Abstract: A method of performing automatic tuning on a deep learning model includes: utilizing an instruction-based learned cost model to estimate a first type of operational performance metrics based on a tuned configuration of layer fusion and tensor tiling; utilizing statistical data gathered during a compilation process of the deep learning model to determine a second type of operational performance metrics based on the tuned configuration of layer fusion and tensor tiling; performing an auto-tuning process to obtain a plurality of optimal configurations based on the first type of operational performance metrics and the second type of operational performance metrics; and configure the deep learning model according to one of the plurality of optimal configurations.

Abstract: A package structure including a photonic, an electronic die, an encapsulant and a waveguide is provided. The photonic die includes an optical coupler. The electronic die is electrically coupled to the photonic die. The encapsulant laterally encapsulates the photonic die and the electronic die. The waveguide is disposed over the encapsulant and includes an upper surface facing away from the encapsulant. The waveguide includes a first end portion and a second end portion, the first end portion is optically coupled to the optical coupler, and the second end portion has a groove on the upper surface.

Abstract: A package includes a photonic layer on a substrate, the photonic layer including a silicon waveguide coupled to a grating coupler; an interconnect structure over the photonic layer; an electronic die and a first dielectric layer over the interconnect structure, where the electronic die is connected to the interconnect structure; a first substrate bonded to the electronic die and the first dielectric layer; a socket attached to a top surface of the first substrate; and a fiber holder coupled to the first substrate through the socket, where the fiber holder includes a prism that re-orients an optical path of an optical signal.

Abstract: A semiconductor package includes a first die stack structure and a second die stack structure, an insulating encapsulation, a redistribution structure, at least one prism structure and at least one reflector. The first die stack structure and the second die stack structure are laterally spaced apart from each other along a first direction, and each of the first die stack structure and the second die stack structure comprises an electronic die; and a photonic die electronically communicating with the electronic die. The insulating encapsulation laterally encapsulates the first die stack structure and the second die stack structure. The redistribution structure is disposed on the first die stack structure, the second die stack structure and the insulating encapsulation, and electrically connected to the first die stack structure and the second die stack structure. The at least one prism structure is disposed within the redistribution structure and optically coupled to the photonic die.

Abstract: A touch panel includes a substrate, first and second touch sensing electrodes, and first and second traces. The substrate includes a display area and a peripheral area. The first and second touch sensing electrodes are formed on the display area, and the first and second traces are formed on the peripheral area. The first touch sensing electrode is formed by a first portion of a metal nanowire layer, which is patterned. The peripheral trace includes a conductive layer and a second portion of the metal nanowire layer, both of which are patterned in a co-etch step. The conductive layer and the second portion of the metal nanowire layer have a coplanar etched surface. The second touch sensing electrode is formed on an insulating layer and connects with the second trace.

Abstract: A flash memory with assistant gates, including two floating gates disposed on a substrate, an insulating layer formed on the two floating gates and the substrate, an assistant gate disposed between the two floating gates, wherein a portion of the assistant gate wraps around the two floating gates, and two select gates disposed respectively outside the two floating gates and partially overlap the two floating gates.

Abstract: A flash memory with assistant gates, including two floating gates disposed on a substrate, an insulating layer formed on the two floating gates and the substrate, an assistant gate disposed between the two floating gates, wherein a portion of the assistant gate wraps around the two floating gates, and two select gates disposed respectively outside the two floating gates and partially overlap the two floating gates.

Abstract: A touch panel includes a substrate, first and second touch sensing electrodes, and first and second traces. The substrate includes a display area and a peripheral area. The first and second touch sensing electrodes are formed on the display area, and the first and second traces are formed on the peripheral area. The first touch sensing electrode is formed by a first portion of a metal nanowire layer, which is patterned. The peripheral trace includes a conductive layer and a second portion of the metal nanowire layer, both of which are patterned in a co-etch step. The conductive layer and the second portion of the metal nanowire layer have a coplanar etched surface. The second touch sensing electrode is formed on an insulating layer and connects with the second trace.

Abstract: A medical alert message handling method for a medical communication device is disclosed. The medical alert message handling method includes receiving a shift, a medical alert message classification configuration and a medical alert message priority configuration, receiving at least one medical alert message, filtering the medical alert message according to the shift to generate a personal alert message, displaying the personal alert message, filtering the medical alert message according to the medical alert message classification configuration and the medical alert message priority configuration, receiving a selection command to select one of the medical alert message, and noting a handling status and a handling problem of the selected medical alert message.

Abstract: A medical alert message handling method for a medical communication device is disclosed. The medical alert message handling method includes receiving a shift, a medical alert message classification configuration and a medical alert message priority configuration, receiving at least one medical alert message, filtering the medical alert message according to the shift to generate a personal alert message, displaying the personal alert message, filtering the medical alert message according to the medical alert message classification configuration and the medical alert message priority configuration, receiving a selection command to select one of the medical alert message, and noting a handling status and a handling problem of the selected medical alert message.

Abstract: A non-volatile memory including the following elements is provided. A first conductive layer and a second conductive layer are disposed on a substrate and separated from each other. A patterned hard mask layer is disposed on the first conductive layer and exposes a sharp tip of the first conductive layer. A third conductive layer is disposed on the substrate at one side of the first conductive layer away from the second conductive layer. The third conductive layer is located on a portion of the first conductive layer and covers the sharp tip, and the third conductive layer and the first conductive layer are isolated from each other. A first doped region is disposed in the substrate below the third conductive layer. A second doped region is disposed in the substrate at one side of the second conductive layer away from the first conductive layer.

Abstract: A method of making a flash memory includes providing a substrate. Then, a first insulating layer, a first conductive layer and a second insulating layer are formed to cover the substrate. Later, a first trench is formed in the first conductive layer and the second insulating layer. After that, a second conductive layer and a mask layer are formed to cover the second insulating layer, and the second conductive layer fills up the first trench. Then, the mask layer are patterned to form patterned mask layers. Subsequently, a spacer is formed on the sidewall of the patterned mask layer. Then, an etching process is carried out by using the patterned mask layers and the spacer as a mask so as to form a first gate structure and a second gate structure.

Abstract: A method for manufacturing a semiconductor device includes the following steps. At first, two gate stack layers are formed on a semiconductor substrate, wherein each of the gate stack layers includes a top surface and two side surfaces. A conductive material layer is deposited to conformally cover the top surface and the two side surfaces of each of the gate stack layers. Then, a cap layer is deposited to conformally cover the conductive material layer. Finally, the cap layer and the conductive material layer above the top surface of each of the gate stack layers are removed to leave the cap layer adjacent to the two side surfaces of each of the gate stack layers and covering a portion of the conductive material layer.

Abstract: A non-volatile memory including the following elements is provided. A first conductive layer and a second conductive layer are disposed on a substrate and separated from each other. A patterned hard mask layer is disposed on the first conductive layer and exposes a sharp tip of the first conductive layer. A third conductive layer is disposed on the substrate at one side of the first conductive layer away from the second conductive layer. The third conductive layer is located on a portion of the first conductive layer and covers the sharp tip, and the third conductive layer and the first conductive layer are isolated from each other. A first doped region is disposed in the substrate below the third conductive layer. A second doped region is disposed in the substrate at one side of the second conductive layer away from the first conductive layer.

Abstract: A non-volatile memory including a substrate, a first stacked structure, a second stacked structure, a fifth conductive layer, a first doped region, and a second doped region is provided. The first stacked structure includes a first conductive layer and a second conductive layer stacked on the substrate in order and isolated from each other. The second stacked structure is separately disposed from the first stacked structure and includes a third conductive layer and a fourth conductive layer stacked on the substrate in order and connected to each other. The fifth conductive layer is disposed on the substrate at one side of the first stacked structure away from the second stacked structure. The first doped region is disposed in the substrate below the fifth conductive layer. The second doped region is disposed in the substrate at one side of the second stacked structure away from the first stacked structure.

Abstract: A method of making a flash memory includes providing a substrate. Then, a first insulating layer, a first conductive layer and a second insulating layer are formed to cover the substrate. Later, a first trench is formed in the first conductive layer and the second insulating layer. After that, a second conductive layer and a mask layer are formed to cover the second insulating layer, and the second conductive layer fills up the first trench. Then, the mask layer are patterned to form patterned mask layers. Subsequently, a spacer is formed on the sidewall of the patterned mask layer. Then, an etching process is carried out by using the patterned mask layers and the spacer as a mask so as to form a first gate structure and a second gate structure.

Abstract: A semiconductor structure and a method for manufacturing the same are provided. The semiconductor structure comprises a substrate, a first gate structure, a second gate electrode, a third gate electrode and a protective layer. The first gate structure comprises a first gate electrode disposed on the substrate and a first gate dielectric covering the first gate electrode. The second gate electrode is disposed on and electrically isolated from the first gate electrode. The first gate structure has an extending portion relative to the second gate electrode. The third gate electrode is disposed adjacent to and electrically isolated from the first gate electrode and the second gate electrode. The third gate has an extending portion between a lower surface of the protective layer and an upper surface of the extending portion of the first gate structure.

Abstract: A method of fabricating a semiconductor device is disclosed. The method includes the steps of: sequentially forming agate dielectric layer and a first gate layer on a semiconductor substrate, wherein the gate dielectric layer is between the first gate layer and the semiconductor substrate; forming at least an opening in the first gate layer; forming a first dielectric layer conformally on the semiconductor substrate wherein the first dielectric layer covers the first gate layer; and forming a second gate layer filling the opening and overlapping the first gate layer.

Abstract: A semiconductor memory device includes a substrate, shallow trench isolations protruding from the substrate, a floating gate formed conformally on the surface of the recess between each shallow trench isolation, a tunnel layer formed between each floating gate and the substrate, a dielectric layer formed conformally on the floating gates, and a control gate formed on the dielectric layer.

Abstract: A method of fabricating a semiconductor device is disclosed. The method includes the steps of: sequentially forming agate dielectric layer and a first gate layer on a semiconductor substrate, wherein the gate dielectric layer is between the first gate layer and the semiconductor substrate; forming at least an opening in the first gate layer; forming a first dielectric layer conformally on the semiconductor substrate wherein the first dielectric layer covers the first gate layer; and forming a second gate layer filling the opening and overlapping the first gate layer.