Qimonda Ag

Infineon Technologies Ag

Qimonda North America Corporation

Infineon Technologies North America Corp.

Josef Schnell

Oliver Kiehl

Wayne Frederick Ellis

Klaus Hummler

Lee Ward Collins

Octavian Beldiman

Jung Pill Kim

Jong Hoon Oh

Michael Dieter Richter

Jong-Hoon Oh

Michael A Killian

Ernst Stahl

Helmut Seitz

Meg Freebern

Jung Pil Kim

A multi-chip package is provided that has at least a first, second and third chip, each comprising a top and bottom surface. The multi-chip package also has a package substrate for interfacing with a printed circuit board (PCB). The chips and the package substrate are housed within an encapsulation material. The bottom surface of the first chip is attached to the package substrate. The top surface of the first chip has a first plurality of landing pads, which serve as a mechanical and electrical interface between the first and second chip. The bottom surface of the second chip has a second plurality of landing pads that serve as a mechanical and electrical interface between the second and first chip. Additionally, the top surface of the second chip has a third plurality of landing pads that serve as a mechanical and electrical interface between the second and third chip.

Multi-chip package with interconnected stacked chips

A refresh scheduler is configured to refresh memory cells of a memory device according to a plurality of refresh intervals. The various refresh intervals are determined in response to refresh errors.

Memory refresh system and method

Embodiments of the present invention generally provide techniques and apparatus for altering the functionality of a multi-chip package (MCP) without requiring entire replacement of the MCP. The MCP may be designed with a top package substrate designed to interface with an add-on package that, when sensed by the MCP, alters the functionality of the MCP.

Method and structure of expanding, upgrading, or fixing multi-chip package

Embodiments of the invention generally provide a system, method and memory device for accessing memory. One embodiment includes synchronization circuitry configured to determine timing skew between a first memory device and a second memory device, and introduce a delta delay to at least one of the first memory device and the second memory device to adjust the timing skew.

Method and apparatus for synchronizing memory enabled systems with master-slave architecture

Embodiments of the invention generally provide a system, method, and memory device for accessing memory. In one embodiment, a first memory device includes command decoding logic configured to decode commands issued to the first memory device and a second memory device, while command decoding logic of the second memory device is bypassed.

Method and apparatus for implementing memory enabled systems using master-slave architecture

The present invention is generally related to integrated circuit devices, and more particularly, to methods and systems of a multi-chip package (MCP) containing a self-diagnostic scheme for detecting errors in the MCP. The MCP generally comprises a controller, at least one volatile memory chip having error detection logic, at least one non-volatile memory chip, and at least one fail signature register for storing fail signature data related to memory errors detected in the MCP. The controller can poll the fail signature register for fail signature data related to memory errors stored therein. Upon detection of fail signature data, the controller can store the fail signature data on a fail signature register located on a non-volatile memory.

Self-diagnostic scheme for detecting errors

Embodiments of the invention generally provide an apparatus and technique for sharing an internally generated voltage between devices of a multi-chip package (MCP). The internally generated voltage may be shared via a conductive structure that electrically couples the devices and carries the internally generated voltage.

Structure to share internally generated voltages between chips in MCP

A memory includes a first macro chip, a spine chip, and a common substrate. The common substrate is configured to pass signals between the first macro chip and the spine chip. The first macro chip, the spine chip, and the common substrate provide a memory.

Memory configured on a common substrate

A memory includes a wafer having at least a first and second edge, at least one memory bank array, a data path, and a plurality of data pads. The data path is coupled to the memory bank array. The plurality of data pads are coupled to the data path and configured with the data path to bus data to and from the memory bank array. The data pads are further configured such that each of the data pads are located adjacent the first and second edges of the wafer. The memory component is configurable for alternative applications such that in a first application all of the data pads used to bus data are located only on the first edge of the wafer and such that in a second application at least one of the data pads used to bus data is located on the first edge of the wafer and at least one of the data pads used to bus data is located on the second edge of the wafer.

Edge pad architecture for semiconductor memory

A write burst stop command function is provided for a semiconductor memory device, and in particular for a memory device having a write latency, such as is common in a low power double data rate (DDR) dynamic random access memory (DRAM) device. In the memory device, when a write stop command is received, pulses that are generated for a column address strobe signal are terminated so that no further data already in the memory device is transferred into a memory array. When the write stop command is received at the beginning of a write operation prior to generation of the pulses in the column address strobe signal, a first-in first-out (FIFO) circuit is reset. The FIFO circuit is used to introduce a predetermined write latency to the write operation. The column address strobe signal is supplied to a column decoder associated with the memory array and to a data path circuit that transfers data to the memory array based on pulses in the column address strobe signal. In one embodiment, the pulses for the column address strobe signal are produced by a latch circuit based on a signal derived from the output of the write latency FIFO circuit and so an input to the latch is disabled in response to the write stop command to stop producing pulses for the column address strobe signal.

Write burst stop function in low power DDR sDRAM

A memory component includes at least one memory bank array, a DQ region, a clock tree, and a voltage generator. The memory component is configured in a semiconductor wafer. The at least one memory bank array is configured such that data is read out of it during a read operation. The clock tree is coupled to the DQ region and is configured for driving data during the read operation. The voltage generator is coupled to at least some components of the clock tree in order to provide at least some of the components of the clock tree with an increased voltage.

Boosted clock circuit for semiconductor memory

Methods and apparatus that determine, at a device (e.g., a DRAM device), a phase difference between two externally supplied timing signals such as a clock signal (CLK) and a data strobe signal (DQS) are provided. Adjustments may be made to timing of one of the signals itself or other internal memory signals that are, perhaps, utilized in circuits controlled by the DQS signal.

Data strobe synchronization for DRAM devices

An integrated circuit chip including a first delay circuit and a second delay circuit. The first delay circuit has a first delay circuit topology configured to delay a signal a first delay. The second delay circuit has a second delay circuit topology configured to provide a second delay in a circuit loop that is configured to be monitored and provide an oscillating signal. The second delay circuit topology is substantially the same as the first delay circuit topology and the first delay circuit is configured to be trimmed to adjust the first delay based on the second delay and the oscillating signal.

Integrated circuit chip having a first delay circuit trimmed via a second delay circuit

A delay lock loop circuit, in accordance with the present invention includes a delay lock loop unit having a power down mode. The delay lock loop unit includes a delay line having an input and an output. The input receives a first clock signal to generate a modified clock signal at the output. A phase detector is coupled to the input and the output of the delay line for comparing the first clock signal and the modified clock signal. A delay line control unit is coupled to the phase detector and the delay line for adjusting delay in the delay line in accordance with a control signal from the phase detector. A counter circuit is included for updating the delay lock loop unit to account for delay line drift during the power down mode by periodically generating an update signal which permits the delay lock loop unit to update in the power down mode.

Delay lock loop and update method with limited drift and improved power savings

A delay lock loop, in accordance with the present invention, includes a plurality of phase detectors each receiving a first clock signal and a second clock signal. Each phase detector includes a specified delay range for detecting phase differences between the first and second clock signals in that range. A delay line includes an input and an output. The first clock signal is received at the input, and the second clock signal includes a delayed first clock signal. An amount of delay is applied to the first clock signal, which is adjusted in the delay line in accordance with control signals of the phase detectors.

Growing community of inventors

Charlotte, VT, United States of America

Josef Schnell