18241062. METHODS AND APPARATUS TO PERFORM CLOUD-BASED ARTIFICIAL INTELLIGENCE OVERCLOCKING simplified abstract (Intel Corporation)

METHODS AND APPARATUS TO PERFORM CLOUD-BASED ARTIFICIAL INTELLIGENCE OVERCLOCKING

Organization Name

Intel Corporation

Inventor(s)

Xia Zhu of Portland OR (US)

Jianfang Zhu of Portland OR (US)

Taeyoung Kim of Fremont CA (US)

Daniel J. Ragland of Sherwood OR (US)

Rodny Rodriguez of Beaverton OR (US)

Louis Draghi of Hillsboro OR (US)

Matthew Paul Fife of Portland OR (US)

Jason Xili Xie of Portland OR (US)

METHODS AND APPARATUS TO PERFORM CLOUD-BASED ARTIFICIAL INTELLIGENCE OVERCLOCKING - A simplified explanation of the abstract

This abstract first appeared for US patent application 18241062 titled 'METHODS AND APPARATUS TO PERFORM CLOUD-BASED ARTIFICIAL INTELLIGENCE OVERCLOCKING

Simplified Explanation

The abstract of this patent application describes a cloud-based artificial intelligence overclocking system. Here is a simplified explanation:

• The patent application discloses a system that collects information about a workload to be executed by a compute platform.
• Based on this workload information, the system generates an output from an artificial intelligence model.
• The output represents a frequency value for overclocking, which is used to optimize the operation of the compute platform.
• The system then performs overclocking on the compute platform using the determined frequency value.

Potential Applications:

• Cloud computing: This technology can be applied in cloud computing environments to optimize the performance of compute platforms by dynamically adjusting their overclocking frequency based on workload information.
• Artificial intelligence: The system can be used to enhance the performance of AI models by overclocking the compute platforms they run on, leading to faster processing and improved efficiency.

Problems Solved:

• Performance optimization: The technology addresses the problem of maximizing the performance of compute platforms by dynamically adjusting their overclocking frequency based on workload characteristics.
• Workload efficiency: By overclocking the compute platform, the technology aims to improve the efficiency of executing workloads, leading to faster processing times and reduced resource utilization.

Benefits:

• Improved performance: By overclocking the compute platform, the technology can enhance the performance of AI models and other workloads, resulting in faster processing and improved overall system performance.
• Resource efficiency: By dynamically adjusting the overclocking frequency based on workload information, the technology helps optimize resource utilization, leading to improved efficiency and cost savings in cloud computing environments.

Original Abstract Submitted

Systems, apparatus, articles of manufacture, and methods to perform cloud-based artificial intelligence overclocking are disclosed that, collect workload information, the workload information representing a workload to be executed by the first compute platform, cause generation of an output from an AI model based on the workload information, the output representing an overclocking frequency value to be used for operation of the first compute platform, and performing overclocking based on the overclocking frequency value.