Abstract: The method and system of the present invention superimposes read and write operations by connecting the global bit lines that are not selected to the Vdd. As a result, the respective local sense amplifiers for the non-selected global bit lines will just read and refresh the respective memory cells. This new approach results in smaller local sense amplifiers and one global sense amplifiers for several memory cells (and local sense amplifiers).

Abstract: A compact dynamic random access memory (DRAM) cell and highly efficient methods for using the DRAM cell are disclosed. The DRAM cell provides reading, writing, and storage of a data bit on an ASIC chip. The DRAM cell includes a first transistor acting as a pass gate and having a first source node, a first gate node, and a first drain node. The DRAM cell also includes a second transistor acting as a storage device and having a second drain node that is electrically connected to the first drain node to form a storage node. The second transistor also includes a second source node and a second gate node. The second source node is electrically floating, thus increasing the effective storage capacitance of the storage node.

Abstract: A method and apparatus are disclosed for efficiently doubling a first frequency of a first clock signal. A second clock signal at a second frequency is generated by dividing the first frequency of the first clock signal by two, such that the second frequency is half of the first frequency and a duty cycle of the second clock signal is 50%. Also, a set of phase-delayed clock signals is generated in response to the second clock signal such that the set of phase-delayed clock signals are delayed in phase with respect to the second clock signal. Further, the set of phase-delayed clock signals is combined to generate a third clock signal at a third frequency, such that the third frequency is twice that of the first frequency and a duty cycle of the third clock signal is 50%.

Abstract: In one aspect, the invention describes a mechanism for refreshing multiple memory words (rows) per refresh cycle, the number of simultaneously refreshed rows being programmable by a small number of inputs. In another aspect, the invention discloses a mechanism for refreshing all banks or a programmable number of banks simultaneously in a multi-bank memory. In yet another aspect, the present invention describes a mechanism for refreshing a programmable multiple memory rows and a programmable multiple banks simultaneously.

Abstract: The present invention relates to a system and method for providing distributed, highly configurable modular predecoding. The system includes a hierarchical memory structure, including a predecoder adapted to perform a first layer of address predecoding and at least one local predecoder interacting with the global predecoder and adapted to perform a second layer of address predecoding.

Abstract: The present invention relates to a synchronous self timed memory device. The device includes a plurality of memory cells forming a cell array, at least one local decoder interfacing with the cell array, at least one local sense amplifier and at least one local controller. The local sense amplifier interfaces with at least the decoder and cell array, and is adapted to precharge and equalize at least one line coupled thereto. The local controller interfaces with and coordinates the activities of at least the local decoder and sense amplifier.

Abstract: A system-on-chip (SOC) device or a random access memory (RAM) chip includes a RAM block. The RAM block includes memory cells, each of which has three transistors. Each memory cell is coupled to both a read bit line and a write bit line. A transparent continuous refresh mechanism has been implemented to read the content of a memory cell and re-write it back to the memory cell without disturbing the access (read/write) cycle, making refresh operations transparent to the system level. The continuous refresh mechanism includes a collision detection mechanism to prevent writing and reading the same memory cell at the same time.

Abstract: A random access memory cell (10) includes a first conductor line (12) and a second conductor line (14). A native device (16) is arranged to store charge. A high voltage threshold transistor (30) couples the native device to the first and second conductors.

Abstract: A random access memory cell (10) comprises a first conductor line (12) and a second conductor line (14). A native device (16) is arranged to store charge. A high voltage threshold transistor (30) couples the native device to the first and second conductors.

Abstract: The method and system of the present invention superimposes read and write operations by connecting the global bit lines that are not selected to the Vdd. As a result, the respective local sense amplifiers for the non-selected global bit lines will just read and refresh the respective memory cells. This new approach results in smaller local sense amplifiers and one global sense amplifiers for several memory cells (and local sense amplifiers).

Abstract: The method and system of the present invention superimposes read and write operations by connecting the global bit lines that are not selected to the Vdd. The respective local sense amplifiers for the non-selected global bit lines just read and refresh the respective memory cells resulting in smaller local sense amplifiers and one global sense amplifiers for several memory cells (and local sense amplifiers). In one embodiment, eight global bit lines are shared by one global sense amplifier and are multiplexed. Only one global bit line pair generates voltage development as an input to a respective local sense amplifier during a write operation, while the other three global bit line pairs are disconnected from their respective local sense amplifiers and thus have no voltage development. Thus, the respective activated sense amplifiers amplify only the cell data which reassembles a read and refresh operation.

Abstract: Present invention describes an efficient implementation of differential sensing in single ended DRAM arrays. According to one embodiment of the present invention, a respective local sense amplifier compares the accessed memory cell data with a dummy cell data in the opposite or adjacent block of the accessed block that is connected to a respective local bit line in the opposite ID block, amplifies the result of the comparison and puts the data on a global bit line. In one embodiment, the invention is process and temperature invariant using reference method and means for canceling cross coupling between read lines and write lines.

Abstract: A system-on-chip (SOC) device or a random access memory (RAM) chip includes a RAM block. The RAM block includes memory cells, each of which has three transistors. Each memory cell is coupled to both a read bit line and a write bit line. A transparent continuous refresh mechanism has been implemented to read the content of a memory cell and re-write it back to the memory cell without disturbing the access (read/write) cycle, making refresh operations transparent to the system level. The continuous refresh mechanism includes a collision detection mechanism to prevent writing and reading the same memory cell at the same time.

Abstract: A system-on-chip (SOC) device or a random access memory (RAM) chip includes a RAM block. The RAM block includes memory cells, each of which has three transistors. Each memory cell is coupled to both a read bit line and a write bit line. A transparent continuous refresh mechanism has been implemented to read the content of a memory cell and re-write it back to the memory cell without disturbing the access (read/write) cycle, making refresh operations transparent to the system level. The continuous refresh mechanism includes a collision detection mechanism to prevent writing and reading the same memory cell at the same time.

Abstract: In one aspect, the invention describes a mechanism for refreshing multiple memory words (rows) per refresh cycle, the number of simultaneously refreshed rows being programmable by a small number of inputs. In another aspect, the invention discloses a mechanism for refreshing all banks or a programmable number of banks simultaneously in a multi-bank memory. In yet another aspect, the present invention describes a mechanism for refreshing a programmable multiple memory rows and a programmable multiple banks simultaneously.