18542308. TECHNOLOGIES FOR DYNAMIC ACCELERATOR SELECTION simplified abstract (Intel Corporation)

TECHNOLOGIES FOR DYNAMIC ACCELERATOR SELECTION

Organization Name

Intel Corporation

Inventor(s)

Francesc Guim Bernat of Barcelona (ES)

TECHNOLOGIES FOR DYNAMIC ACCELERATOR SELECTION - A simplified explanation of the abstract

This abstract first appeared for US patent application 18542308 titled 'TECHNOLOGIES FOR DYNAMIC ACCELERATOR SELECTION

Simplified Explanation: The patent application describes a technology for dynamic accelerator selection using a compute sled, which includes a network interface controller to communicate with a remote accelerator over a network. The compute sled can determine whether to accelerate a function and decide whether to offload the function to the remote accelerator based on network telemetry data.

• The compute sled includes a network interface controller with a local accelerator and a compute engine.
• The compute engine obtains network telemetry data to assess network bandwidth saturation.
• It determines whether to accelerate a function and whether to offload it to the remote accelerator.
• If the function is not offloaded, it assigns it to the local accelerator of the network interface controller.

Potential Applications: This technology can be applied in cloud computing, data centers, and distributed computing systems where dynamic accelerator selection is crucial for optimizing performance.

Problems Solved: 1. Efficient selection of accelerators based on network conditions. 2. Improved performance by offloading functions to remote accelerators when necessary. 3. Enhanced utilization of local and remote accelerators in a network environment.

Benefits: 1. Increased efficiency in accelerator selection. 2. Optimal utilization of network resources. 3. Improved overall system performance. 4. Enhanced scalability in distributed computing environments.

Commercial Applications: Dynamic accelerator selection technology can be utilized in cloud service providers, edge computing systems, and high-performance computing clusters to enhance workload management and resource utilization.

Prior Art: Prior research in the field of dynamic accelerator selection and network telemetry data analysis can provide insights into similar technologies and approaches.

Frequently Updated Research: Stay updated on advancements in network telemetry data analysis, dynamic accelerator selection algorithms, and distributed computing systems to enhance the performance and efficiency of this technology.

Questions about Dynamic Accelerator Selection: 1. How does dynamic accelerator selection improve overall system performance? 2. What are the key factors considered in determining whether to offload a function to a remote accelerator?

Original Abstract Submitted

Technologies for dynamic accelerator selection include a compute sled. The compute sled includes a network interface controller to communicate with a remote accelerator of an accelerator sled over a network, where the network interface controller includes a local accelerator and a compute engine. The compute engine is to obtain network telemetry data indicative of a level of bandwidth saturation of the network. The compute engine is also to determine whether to accelerate a function managed by the compute sled. The compute engine is further to determine, in response to a determination to accelerate the function, whether to offload the function to the remote accelerator of the accelerator sled based on the telemetry data. Also the compute engine is to assign, in response a determination not to offload the function to the remote accelerator, the function to the local accelerator of the network interface controller.