A semiconductor memory device includes a memory cell array including a plurality of memory cell rows, a row hammer management circuit and a refresh control circuit. The row hammer management circuit counts the number of times of access associated with each of the plurality of memory cell rows in response to an active command from an external memory controller to store the counted values in each of the plurality of memory cell rows as count data, determines a hammer address associated with at least one of the plurality of memory cell rows, which is intensively accessed more than a predetermined reference number of times, based on the counted values, and performs an internal read-update-write operation. The refresh control circuit receives the hammer address and to perform a hammer refresh operation on victim memory cell rows which are physically adjacent to a memory cell row corresponding to the hammer address.

A semiconductor device includes a substrate, a first cell having a first functionality, and a second cell having a second functionality. The first cell includes a first portion on a first side of the substrate, wherein the first portion includes a first conductive element; a second portion on a second side of the substrate, wherein the second portion includes a second conductive element; and a first conductive via extending through the substrate and electrically connecting the first conductive element to the second conductive element. The second cell includes a third portion on the first side of the substrate, wherein the third portion includes a third conductive element; a fourth portion on the second side of the substrate, wherein the fourth portion includes a fourth conductive element; and a second conductive via extending through the substrate and electrically connecting the third conductive element to the fourth conductive element.

The invention provides an integrated circuit device and an adaptive power scaling method thereof to reduce and optimize power consumption. The integrated circuit device includes a first die and a second die, wherein the first die and the second die are stacked into a three-dimensional structure. A power circuit provides a power voltage to a first interface circuit of the first die and a second interface circuit of the second die. The first interface circuit transmits data to the second interface circuit via a die-to-die transfer circuit. The control logic controls the power circuit to adjust the power voltage provided to the first interface circuit and the second interface circuit based on the signal quality of the data received by the second interface circuit.

There is provided a memory device including a first chip including a first normal region, the first region including a plurality of first normal connectors on a first surface and configured to be provided with signals used during an operation of memory cells, and a first test region including a plurality of first connectors on the first surface and electrically connected to each other, and a second chip. The second chip includes a second normal region including a plurality of second normal connectors, and configured to provide signals used during the operation of the memory cells to the first normal connectors, and a second test region including a plurality of first and second test connectors on the second surface so as not to overlap the plurality of first connectors in the first direction, and configured to not be provided with signals used during the operation of the memory cells.

Methods, systems, and devices for pass-through power delivery for logic-on-top semiconductor systems are described. A semiconductor system may be configured with a two-dimensional pattern of power delivery conductors that pass through semiconductor components of a stack (e.g., through one or more memory stacks), providing a more-distributed delivery of power to a logic component bonded with the stack. The power delivery conductors may include through-substrate vias that bypass circuitry of the stack, and thus may be allocated for providing power to the logic component. Such techniques may be combined with a redistribution component, such as a package substrate or interposer (e.g., opposite the logic component in the heterogeneous stack), which may include redistribution conductors that convert from relatively fewer interconnections at a surface of the semiconductor system (e.g., for solder interconnection) to relatively more interconnections at a surface bonded with the stack (e.g., for hybrid bonding interconnection).

A semiconductor device includes first, second and third stacked chips with a first, second and third substrate, respectively, at least three first, second and third logical circuits, respectively, and at least two first, second and third vias, respectively, and a fourth chip stacked on the third chip having a fourth substrate, and at least three fourth logical circuits. First and second ones of the first to third logical circuits of the first to fourth chips are each configured to perform a first and second logical operation, respectively, on a first and second address input signal, respectively, received at the respective chip to thereby output a first and second address output signal, respectively. Third ones are each configured to activate the respective chip based on at least the second address output signal transmitted within the respective chip.

A memory device includes a memory die bonded to a logic die. A logic die that is bonded to a memory die via a wafer-on-wafer bonding process can receive signals indicative of input data from a global data bus of the memory die and through a bond of the logic die and memory die. The logic die can also receive signals indicative of kernel data from local input/output (LIO) lines of the memory die and through the bond. The logic die can perform a plurality of operations at a plurality of vector-vector (VV) units utilizing the signals indicative of input data and the signals indicative of kernel data.

A semiconductor package includes a first semiconductor chip including a first semiconductor layer, a first through-electrode that penetrates through the first semiconductor layer, a first bonding pad connected to the first through-electrode, and a first insulating bonding layer, and a second semiconductor chip on the first semiconductor chip and including a second semiconductor layer, a second bonding pad bonded to the first bonding pad, and a second insulating bonding layer bonded to the first insulating bonding layer, wherein the first insulating bonding layer includes a first insulating material, the second insulating bonding layer includes a first insulating layer that forms a bonding interface with the first insulating bonding layer and a second insulating layer on the first insulating layer, the first insulating layer includes a second insulating material, different from the first insulating material, and the second insulating layer includes a third insulating material, different from the second insulating material.

A method of manufacturing a semiconductor package may include: preparing a semiconductor wafer including rear pads and a rear insulating layer surrounding the rear pads, the rear insulating layer including first recesses spaced apart from the rear pads in a first lateral direction; preparing second semiconductor chips including front pads and a front insulating layer surrounding the front pads, the front insulating layer including second recesses spaced apart from the front pads in the first lateral direction; forming an air gap between the first recesses and the second recesses in a vertical direction by disposing the second semiconductor chips on the semiconductor wafer, the rear pads contacting the front pads; and bonding the rear insulating layer and the front insulating layer to each other and bonding the rear pads and the front pads to each other by performing a thermal compression process.

Methods, systems, and devices for redundant bond pads in stacked semiconductor architectures are described. A semiconductor device may be formed one or more redundant structures. A memory chip and a logic die may be formed with a redistribution layer that interconnects multiple bonding pads together. The redistribution layer may couple the bonding pads with a common via, where the common via interfaces with circuitry of a respective device. Additionally, or alternatively, a memory chip and a logic die may be formed with redundant via paths that form parallel electrical paths. The redundant via paths may couple device circuitry with respective bonding pads of a device. The memory chip and the logic die may be bonded together to form a semiconductor device.