MUMBAI, India, May 29 -- Intellectual Property India has published a patent application (202621035393 A) filed by Dr. Ashwin I. Mehta, Maharashtra, on March 24, for 'a method and system for thermal-score based intelligent load balancing scheduling for heterogeneous accelerators data center resource management.'
Inventor(s) include Dr. Ashwin I. Mehta.
The application for the patent was published on May 29, under issue no. 22/2026.
According to the abstract released by the Intellectual Property India: "A method and system for thermally optimized workload scheduling in heterogeneous data center processors are disclosed. The method comprises monitoring computational load and estimating processor temperature based on a thermal model associated with each processing unit. A thermal score is computed using a ratio of predicted processor temperature to a maximum allowable operating temperature. Incoming computational tasks are dynamically assigned to processing units including central processing units (CPU), graphics processing units (GPU), tensor processing units (TPU), and neural processing units (NPU) based on the computed thermal score so as to minimize thermal hotspots and distribute heat generation across the data center infrastructure. The approach reduces localized heat concentration, improves resource utilization, and lowers cooling energy requirements. The system operates in real-time scheduling environments within cloud computing infrastructures, artificial intelligence platforms, and high-performance computing data centers."
Disclaimer: Curated by HT Syndication.
