Direct-bonded native interconnects and active base dies are provided. In a microelectronic architecture, active dies or chiplets connect to an active base die via their core-level conductors. These native interconnects provide short data paths, which forgo the overhead of standard interfaces. The system saves redistribution routing as the native interconnects couple in place. The base die may contain custom logic, allowing the attached dies to provide stock functions. The architecture can connect diverse interconnect types and chiplets from various process nodes, operating at different voltages. The base die may have state elements for drive. Functional blocks aboard the base die receive native signals from diverse chiplets, and communicate with all attached chiplets. The chiplets may share processing and memory resources of the base die. Routing blockages are minimal, improving signal quality and timing. The system can operate at dual or quad data rates. The architecture facilitates ASIC, ASSP, and FPGA ICs and neural networks, reducing footprint and power requirements.

A method of manufacturing a semiconductor device comprises forming a preliminary substrate insulating layer on a lower substrate region; forming buried interconnection lines on the preliminary substrate insulating layer; forming a substrate insulating layer on the buried interconnection lines; forming an upper substrate region on the lower substrate region to form a substrate; forming active regions and a device isolation layer by removing a portion of the substrate; forming sacrificial gate structures and source/drain regions; removing the sacrificial gate structures; forming gate structures; forming a first interlayer insulating layer on the gate structures; forming first contact holes to expose the buried interconnection lines; forming preliminary lower contact plugs by filling the first contact holes; forming lower contact plugs; forming second contact holes to expose a portion of the source/drain regions or a portion of the gate structures; and forming upper contact plugs by filling the second contact holes.

An IC structure includes a first transistor, a second transistor, a dielectric fin, a dielectric cap, a backside metal structure, and a source/drain contact. The first transistor includes a first channel region, a first gate structure, and first source/drain features disposed on opposite sides of the first gate structure. The second transistor includes a second channel region, a second gate structure, and second source/drain features disposed on opposite sides of the second gate structure. The dielectric fin is disposed between the first and second transistors. The dielectric cap interfaces a backside surface of the dielectric fin. The source/drain contact abuts the dielectric fin and is electrically coupled to a first one of the first source/drain features by way of a silicide layer and electrically coupled to the backside metal rail by way of physical contact established by the source/drain contact and the backside metal rail.

A 3D semiconductor device including: a first level including a first single crystal layer and first transistors, which each include a single crystal channel; a first metal layer with an overlaying second metal layer; a second level including second transistors, overlaying the first level; a third level including third transistors, overlaying the second level; a fourth level including fourth transistors, overlaying the third level, where the second level includes first memory cells, where each of the first memory cells includes at least one of the second transistors, where the fourth level includes second memory cells, where each of the second memory cells includes at least one of the fourth transistors, where the first level includes memory control circuits, where second memory cells include at least four memory arrays, each of the four memory arrays are independently controlled, and at least one of the second transistors includes a metal gate.

Methods to fabricate a semiconductor device, the method including: forming a first level, the first level including a single crystal silicon layer, a plurality of transistors, and a plurality of first metal layers, where each transistor of the plurality of transistors includes a single crystal channel, and where the plurality of first metal layers include interconnections between the transistors of the plurality of transistors; thinning the single crystal silicon layer to a thickness of less than two microns; forming a second level, the second level including a plurality of second metal layers, where the second level is disposed underneath the first level; and forming a connection path between at least one of the transistors to at least one of the plurality of second metal layers, where the connective path includes at least one via disposed through at least the single crystal silicon layer.

An integrated circuit (IC) chip is provided. In one aspect, a semiconductor substrate includes active devices at its front surface and power delivery tracks on its back surface. The active devices are powered through mutually parallel buried power rails, with the power delivery tracks running transversely with respect to the power rails, and connected to the power rails by a plurality of Through Semiconductor Via connections, which run from the power rails to the back of the substrate. The TSVs are elongate slit-shaped TSVs aligned to the power rails and arranged in a staggered pattern, so that any one of the power delivery tracks is connected to a first row of mutually parallel TSVs, and any power delivery track directly adjacent to the power delivery track is connected to another row of TSVs which are staggered relative to the TSVs of the first row. A method of producing an IC chip includes producing the slit-shaped TSVs before the buried power rails.

A method including forming an oxide layer on a first substrate and forming a second substrate on the oxide layer. Doping a first section of the second substrate while not doping a second section of the second substrate. Forming a first nano device on the second section of the second substrate and forming a second nano device on first section of the second substrate. Flipping the first substrate over to allow for backside processing of the substrate and forming at least one backside contact connected to the first nano device while backside contacts are not formed or connected to the second nano device.

A method of forming a semiconductor device is provided. The method includes forming an etch stop layer on a substrate having a first thickness, forming an epitaxial layer on the etch stop layer, and forming a wafer device on the epitaxial layer. The wafer device is bonded to a bonding wafer using hybrid bonding. The substrate is then ground to a second thickness less than the first thickness and planarized to a third thickness less than the second thickness. A mask layer is deposited on a bottom surface of the etch stop layer, and at least one via opening is formed in the mask layer. The etch stop layer is selectively removed, and the mask layer is removed to expose the substrate at the third thickness.

A semiconductor structure includes a front-end-of-line level including a plurality of field effect transistors electrically connected to a back-end-of-line interconnect level located on a first side of the front-end-of-line level. A plurality of shallow trench isolation regions are located between adjacent field effect transistors, each of the plurality of shallow trench isolation regions being surrounded by a dielectric isolation liner. A backside power rail is located within a backside interlayer dielectric located on a second side of the front-end-of-line level opposing the first side of the front-end-of-line level. A via-to-backside power rail embedded, at least in part, within a shallow trench isolation region is located between two field effect transistors of a similar polarity, the via-to-backside power rail is adjacent and electrically connected to at least one metal contact and extends from the at least one metal contact to a first surface of the backside power rail.

Power distribution fabrics and methods of forming the same include forming a first layer of parallel conductive lines, having a first width. At least one additional layer of conductive lines is formed over the first layer of conductive lines, with the conductive lines of each successive layer in the at least one additional layer having a different orientation and a different width relative to the conductive lines of the preceding layer.