View source for 17525300. Hardware Accelerator Service Aggregation simplified abstract (Google LLC)

=Hardware Accelerator Service Aggregation=


==Organization Name==

[[:Category:Google LLC|Google LLC]]


[[Category:Google LLC]]

==Inventor(s)==

[[:Category:Shrikant Kelkar of San Jose CA (US)|Shrikant Kelkar of San Jose CA (US)]][[Category:Shrikant Kelkar of San Jose CA (US)]]

[[:Category:Gargi Adhav of San Jose CA (US)|Gargi Adhav of San Jose CA (US)]][[Category:Gargi Adhav of San Jose CA (US)]]

[[:Category:Lakshmi Sharma of Saratoga CA (US)|Lakshmi Sharma of Saratoga CA (US)]][[Category:Lakshmi Sharma of Saratoga CA (US)]]

[[:Category:Manoj Jayadevan of Cupertino CA (US)|Manoj Jayadevan of Cupertino CA (US)]][[Category:Manoj Jayadevan of Cupertino CA (US)]]

[[:Category:Parveen Patel of Cupertino CA (US)|Parveen Patel of Cupertino CA (US)]][[Category:Parveen Patel of Cupertino CA (US)]]

[[:Category:Parthasarathy Ranganathan of San Jose CA (US)|Parthasarathy Ranganathan of San Jose CA (US)]][[Category:Parthasarathy Ranganathan of San Jose CA (US)]]

==Hardware Accelerator Service Aggregation - A simplified explanation of the abstract==


This abstract first appeared for US patent application 17525300 titled 'Hardware Accelerator Service Aggregation

==Simplified Explanation==

The present disclosure describes a system and method for discovering the capabilities of local and remote hardware accelerator cards. These accelerator cards can provide acceleration services for processing workloads.

* The local hardware accelerator card can provide a list of acceleration services it offers, including services provided by accelerators on the card itself and services provided by accelerators on a remote hardware accelerator card.
* A workload instruction, which defines a workload to be processed, is received from a processor of a computing device.
* The workload instruction is then forwarded to the remote hardware accelerator card for processing by one of its acceleration services.

Potential applications of this technology:

* High-performance computing: This technology can be used in applications that require intensive computational tasks, such as scientific simulations, data analysis, and machine learning.
* Cloud computing: By discovering and utilizing the capabilities of remote hardware accelerator cards, cloud service providers can offer more efficient and powerful computing resources to their customers.
* Edge computing: This technology can enable edge devices to offload computationally intensive tasks to remote hardware accelerator cards, improving the performance and efficiency of edge computing systems.

Problems solved by this technology:

* Discovery of hardware accelerator capabilities: This technology provides a mechanism for discovering the available acceleration services provided by both local and remote hardware accelerator cards, allowing for efficient utilization of these resources.
* Offloading of workloads: By forwarding workload instructions to remote hardware accelerator cards, this technology enables the offloading of computationally intensive tasks, reducing the load on the local processor and improving overall system performance.

Benefits of this technology:

* Improved performance: By utilizing hardware accelerator cards, workloads can be processed more quickly and efficiently, leading to improved overall system performance.
* Resource utilization: This technology allows for the efficient utilization of both local and remote hardware accelerator cards, maximizing the available computing resources.
* Scalability: By leveraging remote hardware accelerator cards, this technology enables the scaling of computational capabilities without the need for additional local hardware.


==Original Abstract Submitted==

The present disclosure includes systems, methods, and computer-readable mediums for discovering capabilities of local and remote hardware (HW) accelerator cards. A local hardware (HW) accelerator card may provide, via a communication interface, a listing of acceleration services from the local HW accelerator card. The listing of acceleration services may include a first set of acceleration services provided by one or more accelerators of the local HW accelerator card and a second set of acceleration services provided by one or more accelerators of a remote HW accelerator card. A workload instruction defining a workload for processing by at least one of the acceleration services of the second set of acceleration services may be received from a processor of a computing device. The workload instruction may be forwarded to the remote HW accelerator card.



[[Category:G06F9/50]]
[[Category:G06F9/38]]