# US 2019/0377403 A1 · Systems and methods for autonomous hardware compute resiliency > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2019/0377403 A1 - Document ID: US20190377403A1 - Status: Published application - Kind: A1 - Application number: 16/435,470 - Filed: 2019-06-08 - Published: 2019-12-12 - Earliest claimed: 2018-06-11 - Assignees: Lucid Circuit, Inc. - Applicants: Lucid Circuit, Inc. - Inventors: Michel D. Sika - Web page: https://patentgrep.com/patent/US20190377403A1 ## CPC - G06F 1/3293: by switching to a less power-consuming processor, e.g. sub-CPU - G06F 1/206: {comprising thermal management} - G06N 5/04: Inference or reasoning models - G06N 20/00: Machine learning - G06N 20/10: using kernel methods, e.g. support vector machines [SVM] - H04L 9/3247: {involving digital signatures} - H04L 9/3278: {using physically unclonable functions [PUF]} - H04L 67/12: specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks - G06F 1/3203: Power management, i.e. event-based initiation of a power-saving mode - Y02D 10/00: Energy efficient computing, e.g. low power processors, power management or thermal management - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M) - Y02D: CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE ## Abstract Methods and systems for providing hardware compute resiliency by using a compute fabric that includes sensors and re-programmable data processing components. ## Claims ### Claim 1 A method for controlling a system that includes a data collection system, at least one sensor, and a hardware compute fabric, the method comprising: the data collection system generating data; a first processing component of the compute fabric performing a first sub-process of a data analysis process for the generated data and storing results of the first sub-process at the compute fabric; a third processing component programming a second processing component of the compute fabric to perform the first sub-process and disabling the first sub-process at the first processing component, based on at least one of environment information generated by using the at least one sensor, power consumption information generated by the compute fabric, processing performance information for the data analysis process generated by the compute fabric, and fault detection information for the data analysis process generated by the compute fabric; the second processing component performing the first sub-process for the generated data and storing results of the first sub-process generated by the second processing component at the compute fabric; the compute fabric generating data analytics results by using the stored results of the first sub-process generated by the first processing component and the stored results of the first sub-process generated by the second processing component; and a data transmitter system of the compute fabric transmitting the generated data analytics results to at least one of a terrestrial terminal system and an external second compute fabric. ### Claim 2 (depends on 1) The method of claim 1, further comprising: the at least one sensor generating the environment information for at least a first region of a microelectronic device package that includes the first processing component and a second region of the microelectronic device that includes the second processing component, wherein the environment information includes at least one of environment radiation information and environment temperature information. ### Claim 3 (depends on 2) The method of claim 2, further comprising the third processing component determining that compute performance of the first processing component is susceptible to environmental impact and that the second region is less susceptible to environmental impact than the first region based on the generated environment information, wherein the third processing component programs the second processing component to perform the first sub-process and disables the first sub-process at the first processing component responsive to the determination that compute performance of the first processing component is susceptible to environmental impact and that the second region is less susceptible to environmental impact than the first region. ### Claim 4 (depends on 1) The method of claim 1, further comprising: a fourth processing component of the compute fabric performing a second sub-process of the data analysis process for the generated data and storing results of the second sub-process at the compute fabric, wherein the compute fabric generates the data analytics results by using the stored results of the first sub-process generated by the first processing component, the stored results of the first sub-process generated by the second processing component, and the stored results of the second sub-process generated by the fourth processing component. ### Claim 5 (depends on 4) The method of claim 4, further comprising: the third processing component programming the fourth processing component to perform a third sub-process of the data analysis process and disabling the second sub-process at the fourth processing component, based on at least one of the power consumption information generated by the compute fabric, processing performance information for the data analysis process generated by the compute fabric, fault detection information for the data analysis process generated by the compute fabric, sensor activity information of the first sensor system generated by the compute fabric, and processing activity information of the processing components generated by the compute fabric, wherein the compute fabric generates the data analytics results by using the stored results of the first sub-process generated by the first processing component, the stored results of the first sub-process generated by the second processing component, the stored results of the second sub-process generated by the fourth processing component, and the stored results of the third sub-process generated by the fourth processing component. ### Claim 6 (depends on 4) The method of claim 4, further comprising: the third processing component disabling the second sub-process at the fourth processing component, based on at least one of the power consumption information generated by the compute fabric, processing performance information for the data analysis process generated by the compute fabric, fault detection information for the data analysis process generated by the compute fabric, sensor activity information of the first sensor system generated by the compute fabric, and processing activity information of the processing components generated by the compute fabric. ### Claim 7 (depends on 1) The method of claim 1, wherein the data collection system is an image sensor system that is constructed to generate image data, wherein the data analysis process is an image analysis process, and wherein the image analysis process includes at least one machine learning sub-process. ### Claim 8 (depends on 1) The method of claim 1, wherein the compute fabric tags the data analytics results with identifying information unique to intrinsic physical specificities of at least one of the first sensor system and the compute fabric. ### Claim 9 (depends on 8) The method of claim 8, wherein the identifying information is a Physically Unclonable Function (PUF). ### Claim 10 (depends on 8) The method of claim 8, wherein the identifying information is used to generate a Physically Unclonable Function (PUF). ### Claim 11 (depends on 8) The method of claim 8, wherein the compute fabric tags the data analytics results with information identifying an intended recipient of the analytics results. ### Claim 12 (depends on 8) The method of claim 8, wherein the compute fabric uses the identifying information to generate a digital signature, and signs the analytics results with the generated digital signature. ### Claim 13 (depends on 8) The method of claim 8, wherein the compute fabric encrypts the analytics results analytics results with a public key of an intended recipient of the analytics data. ### Claim 14 (depends on 8) The method of claim 8, wherein the compute fabric generates semiconductor process performance characteristic identifying information. ### Claim 15 (depends on 1) The method of claim 1, wherein the compute fabric disables the first processing component responsive to a determination that the first processing component is not generating data. ### Claim 16 (depends on 1) The method of claim 1, wherein the compute fabric disables the first processing component, responsive to a determination that a sensor communicatively coupled to the first processing component is not generating sensor data. ### Claim 17 (depends on 1) The method of claim 1, wherein the third processing component dynamically reconfigures the compute fabric to improve power consumption while performing the data analysis process, based on compute fabric power consumption information generated by a power monitoring compute fabric component. ### Claim 18 (depends on 1) The method of claim 1, wherein the third processing component dynamically reconfigures the compute fabric to reduce temperature of the compute fabric while performing the data analysis process, based on compute fabric temperature information generated by a temperature monitoring compute fabric component. ### Claim 19 (depends on 1) The method of claim 1, wherein the third processing component dynamically reconfigures the compute fabric to increase processing speed of the compute fabric while performing the data analysis process, based on compute fabric processing speed information generated by a processing speed monitoring compute fabric component. ### Claim 20 (depends on 1) The method of claim 1, wherein the data analysis process is partitioned into a first plurality of sub-processes, and wherein each of a first plurality of processing components of the compute fabric performs a respective one of the first plurality of sub-processes of the data analysis process. ### Claim 21 (depends on 20) The method of claim 20, wherein the third processing component re-partitions the data analysis process into a second plurality of sub-processes based on at least one of the power consumption information generated by the compute fabric, processing performance information for the data analysis process generated by the compute fabric, fault detection information for the data analysis process generated by the compute fabric, sensor activity information of the first sensor system generated by the compute fabric, and processing activity information of the processing components generated by the compute fabric, and wherein the third processing component reprograms at least one processing component during processing of the data analysis process such that each of a second plurality of processing components of the compute fabric performs a respective one of the second plurality of sub-processes of the data analysis process. ### Claim 22 (depends on 20) The method of claim 20, wherein the third processing component re-distributes the data analysis process into a second plurality of sub-processes based on at least one of the power consumption information generated by the compute fabric, processing performance information for the data analysis process generated by the compute fabric, fault detection information for the data analysis process generated by the compute fabric, sensor activity information of the first sensor system generated by the compute fabric, and processing activity information of the processing components generated by the compute fabric, and wherein the third processing component reprograms at least one processing component during processing of the data analysis process such that each of a second plurality of processing components of the compute fabric performs a respective one of the second plurality of sub-processes of the data analysis process.