Abstract: Certain embodiments herein relate to, among other things, designing data cache systems to enhance energy efficiency and performance of computing systems. A data filter cache herein may be designed to store a portion of data stored in a level one (L1) data cache. The data filter cache may reside between the L1 data cache and a register file in the primary compute unit. The data filter cache may therefore be accessed before the L1 data cache when a request for data is received and processed. Upon a data filter cache hit, access to the L1 data cache may be avoided. The smaller data filter cache may therefore be accessed earlier in the pipeline than the larger L1 data cache to promote improved energy utilization and performance. The data filter cache may also be accessed speculatively based on various conditions to increase the chances of having a data filter cache hit.
Type:
Grant
Filed:
March 3, 2017
Date of Patent:
October 2, 2018
Assignee:
Florida State University Research Foundation, Inc.
Inventors:
David Whalley, Magnus Själander, Alen Bardizbanyan, Per Larsson-Edefors
Abstract: Embodiments of the present invention provide a method and a system for sharing a storage resource, where storage resources of a hard disk and a storage array are divided into multiple storage partitions and the multiple storage partitions form a shared storage resource pool, a read-write control module is allocated to each storage partition, and global partition information is generated to record a correspondence between each storage partition in the shared storage resource pool and the read-write control module. The embodiments of the present invention implement fast and simple integration of heterogeneous storage resources, and therefore can efficiently use various storage resources, save cost, and avoid a resource waste.
Type:
Grant
Filed:
August 11, 2017
Date of Patent:
September 25, 2018
Assignee:
HUAWEI TECHNOLOGIES CO., LTD.
Inventors:
Jiongjiong Gu, Xiaoyong Min, Daohui Wang
Abstract: Cluster families for cluster selection and cooperative replication are created. The clusters are grouped into family members of a cluster family base on their relationships and roles. Members of the cluster family determine which family member is in the best position to obtain replicated information and become cumulatively consistent within their cluster family. Once the cluster family becomes cumulatively consistent, the data is shared within the cluster family so that all copies within the cluster family are consistent.
Type:
Grant
Filed:
May 1, 2017
Date of Patent:
September 11, 2018
Assignee:
INTERNATIONAL BUSINESS MACHINES CORPORATION
Inventors:
Thomas W. Bish, Takeshi Nohta, Joseph M. Swingler, Rufus-John Y. Twito
Abstract: Example implementations relate to buffer-based update of state data. In example embodiments, a computing device may obtain current state data from a first buffer and determine next state data based at least on the current state data. The computing device may also obtain last difference data that corresponds, for example to current state data that is different from previous state data. The last difference data may be obtained, for example, using a difference data identifier. The computing device may store the last difference data in the second memory buffer. The computing device may also store in the second memory buffer new difference data corresponding to differences between the next state data and the current state data. The computing device may also update the difference data identifier based on the new difference data.
Type:
Grant
Filed:
June 27, 2014
Date of Patent:
September 4, 2018
Assignee:
Hewlett Packard Enterprise Development LP
Abstract: A method, non-transitory computer readable medium, and computing device that receives metadata for a block associated with an object from a source storage node. The metadata comprises a source object identifier and the object is associated with a source volume of a source aggregate owned by the source storage node. A determination is made when another block associated with the object has been previously received. A destination object identifier is obtained based on the source object identifier, when the determining indicates that the other block associated with the object has been previously received. A new aggregate block number is assigned to the block based on the destination object identifier and another portion of the metadata. Ownership of the source volume is transferred upon receipt of an indication of a cutover from the source storage node in order to migrate the source volume to a destination volume of a destination aggregate.