• GPU: 8x A100 SXM4 40GB
  • System Memory: 1 TB 3,200 MHz DDR4
  • NVIDIA CUDA Cores: 55,296
  • CPU: Dual 64-Core AMD EPYC 7742
Contact sales for pricing

The Universal System for Every AI Workload

With the fastest I/O architecture of any DGX system, NVIDIA DGX A100 is the universal system for all AI infrastructure, from analytics to training to inference. It sets a new bar for compute density, packing 5 petaFLOPS of AI performance into a 6U form factor and replaces legacy compute infrastructure with a single, unified system that can do it all. Its 8 NVIDIA A100 Tensor Core GPUs can be used together to run the largest jobs, or divided into as many as 56 separate and fully isolated instances with dedicated high bandwidth memory, cache, and compute cores. The combination of dense compute power and complete workload flexibility makes DGX A100 ideal for both single node deployments and large scale clusters.

Integrated Access to Unmatched AI Expertise

NVIDIA DGX A100 is more than a server, it is a complete hardware and software platform backed by thousands of AI experts at NVIDIA, and is built upon the knowledge gained from the world’s largest DGX proving ground, NVIDIA DGX SATURNV. Owning a DGX A100 system gives you direct access to a global team of AI-fluent practitioners that offer prescriptive guidance and design expertise to help fastrack AI transformation with know-how and experience from NVIDIA’s decade plus of AI leadership. This ensures mission-critical applications get up and running quickly and stay running smoothly, dramatically improving time to insights.

Fastest Time To Solution

NVIDIA DGX A100 features eight NVIDIA A100 Tensor Core GPUs providing unmatched acceleration 1 at every scale and is fully optimized for NVIDIA CUDA-X software and the end-to-end NVIDIA data center solution stack. The unprecedented compute performance of DGX A100 delivers the fastest time-to-solution allowing users to iterate faster on the world’s most complex AI solutions.

BERT pre training throughput using Pytorch, including (2/3) Phase 1 and (1/3) Phase 2 | Phase 1 Seq Len = 128, Phase 2 Seq Len = 512; V100: NVIDIA DGX-1™ server with 8x V100 using FP32 precision; A100: DGX A100 Server with 8x A100 using TF32 precision.

Unmatched Data Center Scalability

NVIDIA DGX A100 is the foundational building block for large AI clusters such as NVIDIA DGX SuperPOD . DGX A100 debuts next generation NVLink and NVSwitch that are 2X faster than the previous generation and Mellanox ConnectX-6 VPI HDR InfiniBand/Ethernet network adapters with 450 gigabytes per second (GB/s) of peak bi-directional bandwidth. The combination of massive GPU-accelerated compute with state-of-the-art networking and GPU-accelerated software means DGX A100 can scale to hundreds or thousands of nodes to meet the biggest challenges such as conversational AI and large scale image classification.

The World’s Most Secure AI System For Enterprise

NVIDIA DGX A100 delivers the most robust security posture for your AI enterprise, with a multi-layered approach that secures all major hardware and software components. Stretching across the baseboard management controller (BMC), CPU board, GPU board, self-encrypted drives, and secure boot, DGX A100 has security built-in, allowing users to focus on results rather than threat assessment and mitigation.

General Information
Manufacturer Part Number
Manufacturer Website Address
Brand Name
Product Line
Product Name
DGX A100
Product Type
Graphics Computing System
Processor & Chipset
Number of Processors Installed
Processor Manufacturer
Processor Type
Processor Model
Processor Core
64 Core
Processor Speed
2.25 GHz
Number of GPUs Installed
Number of NVIDIA NVSwitches
GPU Manufacturer
Chipset Line
GPU Architecture
Chipset Model
Processor Core
  • 432 Tensor
  • 6912 CUDA
Peak FP64
9.7 TFLOPS (GPU Boost Clock)
Peak FP64 Tensor Core
19.5 TFLOPS (GPU Boost Clock)
Peak FP32
19.5 TFLOPS (GPU Boost Clock)
Peak FP16
78 TFLOPS (GPU Boost Clock)
Peak BF16
39 TFLOPS (GPU Boost Clock)
Peak TF32 Tensor Core
  • 156 TFLOPS (GPU Boost Clock)
  • 312 TFLOPS (Sparsity)
Peak FP16 Tensor Core
  • 312 TFLOPS (GPU Boost Clock)
  • 624 TFLOPS (Sparsity)
Peak BF16 Tensor Core
  • 312 TFLOPS (GPU Boost Clock)
  • 624 TFLOPS (Sparsity)
Peak INT8 Tensor Core
  • 624 TOPS (GPU Boost Clock)
  • 1248 TOPS (Sparsity)
Peak INT4 Tensor Core
  • 1248 TOPS (GPU Boost Clock)
  • 2496 TOPS (Sparsity)
Total GPU Memory
320 GB (8x Tesla Ampere A100 40 GB)
GPU Technology
Standard Memory
1 TB
Memory Technology
Total Hard Drive Capacity
  • OS: 2x 1.92TB M.2 NVMe
  • Internal Storage: 4x 3.84TB U.2 NVMe (15TB)
Storage Type
Solid State Drive
Storage Configuration
Network & Communication
  • 8x Single-Port Mellanox ConnectX-6 VPI 200Gb/s HDR InfiniBand
  • 1x Dual-Port Mellanox ConnectX-6 VPI 10/25/50/100/200Gb/s Ethernet
Operating System
Ubuntu Linux OS
Power Description
Maximum Power Consumption
6500 W
Physical Characteristics
Form Factor
Rack Height
Weight (Approximate)
271 lbs