Edge AI accelerator service

1. A system, comprising:

at least one computing device comprising at least one processor and a data store; and

the data store comprising executable instructions, wherein the instructions, when executed by the at least one processor, cause the at least one computing device to at least:

transmit, from an artificial intelligence (AI) client to an AI broker service, a plurality of periodically-transmitted AI accelerator heartbeat messages that enable tracking availability statuses for at least one AI accelerator connected to an edge device executing the AI client, wherein an AI accelerator heartbeat message comprises: an AI accelerator device identifier of the AI accelerator, a unique device identifier of the edge device to which the AI accelerator is connected, a hardware address of the AI accelerator, an address of the edge device, and an AI technique type used by the AI accelerator, wherein the AI accelerator heartbeat message registers the AI accelerator with the AI broker service, wherein the AI accelerator is locally connected to a bus of the edge device;

receive, by the AI client from an AI agent executed using a networked computing device, an AI processing request comprising the AI accelerator device identifier of the AI accelerator and the AI technique type, wherein the AI agent selects the AI accelerator for an AI workload based at least in part on the AI accelerator being associated with a shortest physical distance among a plurality of available AI accelerators identified in a list of AI accelerators received from the AI broker service;

enable a bus redirect of the AI accelerator that intercepts bus traffic for the AI accelerator and redirects the bus traffic from the edge device to the networked computing device over a network; and

perform, using the AI technique type specified in the AI processing request, the AI workload on the AI accelerator connected to the edge device, the AI accelerator being controlled by the networked computing device using the bus redirect.

2. The system of claim 1 , wherein the AI accelerator is connected to the computing device through a USB connection or a mini PCI-e connection.

3. The system of claim 1 , wherein the AI accelerator heartbeat message comprises information for the AI accelerator based at least in part on detection of a connection of the AI accelerator to the computing device.

4. The system of claim 1 , wherein the AI agent provides a headless redirect functionality.

5. The system of claim 1 , wherein the bus redirect comprises a universal serial bus (USB) redirect.

6. The system of claim 1 , wherein the AI accelerator is accessible to the networked computing device and inaccessible to the computing device based on the bus redirect.

7. The system of claim 1 , wherein the instructions, when executed by the at least one processor, further cause the computing device to at least:

disable the bus redirect of the AI accelerator to the networked computing device, wherein the AI accelerator becomes accessible to the computing device based on the bus redirect being disabled.

8. A non-transitory computer-readable medium comprising executable instructions, wherein the instructions, when executed by at least one processor, cause at least one computing device to at least:

transmit, from an artificial intelligence (AI) client to an AI broker service, a plurality of periodically-transmitted AI accelerator heartbeat messages that enable tracking availability statuses for at least one AI accelerator connected to a local device executing the AI client, wherein an AI accelerator heartbeat message comprises: an AI accelerator device identifier of the AI accelerator, a unique device identifier of the local device, an address of the local device, and an AI technique type used by the AI accelerator, wherein the AI accelerator heartbeat message registers the AI accelerator with the AI broker service, wherein the AI accelerator is locally connected to a bus of the local device;

receive, by the AI client from an AI agent executed using a networked computing device, an AI processing request comprising the AI accelerator device identifier of the AI accelerator and one of the AI algorithm type or the AI technique type, wherein the AI agent selects the AI accelerator for an AI workload based at least in part on the AI accelerator being associated with a shortest physical distance among a plurality of available AI accelerators identified in a list of AI accelerators received from the AI broker service;

enable a bus redirect of the AI accelerator that intercepts bus traffic for the AI accelerator and redirects the bus traffic from the local device to the networked computing device over a network; and

perform, using the one of the AI algorithm type or the AI technique type specified in the AI processing request, the AI workload on the AI accelerator connected to the local device, the AI accelerator being controlled by the networked computing device using the bus redirect.

9. The non-transitory computer-readable medium of claim 8 , wherein the AI accelerator is connected to the computing device through a USB connection or a mini PCI-e connection.

10. The non-transitory computer-readable medium of claim 8 , wherein the AI accelerator heartbeat message comprises information for the AI accelerator based at least in part on detection of a connection of the AI accelerator to the computing device.

11. The non-transitory computer-readable medium of claim 8 , wherein the AI agent provides a headless redirect functionality.

12. The non-transitory computer-readable medium of claim 8 , wherein the bus redirect comprises a universal serial bus (USB) redirect.

13. The non-transitory computer-readable medium of claim 8 , wherein the AI accelerator is accessible to the networked computing device and inaccessible to the computing device based on the bus redirect.

14. The non-transitory computer-readable medium of claim 13 , wherein the instructions, when executed by the at least one processor, further cause the computing device to at least:

disable the bus redirect of the AI accelerator to the networked computing device, wherein the AI accelerator becomes accessible to the computing device based on the bus redirect being disabled.

15. A method performed by instructions executed by at least one computing device, the method comprising:

transmitting, from an artificial intelligence (AI) client to an AI broker service, a plurality of periodically-transmitted AI accelerator heartbeat messages that enable tracking availability statuses for at least one AI accelerator connected to a local device executing the AI client, wherein an AI accelerator heartbeat message comprises: an AI accelerator device identifier of the AI accelerator, a unique device identifier of the local device, and an AI technique type used by the AI accelerator, wherein the AI accelerator heartbeat message is registers the AI accelerator with the AI broker service;

receiving, by the AI client from an AI agent executed using a networked computing device, an AI processing request comprising the AI accelerator device identifier of the AI accelerator and the AI algorithm type, wherein the AI agent identifies a selection of the AI accelerator for an AI workload based at least in part on a plurality of available AI accelerators identified in a list of AI accelerators received from the AI broker service;

enabling a bus redirect of the AI accelerator that intercepts bus traffic for the AI accelerator and redirects the bus traffic from the local device to the networked computing device over a network; and

performing, using the AI algorithm type specified in the AI processing request, the AI workload on the AI accelerator connected to the local device, the AI accelerator being controlled by the networked computing device using the bus redirect.

16. The method of claim 15 , wherein the AI accelerator is connected to the computing device through a USB connection or a mini PCI-e connection.

17. The method of claim 15 , wherein the AI accelerator heartbeat message comprises information for the AI accelerator based at least in part on detection of a connection of the AI accelerator to the computing device.

18. The method of claim 15 , wherein the AI agent provides a graphical user interface that shows the list of AI accelerators, and the selection of the AI accelerator for the AI workload is identified based at least in part on the graphical user interface.

19. The method of claim 15 , wherein the bus redirect comprises a universal serial bus (USB) redirect.

20. The method of claim 15 , wherein the AI accelerator is accessible to the networked computing device and inaccessible to the computing device based on the bus redirect.