Microsoft Technology Licensing, LLC

Other

Badriddine Khessib

Bryan David Kelly

Sriram Govindan

Sriram Sankar

Laura Marie Caulfield

Anirudh Badam

Bikash Sharma

Shaun L Harris

John Joseph Siegler

Mark Santaniello

Mark Edward Shaw

Robin Andrew Alexander

Mihail Gavril Tarta

Xiaozhong Xing

Kushagra Vaid

Mark A Shaw

Woongki Baek

Jian Xu

Zhe Tan

Chris Ong

Iyswarya Narayanan

Jie Jerry Liu

Mallik Bulusu

Di Wang

Suman Kumar Nath

J Michael Andrewartha

Benjamin Franklin Cutler

Brandon Aaron Rubenstein

Jian Huang

Harry Raymond Rogers

Matthew Allen Faist

Aishwarya Ganesan

Jaya Naga Satish Bobba

Christopher W Hoder

Myeongjae Jeon

A memory management system includes a physical memory associated with a computing device and a memory manager. The memory manager is configured to manage a shared memory cache as part of a compression of the physical memory using a cache compression algorithm, wherein a compression block size for the compression is a single cache line size. The physical memory includes a sector translation table (STT) region and a sector memory region. The memory manager uses a memory descriptor defined by an STT entry having a cache line map and a plurality of sector pointers to load cache from the physical memory to a level 3 Cache. The cache line map contains cache line metadata including a size of each cache line, a location of the cache line in one of the sectors pointed to by the STT entry, and a plurality of flags.

Physical memory compression

In the embodiment a determination is made, for one or more applications being executed by the computing system, of an amount of the first or second memory being used by the one or more applications. Based on the determination, a portion of the memory resources of the third memory are configured to function with the first or second memory when it is determined that the amount of the first or second memory being used by the one or more applications is not sufficient for the memory needs of the one or more applications and a portion of the memory resources of the third memory are removed from functioning with the first or second memory when it is determined that the amount of the first or second memory being used by the one or more applications is more than is needed for the memory needs of the one or more applications.

Efficient sharing of non-volatile memory

A system includes reception of a request from a first application to create a virtual open-channel solid state drive associated with a first bandwidth and first capacity, association, in response to the request, of block addresses of a virtual address space of the first application with block addresses of one or more blocks of a first one of a first plurality of channels of a first open-channel solid state drive and with block addresses of one or more blocks of a second one of the first plurality of channels, reception, from the first application, of a first I/O call associated with one or more block addresses of the virtual address space, determination of block addresses of one or more blocks of the first one of the first plurality of channels which are associated with the one or more block addresses of the virtual address space, and execution of the first I/O call on the determined block addresses of one or more blocks of the first one of the first plurality of channels.

Virtualized OCSSDs spanning physical OCSSD channels

A Root of Trust hardware hierarchy provides firmware security for motherboard and peripheral devices. Power is received at a computer system and, in response to the receipt of power, of a standby power rail of a motherboard of the computer system is energized, and a first microcontroller mounted on the motherboard authenticates first firmware associated with a baseboard management controller mounted on the motherboard and coupled to the first microcontroller. If the authentication of the first firmware is successful, the baseboard management controller is powered on, a central processing unit coupled to the first microcontroller is held in reset, and a standby power rail of a peripheral component card coupled to the motherboard is energized. Second firmware associated with the central processing unit is authenticated using the first microcontroller and a second microcontroller mounted on the peripheral component card authenticates third firmware associated with a system on chip mounted on the peripheral component card and coupled to the second microcontroller.

Hardware-enforced firmware security

Technology for handling overcurrent conditions on electrical circuits that power multiple computing modules is disclosed. Aspects of the technology include a power system adapted to provide notifications of overcurrent conditions, and computing modules adapted to reduce an operating speed thereof in response to notification of an overcurrent condition.

Low latency computer system power reduction

A system includes reception of a request from a first application to create a virtual open-channel solid state drive associated with a first bandwidth and first capacity, association, in response to the request, of block addresses of a virtual address space of the first application with block addresses of one or more blocks of a first one of a first plurality of channels of a first open-channel solid state drive and with block addresses of one or more blocks of a second one of a second plurality of channels of a second open-channel solid state drive, reception, from the first application, of a first I/O call associated with one or more block addresses of the virtual address space, determination of block addresses of one or more blocks of the first one of the first plurality of channels which are associated with the one or more block addresses of the virtual address space, and execution of the first I/O call on the determined block addresses of one or more blocks of the first one of the first plurality of channels.

Aspects extend to methods, systems, and computer program products for remediating power loss at a server. Aspects of the invention increase the likelihood of gracefully shutting down a server and associated components in a data center when mains power is lost for a specified amount of time (e.g., an amount of time beyond transition to generator power). A server can include a management module (e.g., a BMC) and a watchdog module. When the management controller detects loss of power at a power supply unit, the management controller orchestrates a graceful shutdown of the server in response to power loss. When the management module is unresponsive, the watchdog module provides backup functionality for orchestrating a graceful shutdown in response to power loss. As such, data can be saved from RAM to more durable storage even when the management module is unresponsive.

Remediating power loss at a server

Various techniques for managing power backup for computing devices are disclosed herein. In one embodiment, a method includes receiving data representing a backup capacity of one or more backup power units and data representing a backup power profile of one or more processing units sharing the one or more backup power units. A portion of the backup capacity may then be assigned to each of the one or more processing units based at least in part on both the received data representing the backup capacity of the one or more backup power units and the received data representing the profile of the one or more processing units.

Backup power management for computing systems

Aspects extend to methods, systems, and computer program products for predicting solid state drive reliability. Aspects of the invention can be used to predict and/or to configure a data center to minimize one or more of: SSD capacity degradation (how much storage an SSD has left), SSD performance degradation (reduced read/write latency/throughput), and SSD failure. Models and data center considerations can be based on device level SSD related operations, such as, for example, read, write, erase. Operations decisions can be made for a data center based on SSD specific features, such as, for example, remaining capacity, write amplification factor, etc. Dependence and/or causality of various different data center factors can be leveraged. The impact of the various data center factors on different SSD failure modes and capacity/performance degradation can be quantified to drive SSD design, SSD provisioning, and SSD operations.

Predicting solid state drive reliability

Operations of a variety of components of a storage system stack are redefined to make the system more efficient when the underlying media has a 'multi-log' type interface such as the case with NAND flash SSD memory or shingled magnetic recording media. The responsibilities of components of the storage system stack are modified such that each responsibility is performed at the most efficient component (level of abstraction) of the storage stack.

Data storage system with data storage devices operative to manage storage device functions specific to a particular data storage device

Various techniques of managing storage devices in a computing system are described in this application. In one embodiment, a method includes receiving an input containing consumption data representing consumption of a storage device in one of the processing units and determining if the storage device in one of the processing units is consumed excessively. In response to determining that the storage device is consumed excessively, an indicator may be generated to indicate a potential program migration from the one of the processing units to another one of the processing units in the computing system.

Storage device management in computing systems

A volatile memory data save subsystem may include a coupling to a shared power source such as a chassis or rack battery, or generator. A data save trigger controller sends a data save command toward coupled volatile memory device(s) such as NVDIMMs and PCIe devices under specified conditions: a programmable amount of time passes without AC power, a voltage level drops below normal but is still sufficient to power the volatile memory device during a data save operation, the trigger controller is notified of an operating system shutdown command, or the trigger controller is notified of an explicit data save command without a system shutdown command. NVDIMMs can avoid reliance on dedicated supercapacitors and dedicated batteries. An NVDIMM may perform an asynchronous DRAM reset in response to the data save command. Voltage step downs may be coordinated among power supplies. After data is saved, power cycles and the system reboots.

Configurable volatile memory without a dedicated power source for detecting a data save trigger condition

Management of infrastructure devices is performed by computing devices that are associated with the processing being provided by the data center, such as chassis managers. A master is first selected through polling or consensus algorithms, and then subsequently the master is endowed with the authority to manage infrastructure devices and generate the control output to such infrastructure devices. Alternatively, no master need be elected and, instead, output to such infrastructure devices is generated by a computing device selected utilizing polling or consensus algorithms, and in accordance with a management decision made through polling or consensus algorithms. The interplay between the cooling apparatuses of individual server computing devices and the cooling apparatuses of the data center as a whole is also managed to increase the portion of the cooling of server computing devices provided by data center air movers. Control of data center air movers can be determined empirically or predictively.

Growing community of inventors

Redmond, WA, United States of America

Badriddine Khessib