Intel Corporation

Robert J Royer, Jr

Sanjeev N Trika

Eng Hun Ooi

Blaise B Fanning

John I Garney

Knut Stener Grimsrud

Richard L Coulson

Richard Paul Mangold

Brian R Mcfarlane

Mark A Schmisseur

Robert W Faber

Amber D Huffman

Raymond Scott Tetrick

David B Minturn

Shane Matthews

Anoop Mukker

Dale J Juenemann

Jeanna N Matthews

Ritesh B Trivedi

James A Boyd

Nitin V Sarangdhar

Michael W Williams

Jeffrey R Wilcox

Zvika Greenfield

Chia-Hung Sophia Kuo

Rick Coulson

Mukesh Kataria

Jason B Akers

Aaron K Martin

Christopher P Mozak

Kirk D Brannock

Lily Pao Looi

Bilal Khalaf

Suryaprasad Kareenahalli

Brent D Chartrand

Chai Huat Gan

Tomer Levy

Sowmiya Jayachandran

Konika Ganguly

Han Chau

Deepak Gandiga Shivakumar

Anthony M Constantine

Rebecca Z Loop

Sahar Khalili

Prasun Ratn

Dimpesh Patel

Virendra Vikramsinh Adsure

Alex Thomas

Suhas Nayak

Gary C Chow

Noam Lupovich

In a memory subsystem, a memory controller can put its physical interface (PHY) into a low power state when an associated memory device is in self-refresh. Instead of powering on the interface and then triggering the memory device to exit self-refresh, or instead waiting for the physical interface to be powered up prior to waking the memory device from self-refresh, the memory controller can instruct the PHY to send a self-refresh exit command to the memory device and power up the physical interface in parallel with the memory device coming out of self-refresh. The memory controller can power down a high speed clock path of the PHY and use a slower clock path to send the self-refresh exit command before powering the high speed clock path back up.

Fast self-refresh exit power state

Examples include techniques to reduce memory power consumption during a system idle state. Cores of a single socket multi-core processor may be mapped to different virtual non-uniform memory architecture (NUMA) nodes and a dynamic random access memory (DRAM) may be partitioned into multiple segments that are capable of having self-refresh operations separately deactivated or activated. Different segments from among the multiple segments of DRAM may be mapped to the virtual NUMA nodes to allow for a mechanism to cause memory requests for pinned or locked pages of data to be directed to a given virtual NUMA node.

Techniques to reduce memory power consumption during a system idle state

Examples are disclosed for configuring a solid state drive (SSD) to operate in a storage mode or a memory mode. In some examples, one or more configuration commands may be received at a controller for an SSD having one or more non-volatile memory arrays. The SSD may be configured to operate in at least one of a storage mode, a memory mode or a combination of the storage mode or the memory mode based on the one or more configuration commands. Other examples are described and claimed.

Techniques to configure a solid state drive to operate in a storage mode or a memory mode

A multilevel memory subsystem includes a persistent memory device that can access data chunks sequentially or randomly to improve read latency, or can prefetch data blocks to improve read bandwidth. A media controller dynamically switches between a first read mode of accessing data chunks sequentially or randomly and a second read mode of prefetching data blocks. The media controller switches between the first and second read modes based on a number of read commands pending in a command queue.

Dynamically changing between latency-focused read operation and bandwidth-focused read operation

Examples include techniques for emulating a non-volatile dual inline memory module (NVDIMM) in a computing platform using a non-volatile storage device. When a power up event occurs for the computing platform, a host memory buffer may be allocated in a system memory device and a backing store for the host memory buffer may be copied from the non-volatile storage device to the host memory buffer in the system memory device. When a power down event or a flush event occurs for the computing platform, the host memory buffer may be copied from the system memory device to the corresponding backing store for the host memory buffer in the non-volatile storage device. Thus, virtual NVDIMM functionality may be provided without having NVDIMM hardware in the computing platform.

NVDIMM emulation using a host memory buffer

A method is described. The method includes receiving notice of a page fault. A page targeted by a memory access instruction that resulted in the page fault residing in persistent memory without system memory status. In response to the page fault, updating page table information to include a translation that points to the page in persistent memory such that the page changes to system memory status without moving the page and system memory expands to include the page in persistent memory.

Fast page fault handling process implemented on persistent memory

The present disclosure provides devices and techniques to dynamically change the operating frequency of an interface where components on the interface have non-common clocks. An interface component may be provided with a frequency negotiation component to negotiate a shift in an operating frequency with other component on an interface where the different components have non-common clocks.

Dynamic adaptive clocking for non-common-clock interfaces

Apparatus, systems, and methods to manage memory latency operations are described. In one embodiment, an electronic device comprises a processor and a memory control logic to receive data from a remote memory device, store the data in a local cache memory, receive an error correction code indicator associated with the data, and implement a data management policy in response to the error correction code indicator. Other embodiments are also disclosed and claimed.

Memory latency management

Technologies for protecting data in an asymmetric volume (ASV) that includes a first storage device that supports device-based encryption and a second storage device that does not support device-based encryption. In embodiments the technologies enable disparate capabilities of the storage devices in an ASV to be exposed to a user. When a complete copy of targeted data identified by a user input for encrypted storage is not present on the first storage device, at least a portion of the targeted data stored on the second storage device is rewritten to the first storage device. When a complete copy of the targeted data is stored on the first storage device, one or more security operations are performed to obfuscate or erase any portion of the targeted data stored on the second storage device.

Technologies for protecting data in an asymmetric storage volume

A hardware platform includes a nonvolatile storage device that can store system firmware as well as code for the primary operating system for the hardware platform. The hardware platform includes a controller that determines the hardware platform lacks functional firmware to boot the primary operating system from the storage device. The controller accesses a firmware image from an external interface that interfaces a device external to the hardware platform, where the external device is a firmware image source. The controller provisions the firmware from the external device to the storage device and initiates a boot sequence from the provisioned firmware.

In-system provisioning of firmware for a hardware platform

A method is described that includes choosing between one of two different ways to cause a memory device to enter a specific one of multiple lower power states that each comprise lower power consumption than a highest low power state. The method also includes asserting a first signal on a first signal line that is coupled to a power management controller of the memory device to indicate to the power management controller that a sideband channel of a memory bus that is coupled to the memory device is activated. The method also includes causing the memory device to enter the specific one of the multiple lower power states by also performing the chosen one of a) sending an in-band signal on said memory bus coupled with said asserting of said first signal, said in-band signal specifying the specific one of the multiple lower power states; or, b) sending a second signal on a second signal line that identifies the specific one of the multiple lower power states.

Power control of a memory device through a sideband channel of a memory bus

An apparatus is described. The apparatus includes a package on package structure. The package on package structure includes an upper package and a lower package. One of the packages contain memory devices of a first type and the other of the packages contain memory devices of a second type. I/O connections on the underside of the upper package's substrate are vertically aligned with their corresponding, first I/O connections on the underside of the lower package's substrate. The first I/O connections are located outside second I/O connections on the underside of the lower package's substrate for the lower package.

I/O layout footprint for multiple 1LM/2LM configurations

Techniques to utilize a very low power state with a memory subsystem that includes one or more non-volatile memory devices and a volatile memory system. A memory controller is coupled with the one or more non-volatile memory devices and the volatile memory system. The memory controller comprising at least an embedded control agent and memory locations to store state information. The memory controller to selectively enable and disable the one or more non-volatile memory devices. The memory controller transfers the state information to the volatile memory system prior to entering a low power state. Control circuitry is coupled with the memory controller. The control circuitry to selectively enable and disable operation of the memory controller.

Architectures and techniques for providing low-power storage mechanisms

Growing community of inventors

Portland, OR, United States of America

Robert J Royer, Jr