199 results about "Reference memory" patented technology

A motion estimation apparatus, which can prevent a system breakdown without evenly lowering the quality of a picture to be decoded, includes: a reference picture setting unit which restricts a data transfer amount of image data to be transferred from an external multi-frame memory, in accordance with a data transfer capability of the external multi-frame memory for storing the image data; a reference local memory; a reference memory control unit which transfers, to the reference local memory, at least a portion of the image data stored in the external multi-frame memory, which is the data transfer amount restricted by the reference picture setting unit; and a motion estimation unit which performs motion estimation of the picture to be coded with reference to the at least the portion of the image data transferred to the reference local memory.

A computer-implemented method of accessing data comprises resetting the value of a register of a first processing core of a multi-core processor, copying the bits of a compressed pointer into the lowest order bits of the register, left shifting the register a predetermined number of bits, and executing on the first processing core a first instruction referencing memory at a virtual address specified by the register.

Methods and apparatus which facilitate the handling of data between platforms interconnected by any of a variety of network environments are disclosed. In general terms, the present invention represents an improvement over conventional packet parsing and searching mechanisms. The parse mechanism sequentially analyzes each character of the packet data and passes relevant characters to the search mechanism as soon as each character is reached. Preferably, the characters of each data field are parsed character-by-character. In one implementation, prior to searching a relevant data field, the parser initializes the appropriate search mechanism based on at least the data field type (e.g., the cookie field or URL field of an HTTP request). Each character of the relevant data is then passed sequentially to the search mechanism, where a search state are obtained for each passed character. Accordingly, the parser passes each character of the relevant data fields to the search mechanism. Since the parser passes the well-defined fields of standard protocols, such as HTTP and FTP, parsing may be efficiently performed without referencing memory (e.g., parsing is implemented in micro-code). When the parser reaches the end of the relevant data field, the parser may then cause the search mechanism to output search results associated with the search state of the last searched character. Alternatively, the parser may initiate another search for another data field, which is subsequently parsed and searched character-by-character as recited above for the first field.

The present disclosure relates to memory arrays with read reference voltage cells. In particular the present disclosure relates to variable resistive memory cell apparatus and arrays that include a high resistance state reference memory cell and a low resistance state reference memory cell that provides a reliable average reference voltage on chip to compare to a read voltage of a selected memory cell and determine if the selected memory cell is in the high resistance state or low resistance state. These memory arrays are particularly suitable for use with spin-transfer torque memory cells and resolves many systematic issues related to generation of a reliable reference voltage.

In a non-volatile memory system, a plurality of main memory cells for storing data is arranged in a data cell array and a plurality of reference memory cells is arranged in a reference cell array. The reference cell array includes first reference word lines connected to first reference memory cells and extending, second reference word lines connected to second reference memory cells and extending alternately with the first reference word lines, reference bit lines to which the first and the second reference memory cells are alternately connected in a line and a combined cell having a pair of the first and second reference memory cells and generating a reference signal for processing the data. The first and the second reference memory cells have different cell characteristics. The stability of the reference signal is improved irrespective of the differentiation of the first and the second reference memory cells.

Video coders use motion prediction, where a reference frame is used to predict a current frame. Most video compression standards require reference frame buffering and accessing. Given the randomized memory accesses to store and access reference frames, there are substantial overlapped areas. Conventional techniques fail to recognize this overlap, and perform duplicate loading, thereby causing increased memory traffic. Techniques disclosed herein reduce the memory traffic by avoiding the duplicated loading of overlapped area, by using a reference cache that is interrogated for necessary reference data prior to accessing reference memory. If the reference data is not in the cache, then that data is loaded from the memory and saved into the cache. If the reference data is in the cache, then that data is used instead of loading it from memory again. Thus, memory traffic is reduced by avoiding duplicated memory access to overlapped areas.

The present invention provides method and apparatus of digital video stream decoding by calculating the complexity of the B-type coded frame or macro-block and decoding whether the motion compensation can be skipped to save the time of accessing the referencing memory. The block with the decompressed accumulative pixel difference is less than a predetermined threshold and the motion vector equals to (0,0) or “Frame movement”, the motion compensation is skipped and one of the P-Type or I-type referencing frames will be accessed as the reference for motion compensation.

Systems and methods for an improved memory allocation service in embedded or wireless devices. Memory is allocated using a combination of container memory items and referencing memory items.

This invention discloses one artificial valuation FPGA physical prototype validation device through micro processor design, wherein the device comprises the following parts: one main control circuit with one processor interface for connection with validation micro processor; one power and clock generation circuit connected with micro process to be tested; one work memory and one reference memory connected to main control circuit through main control circuit memory interface, wherein the main control circuit has one communication interface with one outer work station.

Memory devices and methods are disclosed, such as those facilitating interpolation methods for reference memory cells based on their reference state and/or location in an array of memory cells. For example, a group of reference cells programmed to a subset of possible data states are utilized to interpolate for data states other than those in the subset. Such embodiments might be used to reduce the quantity of reference cells required.

The invention relates to a fast reading device of a phase change memory, which comprises a phase change memory cell to be read, a charging circuit, an overshoot recovery rate detecting circuit, a sensitive amplifier circuit and a reference level or a reference memory cell. On one hand, the phase change memory cell (which can be abstracted as a resistor in the circuit) and a parasitic capacitor ona digit line form a resistor-capacitor (RC) circuit, and on the other hand, a gated metal oxide semiconductor (MOS) pipe instantly enters a linear region from a saturation region when being switched on/of, thereby causing an overshoot phenomenon. For phase change resistors of different states, the corresponding recovery rates after the overshoot are different. By reading the recovery rate of the digit-line level after the overshoot and fast reading the state of the phase change memory cell, the reading rate of the whole memory is accelerated. In addition, fast reading is beneficial to avoiding the damage caused by reading operation to the phase change cell, thereby achieving the purpose of reducing reading interference.

A semiconductor memory device including a memory cell without a capacitor includes: a memory cell array block including first memory cells connected between a first bit line and first word lines and second memory cells connected between a second bit line and second word lines; and a reference memory cell array block including first reference memory cells connected between a first reference bit line connected to the first bit line and a first reference word line and second reference memory cells connected between a second reference bit line connected to the second bit line and a second reference word line. When the first word lines are selected, the second reference memory cells are selected, and when the second word lines are selected, the first reference memory cells are selected. Thus, each bit line includes a reference memory cell and outputs reference signal from the reference memory cell so that data can be precisely sensed during a read operation.

A magnetic memory device includes a memory cell array block and a reference memory cell array block having a plurality of magnetic memory cells arranged, respectively, at intersections of wordlines, digit lines, and bitlines, and reference wordlines, the digit lines, and a reference bitline, a first bitline clamp circuit coupled to a bitline to which a first selected magnetic memory cell is connected and supplying a first current to the first selected magnetic memory cell through the bitline, second and third bitline clamp circuits coupled to respective upper and lower ends of the reference bitline, for supplying a second current to selected magnetic memory cells in the reference memory cell array block through the reference bitline, and a sense amplifier for sensing and amplifying currents on first and second data lines, respectively connected to the bitline and the reference bitline, to judge data of the first selected magnetic memory cell.

The object of the present invention is to provide a DRAM, in which the operation speed for a sense amplifier can be increased. Bit line precharging circuits PCt and PCb are arranged to precharge bit lines BLt and /BLt to a ground voltage GND, and reference word lines RWLo and RWLe and reference memory cells RMC are arranged, so that when a word line WL is activated, a potential difference is always generated between the bit lines BLt and /BLt. The sources of transistors N10 and N11 of an N-type sense amplifier NSAt are connected directly to a ground terminal GND, and the sources of transistors P2 and P3 of a P-type sense amplifier PSA are connected directly to a power source VDD. The gates of the transistors N10 and N11 are connected to the bit lines /BLt and BLt, and the drains are connected to the bit lines BLt and /BLt, respectively. Shift word lines SWL and shift memory cells SMC are arranged, so that the N sense amplifier NSAt can amplify the potential difference between the bit lines BLt and /BLt.

A nonvolatile semiconductor memory device comprises a readout circuit which reads data stored in a selected memory cell by applying predetermined voltage to the selected memory cell and a reference cell such that currents corresponding to the respective threshold voltage may flow, and comparing the current flowing in the selected memory cell with the current flowing in the reference cell. The readout circuit commonly uses the reference cell set in the same storage state for normal readout and for readout for program verification, and when the predetermined voltage is applied to the selected memory cell and the reference memory cell at the time of the readout for the program verification, it sets an applying condition to the reference memory cell such that its storage state may be shifted more in the program state direction than that in an applying condition at the time of the normal readout.

The invention discloses a memory monitoring achievement method based on startup and shutdown of a Linux system, and belongs to the technical field of memory monitoring. The memory monitoring achievement method comprises the following specific steps: (1) ensuring that the Linux system is arranged on a to-be-tested main machine, enabling a user to log in, coping a warmboot script to a desktop, opening linux system termination program, switching a current directory into the desktop, executing the following operations as shown in the specification; (2) after the command is executed, capturing memory capacity under /proc/meminfo by a script as reference memory capacity; (3) generating a RebootResult file folder on the desktop, and recording the memory test condition by a sub-file holder under the file holder; (4) restarting the system according to the set time interval; (5) counting by reading a RebootResult/count file, capturing the memory capacity in the /proc/meminfo and comparing with the reference capacity; and (6) automatically carrying out the test frequency set by restart by the system until execution is ended. According to the memory monitoring achievement method, unattended operation and automatic test are achieved; the test is counted; the memory capacity is monitored; and the stability of the overall machine in a tester restart pressure test is evaluated.