Accelerating artificial intelligence (‘AI’) workflows

1. A method comprising:

receiving, from a computing process of an artificial intelligence workflow, a request for information stored on a data repository;

issuing parallel remote procedure calls from a user space of an operating system environment to the data repository using a network protocol that operates serially at the kernel level of the operating system environment within a distributed network, the parallel remote procedure calls comprising respective remote procedure calls for respective files of a plurality of files, wherein the respective remote procedure calls are directed to a selected representative subset of the requested information;

receiving, from the data repository, one or more responses to the parallel remote procedure calls; and

training an artificial intelligence model based on the one or more responses to the parallel remote procedure calls.

2. The method of claim 1 , wherein the artificial intelligence workflow executes within an artificial intelligence and machine learning infrastructure.

3. The method of claim 2 , wherein the computing process executes within the artificial intelligence and machine learning infrastructure on one or more of: a graphical processing unit, a central processing unit, or a tensor processing unit.

4. The method of claim 2 , wherein the data repository is included within one or more storage systems within the artificial intelligence and machine learning infrastructure.

5. The method of claim 1 , wherein the computing process executes within a client computing device that is connected over a network to a server, wherein the server includes the data repository, and wherein the network protocol is a Network File System (NFS) protocol or a Server Message Block (SMB) protocol.

6. The method of claim 1 , further comprising:

issuing one or more remote procedure calls for file information and directory information to the data repository; and

selecting, based on the file information and directory information received from the data repository, a subset of files or a subset of directories.

7. The method of claim 6 , wherein issuing the parallel remote procedure calls includes generating respective remote procedure calls for respective files of the subset of files.

8. The method of claim 1 , further comprising:

issuing a first request to determine a structure of the data repository; and

receiving a first response that includes a directory structure of the data repository, and

wherein issuing parallel remote procedure calls further comprises based on the first response, issuing the parallel remote procedure calls from the user space of the operating system environment to the data repository using the network protocol that operates serially at the kernel level of the operating system environment, the parallel remote procedure calls comprising respective remote procedure calls for respective files that are identified using the directory structure.

9. An artificial intelligence and machine learning infrastructure system comprising:

a data repository including one or more storage systems comprising, respectively, one or more storage devices; and

one or more graphical processing units, wherein the graphical processing units are configured to communicate with the one or more storage systems over a communication fabric;

wherein an artificial intelligence workflow executing on at least one of the one or more graphical processing units includes:

receiving, from a computing process of an artificial intelligence workflow, a request for information stored on a data repository;

issuing parallel remote procedure calls from a user space of an operating system environment to the data repository using a network protocol that operates serially at the kernel level of the operating system environment within a distributed network, the parallel remote procedure calls comprising respective remote procedure calls for respective files of a plurality of files, wherein the respective remote procedure calls are directed to a selected representative subset of the requested information;

receiving, from the data repository, one or more responses to the parallel remote procedure calls; and

training an artificial intelligence model based on the one or more responses to the parallel remote procedure calls.

10. The artificial intelligence and machine learning infrastructure system of claim 9 , wherein the one or more storage systems and the one or more graphical processing units are implemented within a single chassis.

11. The artificial intelligence and machine learning infrastructure system of claim 9 , wherein the network protocol is a Network File System (NFS) protocol.

12. The artificial intelligence and machine learning infrastructure system of claim 9 , wherein the artificial intelligence workflow further includes:

issuing one or more remote procedure calls for file information and directory information to the data repository; and

selecting, based on the file information and directory information received from the data repository, a subset of files or a subset of directories.

13. The artificial intelligence and machine learning infrastructure system of claim 9 , wherein issuing the parallel remote procedure calls includes generating respective remote procedure calls for respective files of the subset of files.

14. An apparatus comprising a computer processor, a computer memory operatively coupled to the computer processor, the computer memory having disposed within it computer program instructions that, when executed by the computer processor, cause the apparatus to carry out the steps of:

receiving, from a computing process of an artificial intelligence workflow, a request for information stored on a data repository;

issuing parallel remote procedure calls from a user space of an operating system environment to the data repository using a network protocol that operates serially at the kernel level of the operating system environment within a distributed network, the parallel remote procedure calls comprising respective remote procedure calls for respective files of a plurality of files, wherein the respective remote procedure calls are directed to a selected representative subset of the requested information;

receiving, from the data repository, one or more responses to the parallel remote procedure calls; and

training an artificial intelligence model based on the one or more responses to the parallel remote procedure calls.

15. The apparatus of claim 14 , wherein the artificial intelligence workflow executes within an artificial intelligence and machine learning infrastructure.

16. The apparatus of claim 15 , wherein the computing process executes within a graphical processing unit within the artificial intelligence and machine learning infrastructure.

17. The apparatus of claim 15 , wherein the data repository is included within one or more storage systems within the artificial intelligence and machine learning infrastructure.

18. The apparatus of claim 14 , wherein the computing process executes within a client computing device that is connected over a network to a server, wherein the server includes the data repository, and wherein the network protocol is a Network File System (NFS) protocol.

19. The apparatus of claim 14 , wherein the program instructions, when executed by the computer processor, further cause the storage system to carry out the steps of:

issuing one or more remote procedure calls for file information and directory information to the data repository; and

selecting, based on the file information and directory information received from the data repository, a subset of files or a subset of directories.

20. The apparatus of claim 19 , wherein issuing the parallel remote procedure calls includes generating respective remote procedure calls for respective files of the subset of files.