Hui Tao Huang

US Patent:

8108550, Jan 31, 2012

Filed:

Oct 24, 2007

Appl. No.:

11/923513

Inventors:

Ankur Lahoti - Lucknow, IN
Hui Huang - Sunnyvale CA, US
Christian F. Beedgen - Cupertino CA, US

Assignee:

Hewlett-Packard Development Company, L.P. - Houston TX

International Classification:

G06F 15/16
G06F 15/173

US Classification:

709238, 709245, 707E17032

Abstract:

A unique identifier is assigned to a network node and is used to obtain an “asset model” corresponding to the node and to determine whether the node is a member of a particular category. An asset model is a set of information about a node (e. g. , the node's role within the enterprise, software installed on the node, and known vulnerabilities/weaknesses of the node). An identifier lookup module determines a node's identifier based on characteristics of the node (such as IP address, host name, network zone, and/or MAC address), which are used as keys into lookup data structures. A category lookup module determines whether a particular node is a member of (i. e. , within) a particular category using a transitive closure to model the categories (properties) that can be attached to an asset model. A transitive closure for a particular asset category is stored as a bitmap, similar to bitmap indexing.

US Patent:

8386443, Feb 26, 2013

Filed:

Oct 5, 2009

Appl. No.:

12/573690

Inventors:

Eric Brueggemann - Cupertino CA, US
Goutham Rao - San Jose CA, US
Mark Taylor - San Jose CA, US
Murali Bashyam - Fremont CA, US
Hui Huang - Sunnyvale CA, US

Assignee:

Dell Products L.P. - Round Rock TX

International Classification:

G06F 7/00
G06F 17/00

US Classification:

707692, 707609

Abstract:

A data de-duplication system is used with network attached storage and serves to reduce data duplication and file storage costs. Techniques utilizing both symlinks and hardlinks ensure efficient deletion file/data cleanup and avoid data loss in the event of crashes.

US Patent:

20210149729, May 20, 2021

Filed:

Dec 19, 2019

Appl. No.:

16/720717

Inventors:

- Mountain View CA, US
Hui Huang - Palo Alto CA, US

International Classification:

G06F 9/50
G06N 20/00
G06N 3/04

Abstract:

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, are described for scheduling tasks of ML workloads. A system receives requests to perform the workloads and determines, based on the requests, resource requirements to perform the workloads. The system includes multiple hosts and each host includes multiple accelerators. The system determines a quantity of hosts assigned to execute tasks of the workload based on the resource requirement and the accelerators for each host. For each host in the quantity of hosts, the system generates a task specification based on a memory access topology of the host. The specification specifies the task to be executed at the host using resources of the host that include the multiple accelerators. The system provides the task specifications to the hosts and performs the workloads when each host executes assigned tasks specified in the task specifications for the host.