Abstract: Embodiments of the invention relate to providing transposable access to a synapse array using a recursive array layout. One embodiment comprises maintaining synaptic weights for multiple synapses connecting multiple axons and multiple neurons, wherein the synaptic weights are maintained based on a recursive array layout. The recursive array layout facilitates transposable access to the synaptic weights. A neuronal spike event between an axon and a neuron is communicated via a corresponding connecting synapse by accessing the synaptic weight of the corresponding connecting synapse in the recursive array layout.

Abstract: An array organization and architecture for a content addressable memory (CAM) system. More specifically, a circuit is provided for that includes a first portion of the CAM configured to perform a first inequality operation implemented between 1 to n CAM entries. The circuit further includes a second portion of the CAM configured to perform a second inequality operation implemented between the 1 to n CAM entries. The first portion and the second portion are triangularly arranged side by side such that the first inequality operation and the second inequality operation are implemented between the 1 to n CAM entries using the same n wordlines.

Abstract: Embodiments of the invention relate to processor arrays, and in particular, a processor array with interconnect circuits for bonding semiconductor dies. One embodiment comprises multiple semiconductor dies and at least one interconnect circuit for exchanging signals between the dies. Each die comprises at least one processor core circuit. Each interconnect circuit corresponds to a die of the processor array. Each interconnect circuit comprises one or more attachment pads for interconnecting a corresponding die with another die, and at least one multiplexor structure configured for exchanging bus signals in a reversed order.

Abstract: An array organization and architecture for a content addressable memory (CAM) system. More specifically, a circuit is provided for that includes a first portion of the CAM configured to perform a first inequality operation implemented between 1 to n CAM entries. The circuit further includes a second portion of the CAM configured to perform a second inequality operation implemented between the 1 to n CAM entries. The first portion and the second portion are triangularly arranged side by side such that the first inequality operation and the second inequality operation are implemented between the 1 to n CAM entries using the same n wordlines.

Abstract: An array organization and architecture for a content addressable memory (CAM) system. More specifically, a circuit is provided for that includes a first portion of the CAM configured to perform a first inequality operation implemented between 1 to n CAM entries. The circuit further includes a second portion of the CAM configured to perform a second inequality operation implemented between the 1 to n CAM entries. The first portion and the second portion are triangularly arranged side by side such that the first inequality operation and the second inequality operation are implemented between the 1 to n CAM entries using the same n wordlines.

Abstract: Embodiments of the invention relate to faulty recovery mechanisms for a three-dimensional (3-D) network on a processor array. One embodiment comprises a multidimensional switch network for a processor array. The switch network comprises multiple switches for routing packets between multiple core circuits of the processor array. The switches are organized into multiple planes. The switch network further comprises a redundant plane including multiple redundant switches. Multiple data paths interconnect the switches. The redundant plane is used to facilitate full operation of the processor array in the event of one or more component failures.

Abstract: Embodiments of the invention relate to providing transposable access to a synapse array using a recursive array layout. One embodiment comprises maintaining synaptic weights for multiple synapses connecting multiple axons and multiple neurons, wherein the synaptic weights are maintained based on a recursive array layout. The recursive array layout facilitates transposable access to the synaptic weights. A neuronal spike event between an axon and a neuron is communicated via a corresponding connecting synapse by accessing the synaptic weight of the corresponding connecting synapse in the recursive array layout.

Abstract: Embodiments of the invention relate to processor arrays, and in particular, a processor array with interconnect circuits for bonding semiconductor dies. One embodiment comprises multiple semiconductor dies and at least one interconnect circuit for exchanging signals between the dies. Each die comprises at least one processor core circuit. Each interconnect circuit corresponds to a die of the processor array. Each interconnect circuit comprises one or more attachment pads for interconnecting a corresponding die with another die, and at least one multiplexor structure configured for exchanging bus signals in a reversed order.

Abstract: Embodiments of the invention relate to providing transposable access to a synapse array using a recursive array layout. One embodiment comprises maintaining synaptic weights for multiple synapses connecting multiple axons and multiple neurons, wherein the synaptic weights are maintained based on a recursive array layout. The recursive array layout facilitates transposable access to the synaptic weights. A neuronal spike event between an axon and a neuron is communicated via a corresponding connecting synapse by accessing the synaptic weight of the corresponding connecting synapse in the recursive array layout.

Abstract: Embodiments of the invention relate to processor arrays, and in particular, a processor array with interconnect circuits for bonding semiconductor dies. One embodiment comprises multiple semiconductor dies and at least one interconnect circuit for exchanging signals between the dies. Each die comprises at least one processor core circuit. Each interconnect circuit corresponds to a die of the processor array. Each interconnect circuit comprises one or more attachment pads for interconnecting a corresponding die with another die, and at least one multiplexor structure configured for exchanging bus signals in a reversed order.

Abstract: Embodiments of the invention relate to an array of processor core circuits with reversible tiers. One embodiment comprises multiple tiers of core circuits and multiple switches for routing packets between the core circuits. Each tier comprises at least one core circuit. Each switch comprises multiple router channels for routing packets in different directions relative to the switch, and at least one routing circuit configured for reversing a logical direction of at least one router channel.

Abstract: Trench capacitors can be formed between lengthwise sidewalls of semiconductor fins, and source and drain regions of access transistors are formed in the semiconductor fins. A dummy gate structure is formed between end walls of a neighboring pair of semiconductor fins, and limits the lateral extent of raised source and drain regions that are formed by selective epitaxy. The dummy gate structure prevents electrical shorts between neighboring semiconductor fins. Gate spacers can be formed around gate structures and the dummy gate structures. The dummy gate structures can be replaced with dummy replacement gate structures or dielectric material portions, or can remain the same without substitution of any material. The dummy gate structures may consist of at least one dielectric material, or may include electrically floating conductive material portions.

Abstract: After formation of trench capacitors and source and drain regions and gate structures for access transistors, a dielectric spacer is formed on a first sidewall of each source region, while a second sidewall of each source region and sidewalls of drain regions are physically exposed. Each dielectric spacer can be employed as an etch mask during removal of trench top dielectric portions to form strap cavities for forming strap structures. Optionally, selective deposition of a semiconductor material can be performed to form raised source and drain regions. In this case, the raised source regions grow only from the first sidewalls and do not grow from the second sidewalls. The raised source regions can be employed as a part of an etch mask during formation of the strap cavities. The strap structures are formed as self-aligned structures that are electrically isolated from adjacent access transistors by the dielectric spacers.

Abstract: Embodiments of the invention relate to an array of processor core circuits with reversible tiers. One embodiment comprises multiple tiers of core circuits and multiple switches for routing packets between the core circuits. Each tier comprises at least one core circuit. Each switch comprises multiple router channels for routing packets in different directions relative to the switch, and at least one routing circuit configured for reversing a logical direction of at least one router channel.

Abstract: Embodiments of the invention relate to an array of processor core circuits with reversible tiers. One embodiment comprises multiple tiers of core circuits and multiple switches for routing packets between the core circuits. Each tier comprises at least one core circuit. Each switch comprises multiple router channels for routing packets in different directions relative to the switch, and at least one routing circuit configured for reversing a logical direction of at least one router channel.

Abstract: A method of forming a vertically integrated memory cell including a deep trench extending into a substrate, a trench capacitor located within the deep trench, and a vertical transistor at least partially embedded within the deep trench above the trench capacitor, the vertical transistor is in direct contact with and electrically coupled to the trench capacitor.

Abstract: After formation of trench capacitors and source and drain regions and gate structures for access transistors, a dielectric spacer is formed on a first sidewall of each source region, while a second sidewall of each source region and sidewalls of drain regions are physically exposed. Each dielectric spacer can be employed as an etch mask during removal of trench top dielectric portions to form strap cavities for forming strap structures. Optionally, selective deposition of a semiconductor material can be performed to form raised source and drain regions. In this case, the raised source regions grow only from the first sidewalls and do not grow from the second sidewalls. The raised source regions can be employed as a part of an etch mask during formation of the strap cavities. The strap structures are formed as self-aligned structures that are electrically isolated from adjacent access transistors by the dielectric spacers.

Abstract: Trench capacitors can be formed between lengthwise sidewalls of semiconductor fins, and source and drain regions of access transistors are formed in the semiconductor fins. A dummy gate structure is formed between end walls of a neighboring pair of semiconductor fins, and limits the lateral extent of raised source and drain regions that are formed by selective epitaxy. The dummy gate structure prevents electrical shorts between neighboring semiconductor fins. Gate spacers can be formed around gate structures and the dummy gate structures. The dummy gate structures can be replaced with dummy replacement gate structures or dielectric material portions, or can remain the same without substitution of any material. The dummy gate structures may consist of at least one dielectric material, or may include electrically floating conductive material portions.

Abstract: A method of forming a vertically integrated memory cell including a deep trench extending into a substrate, a trench capacitor located within the deep trench, and a vertical transistor at least partially embedded within the deep trench above the trench capacitor, the vertical transistor is in direct contact with and electrically coupled to the trench capacitor.

Abstract: A vertically integrated memory cell including a deep trench extending into a substrate, a trench capacitor located within the deep trench, and a vertical transistor at least partially embedded within the deep trench above the trench capacitor, the vertical transistor is in direct contact with and electrically coupled to the trench capacitor.