A memory controller combines information about which memory component segments are not being refreshed with the information about which rows are going to be refreshed next, to determine, for the current refresh command, the total number of rows that are going to be refreshed. Based on this total number of rows, the memory controller selects how long to wait after the refresh command before issuing a next subsequent command. When the combination of masked segments and the refresh scheme results in less than the nominal number of rows typically refreshed in response to a single refresh command, the waiting period before the next command (e.g., non-refresh command) is issued may be reduced from the nominal minimum time period, thereby allowing the next command to be issued earlier.

In an embodiment, a method of accessing logic data stored in a differential memory using single-ended mode includes: storing second logic data in an auxiliary memory module of the differential memory by copying first logic data stored in a first main memory module of the differential memory into the auxiliary memory module; refreshing the first logic data; receiving a request for reading the first logic data; when refreshing the first logic data, fetching the second logic data when refreshing the first logic data in response to the request for reading the first logic data; and when not refreshing the first logic data, fetching the first logic data in response to the request for reading the first logic data.

Disclosed herein are mechanisms and methods for reducing power consumed by various DRAM technologies (e.g., high-capacity DRAM and/or 3D DRAM) which may impact battery life of the platform. These mechanisms and methods may opportunistically reduce the power consumed by DRAM by inhibiting periodic refresh commands to memory ranks that are not in-use. Since these mechanisms and methods may be based on enhancements to memory controllers, they may accordingly be operating system (OS) agnostic.

A method, computer program product, and computer system for receiving, by a computing device, an I/O request. The I/O request may be processed as a write miss I/O. One or more dirty pages associated with the write miss I/O may be placed into a tree according to a key. It may be determined whether one of a first event and a second event occurs. A data flush may be triggered for the tree when the first event occurs, and the data flush may be triggered for the data flush for the tree when the second event occurs.

Devices and techniques for temperature informed memory refresh are described herein. A temperature counter can be updated in response to a memory device write performed under an extreme temperature. Here, the write is performed on a memory device element in the memory device. The memory device element can be sorted above other memory device elements in the memory device based on the temperature counter. Once sorted to the top of these memory device elements, a refresh can be performed the memory device element.

A method and apparatus for performing opportunistic refreshes of memory banks is disclosed. Refresh circuitry in a memory controller performs a refresh on each bank of a multi-bank memory at least once during a given refresh interval. At the beginning of an interval, memory banks for which there are no pending transactions (e.g., reads or writes) may be refreshed. During a first portion of the interval, refresh may be skipped for memory banks for which transactions are pending. In a second portion of the interval, refreshes are performed on memory banks that have not been refreshed during the interval, which may cause some memory transactions to be delayed.

Devices and techniques for temperature informed memory refresh are described herein. A temperature counter can be updated in response to a memory device write performed under an extreme temperature. Here, the write is performed on a memory device element in the memory device. The memory device element can be sorted above other memory device elements in the memory device based on the temperature counter. Once sorted to the top of these memory device elements, a refresh can be performed the memory device element.

The present embodiments provide a system that supports self-refreshing operations in a memory device. During operation, the system transitions the memory device from an auto-refresh state, wherein a memory controller controls refreshing operations for the memory device, to a self-refresh state, wherein the memory device controls the refreshing operations. While the memory device is in the self-refresh state, the system sends progress information for the refreshing operations from the memory device to the memory controller. Next, upon returning from the self-refresh state to the auto-refresh state, the system uses the progress information received from the memory device to control the sequencing of subsequent operations by the memory controller.

An example apparatus according to an aspect of the present disclosure includes an address scrambler circuit including a sub-wordline scrambler circuit configured to receive a first subset of bits of a row hammer hit address. The sub-wordline scrambler circuit is configured to perform a first set of logical operations on the first subset of bits to provide a second subset of bits, and to perform a second set of logical operations on the first subset of bits and the second subset of bits to provide a third subset of bits of an row hammer refresh address.

A first memory controller receives an access command from a second memory controller, where the access command is timing non-deterministic with respect to a timing specification of a memory. The first memory controller sends at least one access command signal corresponding to the access command to the memory, wherein the at least one access command signal complies with the timing specification. The first memory controller determines a latency of access of the memory. The first memory controller sends feedback information relating to the latency to the second memory controller.