Granted Patent B2

US 11,675,599 · App. 16/984,726 · Granted Jun 13, 2023

Inventors: Chitrak Gupta (Bangalore, IN); Rama Rao Bisa (Bangalore, IN); John R. Palmer (Georgetown, TX)

Assignee: Dell Products L.P.

G06F9/3877G06F9/4411G06F13/20

Abstract

Claims (30)

1. An information handling system comprising:

a processor;

one or more accelerators communicatively coupled to the processor; and

a baseboard management controller (BMC) communicatively coupled to the processor and the one or more accelerators and configured for out-of-band management of the information handling system, the BMC further configured to:

receive statistical information regarding the one or more accelerators wherein the statistical information comprises, for each of the one or more accelerators, information indicative of one or more of: utilization of such accelerator, number of processes executed by such accelerator, page retirement statistics of such accelerator, uptime of such accelerator, and driver loading state of such accelerator;

determine a criticality factor for each of the one or more accelerators based on the statistical information;

determine an accelerator health status for each of the one or more accelerators; and

determine an overall system health of the statistical information handling system based on the criticality factors and the accelerator health statuses.

2. The information handling system of claim 1 , wherein the statistical information comprises physical information regarding the one or more accelerators.

3. The information handling system of claim 2 , wherein the physical information comprises, for each of the one or more accelerators, one or more of cabling issues associated with such accelerator and bus or communication issues associated with such accelerator.

4. The information handling system of claim 1 , wherein the BMC is further configured to receive the statistical information regarding the one or more accelerators from the one or more accelerators.

5. The information handling system of claim 4 , wherein the BMC is further configured to receive the statistical information regarding the one or more accelerators from at least one of a basic input/output system and a server management module executing on the processor.

6. A baseboard management controller (BMC) configured to communicatively couple to a processor and one or more accelerators, configured for out-of-band management of an information handling system, and further configured to:

receive statistical information regarding the one or more accelerators wherein the statistical information comprises, for each of the one or more accelerators, one or more of: utilization of such accelerator, number of processes executed by such accelerator, page retirement statistics of such accelerator, uptime of such accelerator, and driver loading state of such accelerator;

determine a criticality factor for each of the one or more accelerators based on the statistical information;

determine an accelerator health status for each of the one or more accelerators; and

determine an overall system health of the statistical information handling system based on the criticality factors and the accelerator health statuses.

7. The BMC of claim 6 , wherein the statistical information comprises physical information regarding the one or more accelerators.

8. The BMC of claim 7 , wherein the physical information comprises, for each of the one or more accelerators, one or more of cabling issues associated with such accelerator and bus or communication issues associated with such accelerator.

9. The BMC of claim 6 , further configured to receive the statistical information regarding the one or more accelerators from the one or more accelerators.

10. The BMC of claim 9 , further configured to receive the statistical information regarding the one or more accelerators from at least one of a basic input/output system and a server management module executing on the processor.

11. A method comprising, in an information handling system comprising a processor, one or more accelerators communicatively coupled to the processor, and a baseboard management controller (BMC) communicatively coupled to the processor and the one or more accelerators and configured for out-of-band management of the information handling system:

receiving, by the BMC, statistical information regarding the one or more accelerators wherein the statistical information comprises, for each of the one or more accelerators, one or more of: utilization of such accelerator, number of processes executed by such accelerator, page retirement statistics of such accelerator, uptime of such accelerator, and driver loading state of such accelerator;

determining, by the BMC, a criticality factor for each of the one or more accelerators based on the statistical information;

determining, by the BMC, an accelerator health status for each of the one or more accelerators; and

determining, by the BMC, an overall system health of the statistical information handling system based on the criticality factors and the accelerator health statuses.

12. The method of claim 11 , wherein the statistical information comprises physical information regarding the one or more accelerators.

13. The method of claim 12 , wherein the physical information comprises, for each of the one or more accelerators, one or more of cabling issues associated with such accelerator and bus or communication issues associated with such accelerator.

14. The method of claim 11 , wherein receiving the statistical information comprises receiving the statistical information regarding the one or more accelerators from the one or more accelerators.

15. The method of claim 9 , wherein receiving the statistical information comprises receiving the statistical information regarding the one or more accelerators from at least one of a basic input/output system and a server management module executing on the processor.

Assignments (9)