Micron Technology Incorporated

David Hulton

Jeremy Chritz

Tamara Schmitz

Jonathan D Harms

Gregory Edvenson

John L Watson

John Schroeter

Fa-Long Luo

Danilo Caraccio

Kevin T Majerus

Aaron Preston Boehm

Alessandro Orlando

Niccolò Izzo

Angelo Alberto Rovelli

Craig A Jones

Max Vohra

A method can include performing at least one glitch resistance operation and detecting, by a circuit included in a glitch detector coupled via a connection matrix to a first processing unit, an indication of a glitch on a memory system. The method can include notifying, via the connection matrix, at least a second processing unit of the detected indication of the glitch. The method can include subsequent to notifying at least the second processing unit, transmitting via the at least the second processing unit a glitch confirmation signal.

Glitch detection

Apparatuses and methods for managing a coherent memory are described. These may include one or more algorithmic logic units (ALUs) and an input/output (IO) interface. The I/O interface may receive one or more commands and retrieve data from or write data to a memory device. Each command may contain a memory address portion associated with a memory device. The apparatus may also include a memory mapping unit and a device controller. The memory mapping unit may map the memory address to a memory portion of the memory device, and the device controller may communicate with the memory device to retrieve data from or write data to the memory device. The apparatus may be implemented as a processing element in a configurable logic block network, which may additionally include a control logic unit that receives programming instructions from an application and generate the one or more commands based on the instructions.

Configurable logic block networks and managing coherent memory in the same

Methods, systems, and devices for verification of a volatile memory, such as a dynamic random-access memory (DRAM), using a unique identifier (ID) are described. A memory device may store a unique ID for a DRAM component of the memory device in non-volatile memory (e.g., in the DRAM, external to the DRAM). A host device coupled with the memory device may store, to non-volatile memory at the host device, information for verifying the identity of the DRAM component, for example, based on the unique ID. The memory device and host device may perform a procedure for verification of the identity of the DRAM component using the unique ID of the DRAM and the verification information stored at the host device. If the host device detects that the DRAM has been replaced or modified based on the verification procedure, the host device may disable one or more features of the memory device.

Verification of a volatile memory using a unique identifier

An apparatus, and a method therefore, are described, the apparatus according to one embodiment including a security manager and a plurality of clusters of processing elements. Each cluster of the plurality of clusters includes a respective plurality of processing elements. A controller of the apparatus, which may include a security manager, may be configured to control the plurality of clusters to receive a first data stream and a second data stream, control a first plurality of processing elements in a first cluster to process the first data stream using a first security protocol, and control a second plurality of processing elements in a second cluster to process the second data stream using a second security protocol.

Security managers and methods for implementing security protocols in a reconfigurable fabric

Examples of systems and methods described herein provide for accessing memory devices and, concurrently, generating access codes using an authenticated stream cipher at a memory controller. For example, a memory controller may use a memory access request to, concurrently, perform translation logic and/or error correction on data associated with the memory access request; while also utilizing the memory address as an initialization vector for an authenticated stream cipher to generate an access code. The error correction may be performed subsequent to address translation for a write operation (or prior to address translation for a read operation) to improve processing speed of memory access requests at a memory controller; while the memory controller also generates the encrypted access code.

Memory systems and devices including examples of accessing memory and generating access codes using an authenticated stream cipher

Methods, systems, and devices for techniques for managing offline identity upgrades are described. A memory system may receive a command to update a device identifier for a device identifier composition engine (DICE) associated with the memory system. The memory system may generate an updated device identifier, at a first software layer of a set of software layers of the DICE, based on receiving the command. The memory system may decrypt a device specific key (DSK) stored at a read-only memory device of the memory system based on the received command, and sign the updated device identifier using the DSK based on decrypting the DSK. The memory system may execute one or more operations associated with the first software layer of the set of software layers of the DICE based on the signed updated device identifier.

Techniques for managing offline identity upgrades

Methods, apparatuses, and systems for repairing defective memory cells in regions of a memory array associated with high or low priority levels are disclosed. A repair address generator may be configured to generate a memory address map for repair (e.g., blowing fuses at a fuse circuit), depending on whether certain applications may operate at a high priority level indicative of a low bit error rate or a low priority level indicative of a higher bit error rate. For example, a specified error rate associated with a low priority level may correspond to a threshold error rate for certain applications, such as a neural network application that stores trained weights. Such neural network applications may access trained weights being partially stored in defective memory cells, with the least significant bits of such trained weights being stored in defective memory cells that are not repaired according to the memory address map.

Apparatuses and methods for repairing defective memory cells based on a specified error rate for certain memory cells

Examples of systems and method described herein provide for accessing memory devices and, concurrently, generating access codes using an authenticated stream cipher at a memory controller. For example, a memory controller may use a memory access request to, concurrently, perform translation logic and/or error correction on data associated with the memory access request; while also utilizing the memory address as an initialization vector for an authenticated stream cipher to generate an access code. The error correction may be performed subsequent to address translation for a write operation (or prior to address translation for a read operation) to improve processing speed of memory access requests at a memory controller; while the memory controller also generates the encrypted access code.

Examples of systems and method described herein or generating, in a memory controller and/or memory device, access codes for memory regions of the memory device using authentication logic, and for accessing the memory device using the access codes. For example, a memory controller and/or a coupled memory device may generate access codes that a host computing device may include in a memory access request to access one or more memory regions of the memory device. Data read or written at the memory device may in some examples only be accessed in accordance with the access codes for memory regions of the memory device. Accordingly, the systems and methods described herein may provide security for specific memory regions of a memory device because the access code are updated periodically (e.g., based on obtained reset indication) or in accordance with an updated count value from a counter.

Memory systems and devices including examples of generating access codes for memory regions using authentication logic

Methods, apparatuses, and systems for implementing data flows in a processor are described herein. A data flow manager may be configured to generate a configuration packet for a compute operation based on status information regarding multiple processing elements of the processor. Accordingly, multiple processing elements of a processor may concurrently process data flows based on the configuration packet. For example, the multiple processing elements may implement a mapping of processing elements to memory, while also implementing identified paths, through the processor, for the data flows. After executing the compute operation at certain processing elements of the processor, the processing results may be provided. In speech signal processing operations, the processing results may be compared to phonemes to identify such components of human speech in the processing results. Once dynamically identified, the processing elements may continue comparing additional components of human speech to facilitate processing of an audio recording, for example.

Methods, systems, and apparatuses to perform a compute operation according to a configuration packet and comparing the result to data in local memory

A device comprising a plurality of antennas operable to transmit and receive communication packets via a plurality of communication protocols and an integrated circuit chip coupled to the plurality of antennas. The integrated circuit chip comprises a first and a second plurality of processing elements. The first plurality of processing elements operable to receive communication packets via a first one of a plurality of communication protocols and process an optimal route. The second plurality of processing elements communicatively coupled to the first plurality of processing elements and operable to determine the optimal route to transmit the communication packets from a source device to a destination device based, at least in part, on transmission characteristics associated with at least one of the source or destination devices.

System for optimizing routing of communication between devices and resource reallocation in a network

Apparatuses and methods are disclosed for an FPGA architecture that may improve processing speed and efficiency in processing less complex operands. Some applications may utilize operands that are less complex, such as operands that are 1, 2, or 4 bits, for example. In some examples, the DSP architecture may skip or avoid processing all received operands or may process a common operand more frequently than other operands. An example apparatus may include a first configurable logic unit configured to receive a first operand and a second operand; a second configurable logic unit configured to receive a third operand and the first calculated operand; a first switch configured to receive the first operand and a fourth operand and to output a first selected operand; and a second switch configured to receive the second calculated operand and the first selected operand.

DSP execution slice array to provide operands to multiple logic units

Examples of systems and method described herein provide for the processing of image codes (e.g., a binary embedding) at a memory die. Such images codes may generated by various endpoint computing devices, such as Internet of Things (IoT) computing devices, Such devices can generate a Hamming processing command, having an image code of the image, to compare that representation of the image to other images (e.g., in an image dataset) to identify a match or a set of neural network results. Advantageously, examples described herein may be used in neural networks to facilitate the processing of datasets, so as to increase the rate and amount of processing of such datasets. For example, comparisons of image codes can be performed on a memory die itself, like a memory die of a NAND memory device.

Memory including examples of calculating hamming distances for neural network and data center applications

Examples of systems and method described herein provide for the processing of image codes (e.g., a binary embedding) at a memory controller with various memory devices. Such images codes may generated by various endpoint computing devices, such as Internet of Things (IoT) computing devices, Such devices can generate a Hamming processing request, having an image code of the image, to compare that representation of the image to other images (e.g., in an image dataset) to identify a match or a set of neural network results. Advantageously, examples described herein may be used in neural networks to facilitate the processing of datasets, so as to increase the rate and amount of processing of such datasets. For example, comparisons of image codes can be performed 'closer' to the memory devices, e.g., at the memory controller coupled to memory devices.

Memory controllers including examples of calculating hamming distances for neural network and data center applications

Examples of systems and method described herein provide for the processing of image codes (e.g., a binary embedding) at a memory system including a Hamming processing unit. Such images codes may generated by various endpoint computing devices, such as Internet of Things (IoT) computing devices, Such devices can generate a Hamming processing request, having an image code of the image, to compare that representation of the image to other images (e.g., in an image dataset) to identify a match or a set of neural network results. Advantageously, examples described herein may be used in neural networks to facilitate the processing of datasets, so as to increase the rate and amount of processing of such datasets. For example, comparisons of image codes can be performed 'closer' to the memory devices, e.g., at a processing unit having memory devices.

Growing community of inventors

Seattle, WA, United States of America

David Hulton