Abstract: There is disclosed an example of an artificial intelligence (AI) system, including: a first hardware platform; a fabric interface configured to communicatively couple the first hardware platform to a second hardware platform; a processor hosted on the first hardware platform and programmed to operate on an AI problem; and a first training accelerator, including: an accelerator hardware; a platform inter-chip link (ICL) configured to communicatively couple the first training accelerator to a second training accelerator on the first hardware platform without aid of the processor; a fabric ICL to communicatively couple the first training accelerator to a third training accelerator on a second hardware platform without aid of the processor; and a system decoder configured to operate the fabric ICL and platform ICL to share data of the accelerator hardware between the first training accelerator and second and third training accelerators without aid of the processor.

Abstract: There is disclosed an example of an artificial intelligence (AI) system, including: a first hardware platform; a fabric interface configured to communicatively couple the first hardware platform to a second hardware platform; a processor hosted on the first hardware platform and programmed to operate on an AI problem; and a first training accelerator, including: an accelerator hardware; a platform inter-chip link (ICL) configured to communicatively couple the first training accelerator to a second training accelerator on the first hardware platform without aid of the processor; a fabric ICL to communicatively couple the first training accelerator to a third training accelerator on a second hardware platform without aid of the processor; and a system decoder configured to operate the fabric ICL and platform ICL to share data of the accelerator hardware between the first training accelerator and second and third training accelerators without aid of the processor.

Abstract: An error detection and correction method is provided. The method includes: when a pipeline stage error is detected, correcting the pipeline stage error; when it is determined that a plurality of cascaded pipeline stage circuits have continuous pipeline stage errors, stopping all operations of all pipeline stage circuits; flushing the data of the pipeline stage circuits; and re-processing the data of the pipeline stage circuits at a downclocked frequency.

Abstract: Various systems and methods of initiating and performing contextualized AI inferencing, are described herein. In an example, operations performed with a gateway computing device to invoke an inferencing model include receiving and processing a request for an inferencing operation, selecting an implementation of the inferencing model on a remote service based on a model specification and contextual data from the edge device, and executing the selected implementation of the inferencing model, such that results from the inferencing model are provided back to the edge device. Also in an example, operations performed with an edge computing device to request an inferencing model include collecting contextual data, generating an inferencing request, transmitting the inference request to a gateway device, and receiving and processing the results of execution. Further techniques for implementing a registration of the inference model, and invoking particular variants of an inference model, are also described.

Abstract: An error detection and correction method is provided. The method includes: when a pipeline stage error is detected, correcting the pipeline stage error; when it is determined that a plurality of cascaded pipeline stage circuits have continuous pipeline stage errors, stopping all operations of all pipeline stage circuits; flushing the data of the pipeline stage circuits; and re-processing the data of the pipeline stage circuits at a downclocked frequency.

Abstract: There is disclosed an example of an artificial intelligence (AI) system, including: a first hardware platform; a fabric interface configured to communicatively couple the first hardware platform to a second hardware platform; a processor hosted on the first hardware platform and programmed to operate on an AI problem; and a first training accelerator, including: an accelerator hardware; a platform inter-chip link (ICL) configured to communicatively couple the first training accelerator to a second training accelerator on the first hardware platform without aid of the processor; a fabric ICL to communicatively couple the first training accelerator to a third training accelerator on a second hardware platform without aid of the processor; and a system decoder configured to operate the fabric ICL and platform ICL to share data of the accelerator hardware between the first training accelerator and second and third training accelerators without aid of the processor.

Abstract: There is disclosed an example of an artificial intelligence (AI) system, including: a first hardware platform; a fabric interface configured to communicatively couple the first hardware platform to a second hardware platform; a processor hosted on the first hardware platform and programmed to operate on an AI problem; and a first training accelerator, including: an accelerator hardware; a platform inter-chip link (ICL) configured to communicatively couple the first training accelerator to a second training accelerator on the first hardware platform without aid of the processor; a fabric ICL to communicatively couple the first training accelerator to a third training accelerator on a second hardware platform without aid of the processor; and a system decoder configured to operate the fabric ICL and platform ICL to share data of the accelerator hardware between the first training accelerator and second and third training accelerators without aid of the processor.

Abstract: A micro-electro-mechanical system acoustic sensor, a micro-electro-mechanical system package structure and a method for manufacturing the same are provided. The micro-electro-mechanical system acoustic sensor comprises a substrate, a cantilever structure and a diaphragm sensor. The cantilever structure is formed on the substrate, and comprises a fixed end and a free cantilever portion extended from the fixed end. The free cantilever portion comprises a free end. The free end and the fixed end are respectively at opposing sides of the free cantilever portion. The free cantilever portion is capable of generating a vibration wave in an empty space. The diaphragm sensor is formed on the substrate, and comprises a diaphragm film, a back plate, and at least one electrical contact point. The back plate and the diaphragm film have a first empty gap there between. The empty space and the first empty gap communicate to each other.

Abstract: Various systems and methods of initiating and performing contextualized AI inferencing, are described herein. In an example, operations performed with a gateway computing device to invoke an inferencing model include receiving and processing a request for an inferencing operation, selecting an implementation of the inferencing model on a remote service based on a model specification and contextual data from the edge device, and executing the selected implementation of the inferencing model, such that results from the inferencing model are provided back to the edge device. Also in an example, operations performed with an edge computing device to request an inferencing model include collecting contextual data, generating an inferencing request, transmitting the inference request to a gateway device, and receiving and processing the results of execution. Further techniques for implementing a registration of the inference model, and invoking particular variants of an inference model, are also described.

Abstract: There is disclosed an example of an artificial intelligence (AI) system, including: a first hardware platform; a fabric interface configured to communicatively couple the first hardware platform to a second hardware platform; a processor hosted on the first hardware platform and programmed to operate on an AI problem; and a first training accelerator, including: an accelerator hardware; a platform inter-chip link (ICL) configured to communicatively couple the first training accelerator to a second training accelerator on the first hardware platform without aid of the processor; a fabric ICL to communicatively couple the first training accelerator to a third training accelerator on a second hardware platform without aid of the processor; and a system decoder configured to operate the fabric ICL and platform ICL to share data of the accelerator hardware between the first training accelerator and second and third training accelerators without aid of the processor.

Abstract: Various systems and methods of initiating and performing contextualized AI inferencing, are described herein. In an example, operations performed with a gateway computing device to invoke an inferencing model include receiving and processing a request for an inferencing operation, selecting an implementation of the inferencing model on a remote service based on a model specification and contextual data from the edge device, and executing the selected implementation of the inferencing model, such that results from the inferencing model are provided back to the edge device. Also in an example, operations performed with an edge computing device to request an inferencing model include collecting contextual data, generating an inferencing request, transmitting the inference request to a gateway device, and receiving and processing the results of execution. Further techniques for implementing a registration of the inference model, and invoking particular variants of an inference model, are also described.

Abstract: There is disclosed an example of an artificial intelligence (AI) system, including: a first hardware platform; a fabric interface configured to communicatively couple the first hardware platform to a second hardware platform; a processor hosted on the first hardware platform and programmed to operate on an AI problem; and a first training accelerator, including: an accelerator hardware; a platform inter-chip link (ICL) configured to communicatively couple the first training accelerator to a second training accelerator on the first hardware platform without aid of the processor; a fabric ICL to communicatively couple the first training accelerator to a third training accelerator on a second hardware platform without aid of the processor; and a system decoder configured to operate the fabric ICL and platform ICL to share data of the accelerator hardware between the first training accelerator and second and third training accelerators without aid of the processor.

Abstract: A computing method for performing a matrix multiplying-and-accumulating computation by a flash memory array which includes word lines, bit lines and flash memory cells. The computing method includes the following steps: respectively storing a weight value in each of the flash memory cells, receiving a plurality of input voltages via the word lines, performing an computation on one of the input voltages and the weight value by each of the flash memory cells to obtain an output current, outputting the output currents of the flash memory cells via the bit lines, and accumulating the output currents of the flash memory cells connected to the same bit line of the bit lines to obtain a total output current. Each of the flash memory cells is an analog device, and each of the input voltages, each of the output currents and each of the weight values are analog values.

Abstract: An error detection and correction method is provided. The method includes: when a pipeline stage error is detected, correcting the pipeline stage error; when it is determined that a plurality of cascaded pipeline stage circuits have continuous pipeline stage errors, stopping all operations of all pipeline stage circuits; flushing the data of the pipeline stage circuits; and re-processing the data of the pipeline stage circuits at a downclocked frequency.

Abstract: Examples include techniques to manage training or trained models for deep learning applications. Examples include routing commands to configure a training model to be implemented by a training module or configure a trained model to be implemented by an inference module. The commands routed via out-of-band (OOB) link while training data for the training models or input data for the trained models are routed via inband links.

Abstract: Examples include techniques to manage training or trained models for deep learning applications. Examples include routing commands to configure a training model to be implemented by a training module or configure a trained model to be implemented by an inference module. The commands routed via out-of-band (OOB) link while training data for the training models or input data for the trained models are routed via inband links.

Abstract: Various systems and methods of initiating and performing contextualized AI inferencing, are described herein. In an example, operations performed with a gateway computing device to invoke an inferencing model include receiving and processing a request for an inferencing operation, selecting an implementation of the inferencing model on a remote service based on a model specification and contextual data from the edge device, and executing the selected implementation of the inferencing model, such that results from the inferencing model are provided back to the edge device. Also in an example, operations performed with an edge computing device to request an inferencing model include collecting contextual data, generating an inferencing request, transmitting the inference request to a gateway device, and receiving and processing the results of execution. Further techniques for implementing a registration of the inference model, and invoking particular variants of an inference model, are also described.

Abstract: There is disclosed an example of an artificial intelligence (AI) system, including: a first hardware platform; a fabric interface configured to communicatively couple the first hardware platform to a second hardware platform; a processor hosted on the first hardware platform and programmed to operate on an AI problem; and a first training accelerator, including: an accelerator hardware; a platform inter-chip link (ICL) configured to communicatively couple the first training accelerator to a second training accelerator on the first hardware platform without aid of the processor; a fabric ICL to communicatively couple the first training accelerator to a third training accelerator on a second hardware platform without aid of the processor; and a system decoder configured to operate the fabric ICL and platform ICL to share data of the accelerator hardware between the first training accelerator and second and third training accelerators without aid of the processor.

Abstract: A micro-electro-mechanical system acoustic sensor, a micro-electro-mechanical system package structure and a method for manufacturing the same are provided. The micro-electro-mechanical system acoustic sensor comprises a substrate, a cantilever structure and a diaphragm sensor. The cantilever structure is formed on the substrate, and comprises a fixed end and a free cantilever portion extended from the fixed end. The free cantilever portion comprises a free end. The free end and the fixed end are respectively at opposing sides of the free cantilever portion. The free cantilever portion is capable of generating a vibration wave in an empty space. The diaphragm sensor is formed on the substrate, and comprises a diaphragm film, a back plate, and at least one electrical contact point. The back plate and the diaphragm film have a first empty gap there between. The empty space and the first empty gap communicate to each other.

Abstract: There is disclosed an example of an artificial intelligence (AI) system, including: a first hardware platform; a fabric interface configured to communicatively couple the first hardware platform to a second hardware platform; a processor hosted on the first hardware platform and programmed to operate on an AI problem; and a first training accelerator, including: an accelerator hardware; a platform inter-chip link (ICL) configured to communicatively couple the first training accelerator to a second training accelerator on the first hardware platform without aid of the processor; a fabric ICL to communicatively couple the first training accelerator to a third training accelerator on a second hardware platform without aid of the processor; and a system decoder configured to operate the fabric ICL and platform ICL to share data of the accelerator hardware between the first training accelerator and second and third training accelerators without aid of the processor.