Home
Nvidia
Computer Hardware
DGX A100
Nvidia DGX A100 User Manual
5
of 1
of 1 rating
118 pages
Give review
Manual
Specs
To Next Page
To Next Page
To Previous Page
To Previous Page
Loading...
NVIDIA DGX A100
DU-09821-001 _v01|1
Chapter1.
Introduction to the NVIDIA
DGX A100 System
The NVIDIA DGX
™
A100 System is the universal system purpose-built for all AI infrastructure
and workloads, from analytics to training to inference. The system is built on eight NVIDIA
A100 Tensor Core GPUs.
This document is for users and administrators of the DGX A100 system.
1.1.
Hardware Overview
This section provides information about the hardware in DGX A100.
8
10
Table of Contents
default chapter
2
Table of Contents
2
Chapter 1. Introduction to the NVIDIA DGX A100 System
9
Hardware Overview
9
DGX A100 Models and Component Descriptions
10
Mechanical Specifications
11
Power Specifications
12
Support for N+N Redundancy
12
DGX A100 Locking Power Cord Specification
12
Using the Locking Power Cords
13
Environmental Specifications
15
Front Panel Connections and Controls
15
With a Bezel
15
With the Bezel Removed
16
Rear Panel Modules
16
Motherboard Connections and Controls
17
Motherboard Tray Components
18
GPU Tray Components
18
Network Connections, Cables, and Adaptors
19
Network Ports
19
Supported Network Cables and Adaptors
20
DGX A100 System Topology
21
DGX os Software
21
Additional Documentation
22
Customer Support
22
Chapter 2. Connecting to the DGX A100
23
Connecting to the Console
23
Direct Connection
23
Remote Connection through the BMC
25
SSH Connection to the os
27
Chapter 3. First Boot Setup
28
Setting up the System
28
Post Setup Tasks
31
Obtaining Software Updates
31
Enabling the Srp Daemon
31
Chapter 4. Quick Start and Basic Operation
32
Installation and Configuration
32
Registration
32
Obtaining an NGC Account
32
Turning DGX A100 on and off
32
Startup Considerations
33
Shutdown Considerations
33
Verifying Functionality - Quick Health Check
33
Running the Pre-Flight Test
34
Running NGC Containers with GPU Support
34
Using Native GPU Support
35
Using the NVIDIA Container Runtime for Docker
35
Managing CPU Mitigations
36
Determining the CPU Mitigation State of the DGX System
36
Disabling CPU Mitigations
37
Re-Enabling CPU Mitigations
37
Chapter 5. Additional Features and Instructions
38
Managing the DGX Crash Dump Feature
38
Using the Script
38
Connecting to Serial over LAN to View the Console
38
Chapter 6. Managing the DGX A100 Self-Encrypting Drives
39
Overview
39
Installing the Software
40
Configuring Trusted Computing
40
Determining Whether Drives Support SID
40
Enabling the TPM and Preventing the BIOS from Sending Block SID Requests
41
Initializing the System for Drive Encryption
41
Enabling Drive Locking
42
Initialization Examples
42
Example 1: Passing in the JSON File
42
Determining Which Drives Can be Managed as Self-Encrypting
43
Creating the Drive/Password Mapping JSON Files and Using It to Initialize the System
44
Example 2: Generating Random Passwords
44
Example 3: Specifying Passwords One at a Time When Prompted
45
Disabling Drive Locking
45
Exporting the Vault
45
Erasing Your Data
45
Clearing the TPM
46
Changing Disk Passwords, Adding Disks, or Replacing Disks
46
Recovering from Lost Keys
47
Chapter 7. Network Configuration
48
Configuring Network Proxies
48
For the os and most Applications
48
For Apt
48
For Docker
49
Configuring Docker IP Addresses
49
Open Ports
50
Connectivity Requirements for NGC Containers
50
Configuring a Static IP Address for the BMC
51
Configuring a BMC Static Address by Using Ipmitool
51
Configuring a BMC Static IP Address by Using the System BIOS
52
Configuring a BMC Static IP Address for the Network Ports
52
Switching between Infiniband and Ethernet
53
Starting the Mellanox Software Tools and Determining the Current Port Configuration
54
Switching the Port Configuration
54
Chapter 8. Configuring Storage
55
Setting Filesystem Quotas
56
Switching between RAID 0 and RAID 5
56
Configuring Support for Custom Drive Partitioning
57
Chapter 9. Updating and Restoring the Software
58
Updating the DGX A100 Software
58
Connectivity Requirements for Software Updates
58
Update Instructions
59
Restoring the DGX A100 Software Image
59
Obtaining the DGX A100 Software ISO Image and Checksum File
59
Remotely Reimaging the System
60
Creating a Bootable Installation Medium
61
Creating a Bootable USB Flash Drive by Using the DD Command
62
Creating a Bootable USB Flash Drive by Using Akeo Rufus
63
Reimaging the System from a USB Flash Drive
64
Installation Options
64
Retaining the RAID Partition While Installing the os
64
Advanced Installation Option (Encrypted Root - DGX os 5 or Later)
65
Boot into Live Environment (DGX os 5 or Later)
65
Check Disc for Defects (DGX os 5 or Later)
66
Chapter 10. Using the BMC
67
Connecting to the BMC
67
Overview of BMC Controls
68
Common BMC Tasks
70
Changing the BMC Login Credentials
70
Using the Remote Console
71
Setting up Active Directory or LDAP/E-Directory
72
Configuring Platform Event Filters
72
Uploading or Generating SSL Certificates
73
Viewing the SSL Certificate
74
Generating the SSL Certificate
74
Uploading the SSL Certificate
76
Updating the SBIOS Certificate
76
Chapter 11. SBIOS Settings
80
Accessing the SBIOS Setup
80
Configuring the Boot Order
81
Configuring the Local Terminal to Access the SBIOS Settings Screen
82
If Using the IPMI SOL Protocol
83
For Linux Desktop Users, Set the Character Encoding
83
Power on or Reboot the System Using One of the Following Interfaces, Depending on What the State of the System Is
83
Chapter 12. Multi-Instance GPU
84
Chapter 13. Security
85
User Security Measures
85
Securing the BMC Port
85
System Security Measures
85
Secure Flash of DGX A100 Firmware
85
Encryption
85
Signing
86
NVSM Security
86
Secure Data Deletion
86
Prerequisites
86
Instructions
86
Chapter 14. Redfish Apis Support
88
Supported Redfish Features
88
Appendix A. Installing Software on Air-Gapped DGX A100 Systems
90
A.1. Installing NVDIA DGX A100 Software
90
A.2. Reimaging the System
90
A.3. Creating a Local Mirror of the NVIDIA and Canonical Repositories
91
A.3.1. Creating the Mirror in a DGX os 4 System
91
A.3.2. Configuring the Target Air-Gapped DGX os 4 System
93
A.3.3. Configuring the Target Air-Gapped DGX os 5 System
95
A.4. Installing Docker Containers
96
Appendix B. Safety
98
B.1. Safety Information
98
B.2. Safety Warnings and Cautions
98
B.3. Intended Application Uses
100
B.4. Site Selection
100
B.5. Equipment Handling Practices
100
B.6. Electrical Precautions
101
B.7. System Access Warnings
102
B.8. Rack Mount Warnings
102
B.9. Electrostatic Discharge
103
B.10. Other Hazards
104
Appendix C. Compliance
106
C.1. United States
106
C.2. United States/Canada
106
C.3. Canada
107
C.5. Australia and New Zealand
108
C.6. Brazil
108
C.7. Japan
108
C.8. South Korea
111
C.9. China
112
C.10. Taiwan
114
C.11. Russia/Kazakhstan/Belarus
115
C.12. Israel
116
C.13. India
116
C.14. South Africa
117
C.15. Great Britain (England, Wales, and Scotland)
117
Other manuals for Nvidia DGX A100
Service Manual
108 pages
5
Based on 1 rating
Ask a question
Give review
Questions and Answers:
Need help?
Do you have a question about the Nvidia DGX A100 and is the answer not in the manual?
Ask a question
Nvidia DGX A100 Specifications
General
Brand
Nvidia
Model
DGX A100
Category
Computer Hardware
Language
English
Related product manuals
Nvidia DGX H100
92 pages
Nvidia JETSON DEVELOPER KIT
32 pages
Nvidia DGX Station A100
72 pages
Nvidia RTX 5000
145 pages
Nvidia RTX 4000
145 pages
Nvidia TESLA K20
27 pages
Nvidia Jetson Orin NX
90 pages
Nvidia GeForce3 Ti200
18 pages
Nvidia JETSON XAVIER NX
15 pages
Nvidia Jetson AGX Orin Series
51 pages
Geforce PCI-Express Series
26 pages
Mellanox ConnectX-6 MCX654106A-ECAT
90 pages
To Next PageTo Next PageTo Previous PageTo Previous Page
Loading...
Need help?
General
