A method according to the present disclosure includes forming a plurality of transistors in a first wafer and forming a memory array in a second wafer. A first surface of the first wafer includes a first plurality of bonding pads electrically coupled to the transistors. The memory array includes a plurality of ferroelectric tunnel junction (FTJ) stacks. A second surface of the second wafer includes a second plurality of bonding pads electrically coupled to the FTJ stacks. The method also includes performing a thermal treatment to the FTJ stacks in the second wafer, and after the performing of the thermal treatment, bonding the first surface of the first wafer with the second surface of the second wafer. The transistors are coupled to the memory cells through the first plurality of bonding pads and the second plurality of bonding pads.

A packaging technology to improve performance of an AI processing system resulting in an ultra-high bandwidth system. An IC package is provided which comprises: a substrate; a first die on the substrate, and a second die stacked over the first die. The first die can be a first logic die (e.g., a compute chip, CPU, GPU, etc.) while the second die can be a compute chiplet comprising ferroelectric or paraelectric logic. Both dies can include ferroelectric or paraelectric logic. The ferroelectric/paraelectric logic may include AND gates, OR gates, complex gates, majority, minority, and/or threshold gates, sequential logic, etc. The IC package can be in a 3D or 2.5D configuration that implements logic-on-logic stacking configuration. The 3D or 2.5D packaging configurations have chips or chiplets designed to have time distributed or spatially distributed processing. The logic of chips or chiplets is segregated so that one chip in a 3D or 2.5D stacking arrangement is hot at a time.

Systems, apparatuses, and methods related to dynamic random access memory (DRAM), such as finer grain DRAM, are described. For example, an array of memory cells in a memory device may be partitioned into regions. Each region may include a plurality of banks of memory cells. Each region may be associated with a data channel configured to communicate with a host device. In some examples, each channel of the array may include two or more data pins. The ratio of data pins per channel may be two or four in various examples. Other examples may include eight data pins per channel.

Described is a packaging technology to improve performance of an AI processing system. An IC package is provided which comprises: a substrate; a first die on the substrate, and a second die stacked over the first die. The first die includes memory and the second die includes computational logic. The first die comprises DRAM having bit-cells. The memory of the first die may store input data and weight factors. The computational logic of the second die is coupled to the memory of the first die. In one example, the second die is an inference die that applies fixed weights for a trained model to an input data to generate an output. In one example, the second die is a training die that enables learning of the weights. Ultra high-bandwidth is changed by placing the first die below the second die. The two dies are wafer-to-wafer bonded or coupled via micro-bumps.

Systems and devices for routing signals between a memory device and an interface of a host device are described. Some memory technologies may have a defined, preconfigured interface (e.g., bumpout), where each interface terminal may have a specific location and a specific function. Using preconfigured interfaces may allow device maker and memory makers to make parts that are able to connect with one another without special designs. In some cases, a memory device may include a redistribution layer that includes a plurality of interconnects that may be configured couple channel terminals of the memory device with an interface associated with the host device.

A semiconductor package including a core substrate, a semiconductor chip in the core substrate and having chip pads, a redistribution wiring layer covering a lower surface of the core substrate and including redistribution wirings electrically connected to the chip pads and a pair of capacitor pads exposed from an outer surface of the redistribution wiring layer, conductive pastes on the capacitor pads, respectively, and a capacitor via the conductive pastes and having first and second outer electrodes on the capacitor pads, respectively, may be provided. Each of the capacitor pads includes a pad pattern exposed from the outer surface of the redistribution wiring layer, and at least one via pattern at a lower portion of the pad pattern and electrically connected to at least one of the redistribution wirings. The via pattern is eccentric by a distance from a center line of the pad pattern.

In certain aspects, a memory device includes a vertical transistor, a storage unit, a bit line, and a body line. The vertical transistor includes a semiconductor body extending in a first direction. The semiconductor body includes a doped source, a doped drain, and a channel portion. The storage unit is coupled to a first terminal. The first terminal is one of the source and the drain. The bit line extends in a second direction perpendicular to the first direction and coupled to a second terminal. The second terminal is another one of the source and the drain. The body line is coupled to the channel portion of the semiconductor body.

Methods, systems, and devices for power distribution for stacked memory are described. A memory die may be configured with one or more conductive paths for providing power to another memory die, where each conductive path may pass through the memory die but may be electrically isolated from circuitry for operating the memory die. Each conductive path may provide an electronic coupling between at least one of a first set of contacts of the memory die (e.g., couplable with a power source) and at least one of a second set of contacts of the memory die (e.g., couplable with another memory die). To support operations of the memory die, a contact of the first set may be coupled with circuitry for operating a memory array of the memory die, and to support operations of another memory die, another contact of the first set may be electrically isolated from the circuitry.

A semiconductor package including a core substrate, a semiconductor chip in the core substrate and having chip pads, a redistribution wiring layer covering a lower surface of the core substrate and including redistribution wirings electrically connected to the chip pads and a pair of capacitor pads exposed from an outer surface of the redistribution wiring layer, conductive pastes on the capacitor pads, respectively, and a capacitor via the conductive pastes and having first and second outer electrodes on the capacitor pads, respectively, may be provided. Each of the capacitor pads includes a pad pattern exposed from the outer surface of the redistribution wiring layer, and at least one via pattern at a lower portion of the pad pattern and electrically connected to at least one of the redistribution wirings. The via pattern is eccentric by a distance from a center line of the pad pattern.

Methods, systems, and devices for power distribution for stacked memory are described. A memory die may be configured with one or more conductive paths for providing power to another memory die, where each conductive path may pass through the memory die but may be electrically isolated from circuitry for operating the memory die. Each conductive path may provide an electronic coupling between at least one of a first set of contacts of the memory die (e.g., couplable with a power source) and at least one of a second set of contacts of the memory die (e.g., couplable with another memory die). To support operations of the memory die, a contact of the first set may be coupled with circuitry for operating a memory array of the memory die, and to support operations of another memory die, another contact of the first set may be electrically isolated from the circuitry.