 GB modules)
Data Storage
7.68 TB NVMe U.2 drive for cache/data storage
OS Storage
1.92 TB NVMe M.2 drive for the operating system
Power Supply
1500W; 100-115VAC/15A, 115-120VAC/12A, or 200-240VAC/10A at 50/60Hz
Dimensions
Height: 26.1 in (639 mm), Width: 10.1 in (256 mm), Depth: 20.4 in (518 mm)
Weight
91 lbs (43.1 kg)
Operating Temperature
10°C to 35°C (50°F to 95°F)
Non-Operating Temperature
5°C to 40°C (41°F to 104°F)
Humidity
10% to 80% (non-condensing) for both operating and non-operating conditions
Networking
Dual 10 Gb Ethernet ports
Cooling
Air-cooled system
Software
DGX OS with pre-installed software for deep learning frameworks, supports NVIDIA Docker