Qualcomm Incorporated

Other

Ye Lu

Junjing Bao

Haitao Cheng

Chao Song

Bin Yang

Lixin Ge

Yun Yue

Zhongze Wang

Peijie Feng

Chenjie Tang

Xia Li

Yandong Gao

Chuan-Hsing Chen

Xiaochun Zhu

Haining S Yang

John Jianhong Zhu

Periannan Chidambaram

Xiaonan Chen

Giridhar Nallapati

Jun Chen

Dongjiang Qiao

Zhengzheng Wu

Kwanyong Lim

Phanikumar Konkapaka

Chuan-cheng Cheng

Ken Liao

A multi-gate HEMT includes at least two gates, with at least one recessed the same depth or at a deeper depth in a barrier layer than at least one other gate. Recessing a gate decreases the thickness of the barrier layer beneath the gate, reducing a density of high mobility carriers in a two-dimensional electron gas layer (2DEG) conductive channel formed at the heterojunction of a barrier layer and a buffer layer below the recessed gate. The recessed gate can increase gate control of the 2DEG conductive channel. The multi-gate HEMT has at least one gate recessed the same depth or a deeper depth into the buffer layer than another gate, which forms at least two different turn-on voltages for different gates. This can achieve improvement of transconductance linearity and a positive shift of the threshold voltage.

Multi-gate high electron mobility transistors (HEMTs) employing tuned recess depth gates for improved device linearity

An apparatus includes first and second compute-in-memory (CIM) arrays. The first CIM array is configured to store weights corresponding to a filter tensor, to receive a first set of activations corresponding to a first receptive field of an input, and to process the first set of activations with the weights to generate a corresponding first tensor of output values. The second CIM array is configured to store a first copy of the weights corresponding to the filter tensor and to receive a second set of activations corresponding to a second receptive field of the input. The second CIM array is also configured to process the second set of activations with the first copy of the weights to generate a corresponding second tensor of output values. The first and second compute-in-memory arrays are configured to process the first and second receptive fields in parallel.

Parallel processing of a convolutional layer of a neural network with compute-in-memory array

Certain aspects provide a circuit for in-memory computation. The circuit generally includes an in-memory computation array having a plurality of computation circuits, each of the computation circuits being configured to perform a dot product computation. In certain aspects, each of the computation circuits includes a memory cell, a capacitive element, a precharge transistor coupled between an output of the memory cell and the capacitive element, and a read transistor coupled between a read bit line (RBL) and the capacitive element.

Memory cell for dot product operation in compute-in-memory chip

A bit cell circuit of a most-significant bit (MSB) of a multi-bit product generated in an array of bit cells in a compute-in-memory (CIM) array circuit is configured to receive a higher supply voltage than a supply voltage provided to a bit cell circuit of another bit cell corresponding to another bit of the multi-bit product. A bit cell circuit receiving a higher supply voltage increases a voltage difference between increments of an accumulated voltage, which can increase accuracy of an analog-to-digital converter determining a pop-count. A bit cell circuit of the MSB in the CIM array circuit receives the higher supply voltage to increase accuracy of the MSB which increases accuracy of the CIM array circuit output. A capacitance of a capacitor in the bit cell circuit of the MSB is smaller to avoid an increase in energy consumption due to the higher voltage.

Multi-bit compute-in-memory (CIM) arrays employing bit cell circuits optimized for accuracy and power efficiency

An integrated device that includes a substrate and a first transistor formed over the substrate. The first transistor includes a first source disposed over the substrate, a first drain disposed over the substrate, a first plurality of channels coupled to the first source and the first drain, where the first plurality of channels is located between the first source and the first drain; at least one inner spacer located between two adjacent channels from the first plurality of channels; at least two voids located between the two adjacent channels; and a first gate surrounding the first plurality of channels.

Field effect transistor (FET) comprising inner spacers and voids between channels

Certain aspects of the present disclosure relate to a semiconductor device (e.g., a gate-all-around (GAA) semiconductor device) comprising at least one superlattice fin. One example superlattice fin includes a first plurality of nanosheets composed of a first semiconductor material and a second plurality of nanosheets composed of a second semiconductor material, the second semiconductor material being different from the first semiconductor material, wherein a width of a first nanosheet in the first plurality of nanosheets differs from a width of a second nanosheet in the second plurality of nanosheets, the second nanosheet being adjacent to the first nanosheet.

Semiconductor device with superlattice fin

A low cost capacitor (e.g., metal-insulator-metal (MIM) capacitor) is included in the back-end-of-line layers for effective routing and area savings. The capacitor has a first electrode (e.g., a first terminal of the capacitor) including a conductive back-end-of-line (BEOL) layer and a second electrode (e.g., a second terminal of the capacitor) including a nitride-based metal. The capacitor also has an etch stop layer (e.g., a dielectric of the capacitor) between the first electrode and the second electrode.

Back-end-of-line integrated metal-insulator-metal capacitor

3D vertically-integrated FETs electrically coupled by integrated vertical FET-to-FET interconnects for reducing an area of CMOS cell circuits are disclosed. Vertically integrated FETs reduce a footprint area of an integrated circuit chip. The FETs include horizontal channel structures that are vertically integrated by stacking a second channel structure of a second FET above a first channel structure of a first FET. The first and second FETs can include a combination of a PFET and NFET that can be used to form a 3D CMOS cell circuit as an example. The area occupied by the 3D CMOS cell circuit includes interconnects for electrically coupling terminal regions of the FETs internally and externally. Vertical FET-to-FET interconnects extend between the FETs to electrically couple terminal regions of the FETs to reduce a number of vias from a semiconductor layer of the 3D CMOS cell circuit to metal interconnect layers above the vertically-integrated FETs.

Three-dimensional (3D), vertically-integrated field-effect transistors (FETs) electrically coupled by integrated vertical FET-to-FET interconnects for complementary metal-oxide semiconductor (CMOS) cell circuits

Certain aspects of the present disclosure generally relate to a semiconductor device including an air gap underneath passive devices. The semiconductor device generally includes a substrate layer, a passive device layer, and a dielectric layer disposed between the substrate layer and the passive device layer, wherein the dielectric layer includes an air gap disposed beneath at least one passive device in the passive device layer.

Air gap underneath passive devices

Certain aspects of the present disclosure generally relate to a metal-oxide-metal (MOM) capacitor formed from a subtractive back-end-of-line (BEOL) scheme. One example method of fabricating a semiconductor device generally includes forming an active layer and forming a capacitive element above the active layer with a back-end-of-line subtractive process for conductive materials.

Metal-oxide-metal capacitor from subtractive back-end-of-line scheme

Certain aspects of the present disclosure generally relate to a gate-all-around (GAA) semiconductor device. The GAA semiconductor device generally includes a substrate, a first nanosheet stack structure, a second nanosheet stack structure, the first and second nanosheet stack structures being disposed above a horizontal plane of the substrate and each comprising one or more nanosheet structures, and a dielectric structure disposed between the first nanosheet stack structure and the second nanosheet stack structure.

Gate-all-around devices with reduced parasitic capacitance

A gate all around transistor may be improved to provide better transistor circuits performance. In one example, a transistor circuit may include a dielectric or air gap as an insulator between the channels of the transistors in the circuit. In another example, a transistor may include a first channel surrounded by a first metal, a second channel surrounded by a second metal proximate to the first channel, and an insulator, such as a dielectric or air gap, between the first metal and the second metal. The insulator helps reduce the parasitic capacitance between the source/drain regions and the metal fill regions of the transistor.

Transistor with insulator

Certain aspects of the present disclosure provide a metal-on-metal (MoM) capacitor with metal layers, each layer having two different electrical conductors with orthogonally-arranged conductive arteries and orthogonally-oriented conductive fingers. One exemplary MoM capacitor generally includes a plurality of metal layers, wherein a first metal layer in the plurality of metal layers comprises a first electrical conductor providing a first node of the MoM capacitor and a second electrical conductor providing a second node of the MoM capacitor. According to aspects, the first electrical conductor comprises a first plurality of conductive fingers and the second electrical conductor comprises a second plurality of conductive fingers. Further, conductive fingers of the first plurality of conductive fingers are interdigitated with conductive fingers of the second plurality of conductive fingers. Additionally, the first electrical conductor in the first metal layer is oriented orthogonal to the second electrical conductor in the first metal layer.

Metal-on-metal capacitor

An integrated circuit (IC) device includes a first resistive strip having an input terminal and an output terminal. The IC device further includes a second resistive strip having a terminal coupled to a voltage. The second resistive strip may be coplanar with the first resistive strip. The IC device further includes a capacitor formed by the first resistive strip and the second resistive strip.

Process-invariant delay cell

A charge sharing Compute In Memory (CIM) may comprise an XNOR bit cell with an internal capacitor between the XNOR output node and a system voltage. Alternatively, a charge sharing CIM may comprise an XNOR bit cell with an internal capacitor between the XNOR output node and a read bit line. Alternatively, a charge sharing CIM may comprise an XNOR bit cell with an internal cap between XNOR and read bit line with a separate write bit line and write bit line bar.

Growing community of inventors

San Diego, CA, United States of America

Ye Lu