Granted Patent B1

US 12,206,243 · App. 18/909,389 · Granted Jan 21, 2025

Inventors: Karl Andersen (New York, NY); Vitaly Leokumovich (New York, NY)

Assignee: Greenlight AI LLC

H02J3/003H02J3/381G06F1/3203

Abstract

Claims (40)

1. A smart, bi-directional electrical microgrid comprising:

a) a plurality of processor-on-demand systems, wherein each processor-on-demand system comprises:

1) a computing device including a processing unit and a memory, wherein the processing unit is selected from a central processing unit, a graphics processing unit, a data processing unit, tensor processing unit, neural processing unit, accelerated processing unit, floating point unit, microcontroller unit, vision processing unit, and processing in memory;

2) an Energy Management System (EMS) configured for regulating power usage and optimizing energy efficiency;

3) a distributed power resource configured for providing a stable and efficient energy supply;

4) a database configured to store energy metrics gathered from the processor-on-demand system, the energy metrics including at least energy expenditure of the processor-on-demand system;

5) a Large Language Model (LLM) executing on the computing device, the LLM configured for processing the energy metrics stored in the database so as to generate an energy management plan that is configured to: i) manage when the distributed power resource shall be used by the processor-on-demand system, ii) dynamically adjust computing tasks of the computing device based on current energy availability and predicted workload, and iii) maximize processing unit utilization while adhering to energy efficiency goals;

6) an API gateway comprising an API coupled to a network connection providing access to a communications network, the API gateway configured for providing external systems secure, on-demand access to the processing unit, and wherein the API gateway is configured to manage load on the computing device; and

7) a software module executing on the computing device, the software module configured for managing the processor-on-demand system according to the energy management plan generated by the LLM by: i) controlling usage of the distributed power resource, ii) rate limiting the API gateway and iii) dynamically adjusting the computing tasks of the computing device;

b) one or more management servers configured for managing the delivery and distribution of power among the plurality of processor-on-demand systems to optimize for efficiency and uptime; and

c) a network of power lines that interconnect the plurality of processor-on-demand systems.

2. The electrical microgrid of claim 1 , wherein the EMS utilizes predictive analytics to forecast energy demand of the computing device and proactively adjust power usage of the computing device.

3. The electrical microgrid of claim 2 , wherein the distributed power resource further comprises a battery and a battery management system (BMS), wherein the battery comprises battery cells, and the BMS includes algorithms for state-of-charge (SoC) and state-of-health (SoH) monitoring to extend battery life and performance.

4. The electrical microgrid of claim 3 , wherein the distributed power resource further includes renewable energy sources integrated with the BMS to prioritize the use of renewable energy.

5. The electrical microgrid of claim 4 , wherein database is further configured to encrypt sensitive data, including energy consumption patterns and operational parameters, ensuring data privacy and security.

6. The electrical microgrid of claim 5 , wherein the energy management plan is configured to optimize energy consumption and reduce operational costs.

7. The electrical microgrid of claim 6 , wherein the energy management plan is configured to dynamically adjust computing tasks of the computing device based on current energy availability and predicted workload, to maximize processing unit utilization while adhering to energy efficiency goals.

8. The electrical microgrid of claim 6 , wherein the API gateway comprises security features including encryption and authentication protocols to ensure secure data transmission and access control.

9. The electrical microgrid of claim 8 , wherein the API gateway supports multiple protocols for integration with various external systems, including REST, GraphQL, and WebSocket, so as to facilitate on-demand processing services.

10. The electrical microgrid of claim 9 , wherein the software module includes a user interface (UI) that provides real-time monitoring and control over the energy management plan, allowing users to manually adjust operational parameters and set energy efficiency goals.

11. The electrical microgrid of claim 10 , wherein the software module provides automated reports on energy savings and system performance that allow users to track improvements over time and identify areas for further optimization.

12. The electrical microgrid of claim 11 , further comprising a second power source connected to the EMS and the computing device.

13. The electrical microgrid of claim 12 , wherein the EMS is further configured to automatically switch between the power source and the second power source based on predefined criteria, including cost of electricity and carbon footprint, so as to meet defined energy efficiency goals.

14. The electrical microgrid of claim 13 , wherein the LLM is further configured for identifying problem issues and providing troubleshooting advice to a user.

15. The electrical microgrid of claim 13 , wherein the one or more management servers are further configured to: a) monitor real-time energy production and consumption data from each of the plurality of processor-on-demand systems; b) utilize advanced predictive algorithms to forecast energy demands based on historical data and current computational load; c) dynamically adjust distribution of power among the plurality of processor-on-demand systems to optimize for energy efficiency and system uptime; d) coordinate charging and discharging cycles of batteries within the plurality of processor-on-demand systems to maximize battery lifespan and efficiency; e) integrate data from weather forecasting systems to anticipate changes in energy production from renewable sources, and adjust the microgrid operations accordingly; f) provide a user interface that allows operators to manually override automated controls and make adjustments to energy distribution and computational load management based on operational need.

16. The electrical microgrid of claim 13 , wherein the network of power lines is further configured to support bi-directional flow of electricity, allowing distribution of power to and from each of the plurality of processor-on-demand systems.

17. A method on a server for managing a smart, bi-directional electrical microgrid of a plurality of processor-on-demand systems, the method comprising:

a) collecting energy metrics from each of the plurality of processor-on-demand systems, wherein each processor-on-demand system includes:

1) a computing device with a processing unit and a memory, wherein the processing unit is selected from a central processing unit, a graphics processing unit, a data processing unit, tensor processing unit, neural processing unit, accelerated processing unit, floating point unit, microcontroller unit, vision processing unit, and processing in memory;

2) an Energy Management System (EMS) configured for regulating power usage and optimizing energy efficiency;

3) a distributed power resource providing a stable and efficient energy supply;

4) a database for storing energy metrics, including at least energy expenditure;

5) a Large Language Model (LLM) configured for processing the energy metrics so as to generate an energy management plan that is configured to: i) manage when the distributed power resource shall be used by the processor-on-demand system, ii) dynamically adjust computing tasks of the computing device based on current energy availability and predicted workload, and iii) maximize processing unit utilization while adhering to energy efficiency goals;

6) an API gateway comprising an API coupled to a network connection providing access to a communications network, the API gateway configured for providing external systems secure, on-demand access to the processing unit, and wherein the API gateway is configured to manage load on the computing device; and

7) a software module configured for managing the processor-on-demand system according to the energy management plan generated by the LLM by: i) controlling usage of the distributed power resource, ii) rate limiting the API gateway and iii) dynamically adjusting the computing tasks of the computing device;

b) reading the collected energy metrics from each of the plurality of processor-on-demand systems;

c) transmitting control signals from a centralized management server to each of the plurality of processor-on-demand systems to adjust power usage dynamically and optimize distribution of energy resources across the microgrid;

d) monitoring and adjusting operational parameters of each of the plurality of processor-on-demand systems in real-time to ensure continuous optimization of energy efficiency and computational performance; and

e) managing interconnection and data exchange between the plurality of processor-on-demand systems via a network of power lines, ensuring bi-directional flow and dynamic routing of power based on system-wide energy management strategies.

18. The method of claim 17 , wherein the control signals are configured for coordinating energy distribution adjustments across the plurality of processor-on-demand systems in order to maintain system balance during peak demand periods and in response to microgrid instability.

Assignments (2)

Continuity (2)

References Cited (39)