63 results about "Interleaved memory" patented technology

A memory controller provides programmable flexibility, via one or more configuration registers, for the configuration of the memory. The memory may be optimized for a given application by programming the configuration registers. For example, in one embodiment, the portion of the address of a memory transaction used to select a storage location for access in response to the memory transaction may be programmable. In an implementation designed for DRAM, a first portion may be programmably selected to form the row address and a second portion may be programmable selected to form the column address. Additional embodiments may further include programmable selection of the portion of the address used to select a bank. Still further, interleave modes among memory sections assigned to different chip selects and among two or more channels to memory may be programmable, in some implementations. Furthermore, the portion of the address used to select between interleaved memory sections or interleaved channels may be programmable. One particular implementation may include all of the above programmable features, which may provide a high degree of flexibility in optimizing the memory system.

A staggered memory cell architecture staggers memory cells on opposite sides of a shared bit line preserves memory cell density, while increasing the distance between such memory cells, thereby reducing the possibility of a disturb. In one implementation, the memory cells along a first side of a shared bit line are connected to a set of global word lines provided underneath the memory structure, while the memory cells on the other side of the shared bit line —which are staggered relative to the memory cells on the first side—are connected to global word lines above the memory structure.

A ferroelectric memory with interleaved pairs of ferroelectric memory cells of the two-transistor, two-capacitor (2T2C) type. Each memory cell in a given pair is constructed as first and second portions, each portion including a transistor and a ferroelectric capacitor. Within each pair, a first portion of a second memory cell is physically located between the first and second portions of the first memory cell. As a result, complementary bit lines for adjacent columns are interleaved with one another. Each sense amplifier is associated with a multiplexer, so that the adjacent columns of the interleaved memory cells are supported by a single sense amplifier. Noise coupling among the bit lines is reduced, and the sense amplifiers can be placed along one side of the array, reducing the number of dummy cells required to eliminate edge cell effects.

A predictive memory performance optimizing unit for use with an interleaved memory, for example a DDR SDRAM memory, and suitable for use in a computer graphics system, among others, is described. The unit maintains a queue of pending requests for data from the memory, and prioritizes precharging and activating interleaves with pending requests. Interleaves which are in a ready state may be accessed independently of the precharging and activation of non-ready interleaves. The unit utilizes idle cycles occurring between consecutive requests to activate interleaves with pending requests.

An interconnect for an integrated circuit communicating transactions between initiator Intellectual Property (IP) cores and multiple target IP cores coupled to the interconnect is generally described. The interconnect routes the transactions between the target IP cores and initiator IP cores in the integrated circuit. A first aggregate target of the target IP cores includes two or more memory channels that are interleaved in an address space for the first aggregate target in the address map. Each memory channel is divided up in defined memory interleave segments and then interleaved with memory interleave segments from other memory channels. An address map is divided up into two or more regions. Each interleaved memory interleave segment is assigned to at least one of those regions and populates the address space for that region, and parameters associated with the regions and memory interleave segments are configurable.

A heat sink comprises a heat sink base and a row of heat sink extensions that are attached to one side of the heat sink base. An interleaved heat sink structure includes a first row and a second row of heat sink extensions. The first row and the second row of heat sink extensions are coupled respectively to a first and a second heat sink bases. The first and the second heat sink bases are thermally coupled to a first plurality of memory packages and a second plurality of memory packages, respectively. The first row of heat sink extensions is parallel to and at least partially interleaved with the second row of heat sink extensions. A memory heat dissipation control system and a method for assembly a memory part that includes a DIME and two heat sinks are also described.

An interleaved memory programming and verification method, device and system includes a memory array includes first and second memory banks of memory cells. The memory device further includes a controller configured to concurrently program a first data into the first memory bank and a second data into the second memory bank using iterative programming and verification operations in each of the first and second memory banks with the programming and verification operations in the second memory bank being offset from the programming and verification operations in the first memory bank.

A dual-mode memory chip supports a first operation mode in which received data access commands contain chip select data to identify the chip addressed by the command, and control logic in the memory chip determines whether the command is addressed to the chip, and a second operation mode in which the received data access command addresses a set of multiple chips. Preferably, the first mode supports a daisy-chained configuration of memory chips. Preferably the second mode supports a hierarchical interleaved memory subsystem, in which each addressable set of chips is configured as a tree, command and write data being propagated down the tree, the number of chips increasing at each succeeding level of the tree.

In exemplary embodiments, a skewed interleaving function for iterative code systems is described. The skewed interleaving function provides a skewed row and column memory partition and a layered structure for re-arranging data samples read from, for example, a first channel detector. An iterative decoder, such as an iterative decoder based on a low-density parity-check code (LDPC), might employ an element to de-skew the data from the interleaved memory partition before performing iterative decoding of the data, and then re-skew the information before passing decoded samples to the de-interleaver. The de-interleaver re-arranges the iterative decoded data samples in accordance with an inverse of the interleaver function before passing the decoded data samples to, for example, a second channel detector.

Memory transactions are carried out in an order that maximizes concurrency in a memory system such as a multi-bank interleaved memory system. Read data is collected in a buffer memory to be presented back to the bus in the same order as read transactions were requested. An adaptive algorithm groups writes to minimize overhead associated with transitioning from reading to writing into memory.

A method for optimizing performance of memory in an information handling system which includes determining whether memory within the information handing system is being accessed in a flexible interleaving memory mode of operation, when the memory is being accessed in the flexible interleaving memory mode of operation, identifying which of the memory is configured as interleaved memory and which of the memory is configured as non-interleaved memory, and configuring the memory such that the interleaved memory is accessed prior to the non-interleaved memory being accessed is disclosed.

Embodiments of the present invention may provide a method of sparing and removing pinned or interleaved memory. When a memory device failure is predicted in a device containing pinned memory, a request may be made for the de-allocation of a freeable memory range 304. When the request for de-allocating the freeable range of memory is accepted 306, the memory data from the failing memory device may be copied to one or more de-allocated memory devices 308. Requests directed to the failing memory device may be re-routed to the replacement memory device(s) 310 and the memory without the deactivated memory device 312 may be re-interleaved.

A switch/network adapter port (“SNAP™”) in a dual in-line memory module (“DIMM”) or Rambus™ in-line memory module (“RIMM”) format for clustered computers employing multi-adaptive processor (“MAP®”, both trademarks of SRC Computers, Inc.) elements for use with interleaved memory controllers. Particularly disclosed is a microprocessor based computer system utilizing either a DIMM or RIMM physical format adapter port coupled to a reconfigurable processor element for the purpose of implementing a connection to an external switch, network, or other device. In a particular embodiment, connections may be provided to either the PCI, accelerated graphics port (“AGP”) or system maintenance (“SM”) bus for purposes of passing control information to the host microprocessor or other control chips. The field programmable gate array (“FPGA”) based processing elements have the capability to alter data passing through it to and from an external interconnect fabric or device.

A design structure is provided for a dual-mode memory chip supporting a first operation mode in which received data access commands contain chip select data to identify the chip addressed by the command, and control logic in the memory chip determines whether the command is addressed to the chip, and a second operation mode in which the received data access command addresses a set of multiple chips. Preferably, the first mode supports a daisy-chained configuration of memory chips. Preferably the second mode supports a hierarchical interleaved memory subsystem, in which each addressable set of chips is configured as a tree, command and write data being propagated down the tree, the number of chips increasing at each succeeding level of the tree.

A convolutional neural network (CNN) for an image processing system comprises an image cache responsive to a request to read a block of N*M pixels extending from a specified location within an input map to provide a block of N*M pixels at an output port. A convolution engine reads blocks of pixels from the output port, combines blocks of pixels with a corresponding set of weights to provide a product, and subjects the product to an activation function to provide an output pixel value. The image cache comprises a plurality of interleaved memories capable of simultaneously providing the N*M pixels at the output port in a single clock cycle. A controller provides a set of weights to the convolution engine before processing an input map, causes the convolution engine to scan across the input map by incrementing a specified location for successive blocks of pixels and generates an output map within the image cache by writing output pixel values to successive locations within the image cache.

A mechanism implemented by a controller enables efficient access to an interleaved memory system that includes M modules, M being (2ⁿ+1) or (2ⁿ−1), n being a positive integer number. Upon receiving an address N, the controller performs shift and add/subtract operations to obtain a quotient of N divided by M based on a binomial series expansion of N over M. The controller computes a remainder of N divided by M based on the quotient. The controller then accesses one of the modules in the memory based on the remainder.

An interleaving method and apparatus provides parallel access in a linear and interleaved order to a predetermined number of stored data samples. A memory array with a plurality of memory devices is addressed by applying a first portion of an address to memory devices and by using a second portion of the address to select at least one memory device to be accessed, wherein the position of the first and second portions within the address is changed in response to a change between the linear order and the interleaved order. Due to the fact that the memory array is split into several individually addressable memory devices, each of these memory devices can be accessed in a linear and interleaved order by changing an allocation of a chip selection portion and a chip addressing portion of the address.

Successive sequences of interleaved data samples extracted from a virtual memory having L0 columns and C0 rows are de-interleaved. The de-interleaving includes receiving each sequence of the interleaved data samples, and writing row by row the received sequences of interleaved data samples in a de-interleaving memory array having L rows and C columns, with L being greater or equal to L0 and C being greater or equal to C0. The data samples stored in the de-interleaving memory array are de-interleaved sub-array by sub-array. Each sub-array is a square cluster array having a number SQ of rows and columns. A cluster array is a row of the square cluster array comprising SQ data samples, with the number L of rows and the number C of columns of the de-interleaving memory array being multiples of the number SQ of rows and columns.

An apparatus and method for storing log likelihood ratios in an interleaved form comprising receiving a plurality of interleaved codewords (710); obtaining at least one log likelihood ratio (LLR) for the plurality of interleaved codewords (720); storing the at least one LLR in a memory (730); deinterleaving the plurality of interleaved codewords after the at least one LLR has been stored in the memory (740); and performing a bit decision of the deinterleaved codewords using the stored at least one LLR (750).

A switch/network adapter port (“SNAP™”) in a dual in-line memory module (“DIMM”) or Rambus™ in-line memory module (“RIMM”) format for clustered computers employing multi-adaptive processor (“MAP®”, both trademarks of SRC Computers, Inc.) elements for use with interleaved memory controllers. Particularly disclosed is a microprocessor based computer system utilizing either a DIMM or RIMM physical format adapter port coupled to a reconfigurable processor element for the purpose of implementing a connection to an external switch, network, or other device. In a particular embodiment, connections may be provided to either the PCI, accelerated graphics port (“AGP”) or system maintenance (“SM”) bus for purposes of passing control information to the host microprocessor or other control chips. The field programmable gate array (“FPGA”) based processing elements have the capability to alter data passing through it to and from an external interconnect fabric or device.