Nvidia DGX H100 User Manual

NVIDIA DGX H100 User Guide
curl -k -u <bmc-user>:<password> --location --request PATCH 'https:∕∕<bmc-ip-address>∕
,→redfish∕v1∕Systems∕DGX∕Bios∕SD' -H 'Content-Type: application∕json' -H 'If-Match:*
,→' --data-raw '{ "Attributes" : { "FBO201" :
,→"FBO201NetworkUEFIPXEIPv4IntelREthernetNetworkAdapterX710TL" }}'
The following example changes the boot order back to boot from M.2:
curl -k -u <bmc-user>:<password> --location --request PATCH 'https:∕∕<bmc-ip-address>∕
,→redfish∕v1∕Systems∕DGX∕Bios∕SD' -H 'Content-Type: application∕json' -H 'If-Match:*
,→' --data-raw '{ "Attributes" : { "FBO201" :
,→"FBO201NVMEubuntuSAMSUNGMZ1L21T9HCLS00A07" }}'
8.2.4. Telemetry
▶ HGX platform telemetry
curl -k -u <bmc-user>:<password> --location --request GET 'https:∕∕<bmc-
,→ip-address>∕redfish∕v1∕TelemetryService∕MetricReportDefinitions∕HGX_
,→PlatformEnvironmentMetrics_0'
▶ DGX platform sensors
curl -k -u <bmc-user>:<password> --location --request GET 'https:∕∕<bmc-
,→ip-address>∕redfish∕v1∕Chassis∕DGX∕Sensors'
The endpoint returns 75 members at a time. To page through the results, use the URI
in the Members@odata.nextLink eld. For example, ∕redfish∕v1∕Chassis∕DGX∕
Sensors?$skip=75.
8.2.5. Chassis
▶ Chassis Restart (IPMI chassis power cycle)
curl -k -u <bmc-user>:<password> --request POST --location 'https:∕∕<bmc-
,→ip-address>∕redfish∕v1∕Systems∕DGX∕Actions∕ComputerSystem.Reset' --
,→header 'Content-Type: application∕json' --data '{"ResetType":
,→"ForceRestart"}'
▶ Chassis Start (IPMI chassis power on)
curl -k -u <bmc-user>:<password> --request POST --location 'https:∕∕<bmc-
,→ip-address>∕redfish∕v1∕Systems∕DGX∕Actions∕ComputerSystem.Reset' --
,→header 'Content-Type: application∕json' --data '{"ResetType": "On"}'
▶ Chassis O (IPMI chassis power o)
curl -k -u <bmc-user>:<password> --request POST --location 'https:∕∕<bmc-
,→ip-address>∕redfish∕v1∕Systems∕DGX∕Actions∕ComputerSystem.Reset' --
,→header 'Content-Type: application∕json' --data '{"ResetType":
,→"ForceOff"}'
▶ Chassis O Gracefully (IPMI chassis soft)
60 Chapter 8. Redsh APIs Support

Nvidia DGX H100 Specifications

General

GPU	8x NVIDIA H100 Tensor Core GPUs
GPU Memory	640 GB HBM3 (80GB per GPU)
System Memory	2 TB DDR5
Form Factor	6U rackmount
Storage	30TB NVMe SSD
Power Supply	10kW
Interconnect	NVLink 4.0