Abstract: A quasi-volatile memory (QV memory) stack includes at least one semiconductor die, having formed thereon QV memory circuits, bonded to a second semiconductor on which a memory controller for the QV memory (“QV memory controller”) is formed. The circuits in the bonded semiconductor dies are electrically connected using numerous copper interconnect conductors and conductive through-silicon vias (TSVs). The QV memory controller may include one or more interfaces to additional devices (“back-channel devices”) to enable the QV memory controller to also serve as a controller for each back-channel device and to provide additional services. The QV memory controller performs data transfers between a back-channel device and the QV memory without intervention by the host CPU.

Abstract: A memory structure including three-dimensional NOR memory strings and method of fabrication is disclosed. In one embodiment, a memory system includes a set of memory modules of quasi-volatile memory circuits interconnected to a memory controller and having a set of memory ports. The memory system includes a first processor port, a second processor port, and one or more DIMM interface ports to be coupled to respective processors for providing access to the set of memory modules. In another embodiment, an artificial intelligence (AI) computing system includes a set of memory modules of quasi-volatile memory circuits interconnected to a memory controller and an arithmetic function block performing multiply and accumulate functionalities using data stored in the memory modules. The set of memory modules are accessed to perform read, write and erase memory operations in a rotating manner in each computing cycle.

Abstract: Semiconductor memory systems and architectures for shared memory access implements memory-centric structures using a quasi-volatile memory. In one embodiment, a memory processor array includes an array of memory cubes, each memory cube in communication with a processor mini core to form a computational memory. In another embodiment, a memory system includes processing units and one or more mini core-memory module both in communication with a memory management unit. Mini processor cores in each mini core-memory module execute tasks designated to the mini core-memory module by a given processing unit using data stored in the associated quasi-volatile memory circuits of the mini core-memory module.

Abstract: A first circuit formed on a first semiconductor substrate is wafer-bonded to a second circuit formed on a second memory circuit, wherein the first circuit includes quasi-volatile or non-volatile memory circuits and wherein the second memory circuit includes fast memory circuits that have lower read latencies than the quasi-volatile or non-volatile memory circuits, as well as logic circuits. The volatile and non-volatile memory circuits may include static random-access memory (SRAM) circuits, dynamic random-access memory (DRAM) circuits, embedded DRAM (eDRAM) circuits, magnetic random-access memory (MRAM) circuits, embedded MRAM (eMRAM), or any suitable combination of these circuits.

Abstract: A first circuit formed on a first semiconductor substrate is wafer-bonded to a second circuit formed on a second memory circuit, wherein the first circuit includes quasi-volatile or non-volatile memory circuits and wherein the second memory circuit includes fast memory circuits that have lower read latencies than the quasi-volatile or non-volatile memory circuits, as well as logic circuits. The volatile and non-volatile memory circuits may include static random-access memory (SRAM) circuits, dynamic random-access memory (DRAM) circuits, embedded DRAM (eDRAM) circuits, magnetic random-access memory (MRAM) circuits, embedded MRAM (eMRAM), or any suitable combination of these circuits.

Abstract: A high-capacity system memory may be built from both quasi-volatile (QV) memory circuits, logic circuits, and static random-access memory (SRAM) circuits. Using the SRAM circuits as buffers or cache for the QV memory circuits, the system memory may achieve access latency performance of the SRAM circuits and may be used as code memory. The system memory is also capable of direct memory access (DMA) operations and includes an arithmetic logic unit for performing computational memory tasks. The system memory may include one or more embedded processor. In addition, the system memory may be configured for multi-channel memory accesses by multiple host processors over multiple host ports. The system memory may be provided in the dual-in-line memory module (DIMM) format.

Abstract: A high-capacity system memory may be built from both quasi-volatile (QV) memory circuits, logic circuits, and static random-access memory (SRAM) circuits. Using the SRAM circuits as buffers or cache for the QV memory circuits, the system memory may achieve access latency performance of the SRAM circuits and may be used as code memory. The system memory is also capable of direct memory access (DMA) operations and includes an arithmetic logic unit for performing computational memory tasks. The system memory may include one or more embedded processor. In addition, the system memory may be configured for multi-channel memory accesses by multiple host processors over multiple host ports. The system memory may be provided in the dual-in-line memory module (DIMM) format.

Abstract: Semiconductor memory systems and architectures for shared memory access implements memory-centric structures using a quasi-volatile memory. In one embodiment, a memory processor array includes an array of memory cubes, each memory cube in communication with a processor mini core to form a computational memory. In another embodiment, a memory system includes processing units and one or more mini core-memory module both in communication with a memory management unit. Mini processor cores in each mini core-memory module execute tasks designated to the mini core-memory module by a given processing unit using data stored in the associated quasi-volatile memory circuits of the mini core-memory module.

Abstract: A shared memory provides multi-channel access from multiple computing or host devices. A priority circuit prioritizes the multiple memory requests that are submitted as bids from the multiple host channels, such that those memory access requests that do not give rise to a conflict may proceed in parallel. The shared memory may be multi-ported and a routing circuit routes the prioritized memory access request to the appropriate memory ports where the allowed memory access requests may be carried out.

Abstract: A memory device includes a first semiconductor wafer portion including two or more adjacent quasi-volatile memory circuits formed on a common semiconductor substrate where each quasi-volatile memory circuit being isolated from an adjacent quasi-volatile memory circuit by scribe lines; and a second semiconductor wafer portion including at least one memory controller circuit formed on a semiconductor substrate. The memory controller circuit includes logic circuits and interface circuits. The memory controller circuit is interconnected to the two or more adjacent quasi-volatile memory circuits of the first semiconductor wafer portion through interconnect structures and the memory controller circuit operates the two or more quasi-volatile memory circuits as one or more quasi-volatile memories.

Abstract: A semiconductor memory module for shared memory access implements memory-centric structures using a quasi-volatile memory. In one embodiment, the memory module for shared memory access includes a memory cube providing high capacity memory coupled to multiple multi-port memories to support simultaneous memory access at multiple memory interfaces. In other embodiments, a memory module incorporates a processor to implement computational memory architecture. In some embodiments, a mini core memory system implements a memory architecture for providing direct and parallel memory access to a mini processor core array.

Abstract: NOR memory strings may be used for implementations of logic functions involving many Boolean variables, or to generate analog signals whose magnitudes are each representative of the bit values of many Boolean variables. The advantage of using NOR memory strings in these manners is that the logic function or analog signal generation may be accomplished within one simultaneous read operation on the NOR memory strings.

Abstract: A quasi-volatile memory (QV memory) stack includes at least one semiconductor die, having formed thereon QV memory circuits, bonded to a second semiconductor on which a memory controller for the QV memory (“QV memory controller”) is formed. The circuits in the bonded semiconductor dies are electrically connected using numerous copper interconnect conductors and conductive through-silicon vias (TSVs). The QV memory controller may include one or more interfaces to additional devices (“back-channel devices”) to enable the QV memory controller to also serve as a controller for each back-channel device and to provide additional services. The QV memory controller performs data transfers between a back-channel device and the QV memory without intervention by the host CPU.

Abstract: A quasi-volatile memory (QV memory) stack includes at least one semiconductor die, having formed thereon QV memory circuits, bonded to a second semiconductor on which a memory controller for the QV memory (“QV memory controller”) is formed. The circuits in the bonded semiconductor dies are electrically connected using numerous copper interconnect conductors and conductive through-silicon vias (TSVs). The QV memory controller may include one or more interfaces to additional devices (“back-channel devices”) to enable the QV memory controller to also serve as a controller for each back-channel device and to provide additional services. The QV memory controller performs data transfers between a back-channel device and the QV memory without intervention by the host CPU.

Abstract: A memory device includes a first semiconductor wafer portion including two or more adjacent quasi-volatile memory circuits formed on a common semiconductor substrate where each quasi-volatile memory circuit being isolated from an adjacent quasi-volatile memory circuit by scribe lines; and a second semiconductor wafer portion including at least one memory controller circuit formed on a semiconductor substrate. The memory controller circuit includes logic circuits and interface circuits. The memory controller circuit is interconnected to the two or more adjacent quasi-volatile memory circuits of the first semiconductor wafer portion through interconnect structures and the memory controller circuit operates the two or more quasi-volatile memory circuits as one or more quasi-volatile memories.

Abstract: A shared memory provides multi-channel access from multiple computing or host devices. A priority circuit prioritizes the multiple memory requests that are submitted as bids from the multiple host channels, such that those memory access requests that do not give rise to a conflict may proceed in parallel. The shared memory may be multi-ported and a routing circuit routes the prioritized memory access request to the appropriate memory ports where the allowed memory access requests may be carried out.

Abstract: A semiconductor memory module for shared memory access implements memory-centric structures using a quasi-volatile memory. In one embodiment, the memory module for shared memory access includes a memory cube providing high capacity memory coupled to multiple multi-port memories to support simultaneous memory access at multiple memory interfaces. In other embodiments, a memory module incorporates a processor to implement computational memory architecture. In some embodiments, a mini core memory system implements a memory architecture for providing direct and parallel memory access to a mini processor core array.

Abstract: A first circuit formed on a first semiconductor substrate is wafer-bonded to a second circuit formed on a second memory circuit, wherein the first circuit includes quasi-volatile or non-volatile memory circuits and wherein the second memory circuit includes fast memory circuits that have lower read latencies than the quasi-volatile or non-volatile memory circuits, as well as logic circuits. The volatile and non-volatile memory circuits may include static random-access memory (SRAM) circuits, dynamic random-access memory (DRAM) circuits, embedded DRAM (eDRAM) circuits, magnetic random-access memory (MRAM) circuits, embedded MRAM (eMRAM), or any suitable combination of these circuits.

Abstract: A high-capacity system memory may be built from both quasi-volatile (QV) memory circuits, logic circuits, and static random-access memory (SRAM) circuits. Using the SRAM circuits as buffers or cache for the QV memory circuits, the system memory may achieve access latency performance of the SRAM circuits and may be used as code memory. The system memory is also capable of direct memory access (DMA) operations and includes an arithmetic logic unit for performing computational memory tasks. The system memory may include one or more embedded processor. In addition, the system memory may be configured for multi-channel memory accesses by multiple host processors over multiple host ports. The system memory may be provided in the dual-in-line memory module (DIMM) format.

Abstract: A method in a memory circuit for rapidly determining the location of a file includes (a) associating the file with a timestamp and a unique identifier index number when the file is stored or updated in the memory circuit, and storing in a look-up table in the memory circuit the associated timestamp and an address associated with where the file is stored; (b) receiving a search request that specifies a unique identifier index number of a file to be located; and (c) using exclusive-or (XOR) circuits or content addressable memory (CAM) circuits to compare the unique identifier index number in the search request with the unique identifier index number stored in the look-up table, and reporting, when a match is found between the unique identifier index number in the search request and the unique identifier index number stored in the look-up table, the timestamp and address associated with the match.