A discrete three-dimensional (3-D) processor comprises first and second dice. The first die comprises 3-D memory (3D-M) arrays and in-die peripheral-circuit components thereof, whereas the second die comprises processing circuits and off-die peripheral-circuit components of the 3D-M arrays. The first and second dice are communicatively coupled by a plurality of inter-die connections.

A second semiconductor switching element is connected in series with a first semiconductor switching element, and is at least partially stacked on the first semiconductor switching element in the thickness direction. A first control element controls the first semiconductor switching element and the second semiconductor switching element, and performs an overcurrent protection operation with reference to a shunt voltage. The first control element is arranged outside the first semiconductor switching element and the second semiconductor switching element in the in-plane direction.

A discrete three-dimensional (3-D) processor comprises first and second dice. The first die comprises 3-D memory (3D-M) arrays and in-die peripheral-circuit components thereof, whereas the second die comprises processing circuits and off-die peripheral-circuit components of the 3D-M arrays. The first and second dice are communicatively coupled by a plurality of inter-die connections.

A second semiconductor switching element is connected in series with a first semiconductor switching element, and is at least partially stacked on the first semiconductor switching element in the thickness direction. A first control element controls the first semiconductor switching element and the second semiconductor switching element, and performs an overcurrent protection operation with reference to a shunt voltage. The first control element is arranged outside the first semiconductor switching element and the second semiconductor switching element in the in-plane direction.

Some embodiments provide a three-dimensional (3D) circuit structure that has two or more vertically stacked bonded layers with a machine-trained network on at least one bonded layer. As described above, each bonded layer can be an IC die or an IC wafer in some embodiments with different embodiments encompassing different combinations of wafers and dies for the different bonded layers. The machine-trained network in some embodiments includes several stages of machine-trained processing nodes with routing fabric that supplies the outputs of earlier stage nodes to drive the inputs of later stage nodes. In some embodiments, the machine-trained network is a neural network and the processing nodes are neurons of the neural network. In some embodiments, one or more parameters associated with each processing node (e.g., each neuron) is defined through machine-trained processes that define the values of these parameters in order to allow the machine-trained network (e.g., neural network) to perform particular operations (e.g., face recognition, voice recognition, etc.). For example, in some embodiments, the machine-trained parameters are weight values that are used to aggregate (e.g., to sum) several output values of several earlier stage processing nodes to produce an input value for a later stage processing node.

Some embodiments provide a three-dimensional (3D) circuit structure that has two or more vertically stacked bonded layers with a machine-trained network on at least one bonded layer. As described above, each bonded layer can be an IC die or an IC wafer in some embodiments with different embodiments encompassing different combinations of wafers and dies for the different bonded layers. The machine-trained network in some embodiments includes several stages of machine-trained processing nodes with routing fabric that supplies the outputs of earlier stage nodes to drive the inputs of later stage nodes. In some embodiments, the machine-trained network is a neural network and the processing nodes are neurons of the neural network. In some embodiments, one or more parameters associated with each processing node (e.g., each neuron) is defined through machine-trained processes that define the values of these parameters in order to allow the machine-trained network (e.g., neural network) to perform particular operations (e.g., face recognition, voice recognition, etc.). For example, in some embodiments, the machine-trained parameters are weight values that are used to aggregate (e.g., to sum) several output values of several earlier stage processing nodes to produce an input value for a later stage processing node.

A manufacturing method of a semiconductor device includes: a first step of, after joining a wire to an electrode using a capillary, forming a wire part by moving the capillary to a third target pointwhile feeding out the wire; a second step of forming a bent part by moving the capillary to a fourth target point while feeding out the wire; a third step of processing the bent part into a planned cut part by repeating lowering and raising of the capillary for multiple times; and a fourth step of cutting the wire at the planned cut part by raising the capillary with a wire clamper closed to form a pin wire.

A method for manufacturing a semiconductor device is provided, including: preparing a first chip forming portion having a first semiconductor substrate, first metal pads provided at the substrate and a first circuit electrically connected to at least a part of the pads, and a second chip forming portion having a second semiconductor substrate, second metal pads provided at substrate and a second circuit electrically connected to at least a part of the pads; bonding the first and the second chip forming portions while joining the first and the second pads to form a bonding substrate having a non-bonded region between the first and the second chip forming portions at an outer peripheral portion thereof; and filling an insulating film into the non-bonded region, at least a part of the insulating film containing at least one selected from the group consisting of silicon nitride and nitrogen-containing silicon carbide.

A semiconductor device according to an embodiment includes: a bonding substrate which includes a first chip forming portion having first metal pads provided at a semiconductor substrate and a first circuit connected to the first metal pads, and a second chip forming portion having second metal pads joined to the first metal pads and a second circuit connected to the second metal pads and being bonded to the first chip forming portion; and an insulating film which is filled into a non-bonded region between the first chip forming portion and the second chip forming portion at an outer peripheral portion of the bonding substrate. At least a part of the insulating film contains at least one selected from the group consisting of silicon nitride and nitrogen-containing silicon carbide.

A semiconductor device includes a first semiconductor element, a first connection terminal formed on a lower surface of the first semiconductor element, a second semiconductor element mounted on the lower surface of the first semiconductor element so that the second semiconductor element partially overlaps the first semiconductor element in plan view, a second connection terminal formed on a lower surface of the second semiconductor element, and a wiring substrate on which the first and second semiconductor elements are mounted. The wiring substrate includes first and second connection pads electrically connected to the first connection terminal and the second connection terminal, respectively. The semiconductor device further includes a third connection terminal formed on the first connection pad and electrically connected to the first connection terminal. One of the first connection terminal and the third connection terminal is a metal post, and the other is a solder ball.