Patents by Inventor Praveen Krishnamurthy

Praveen Krishnamurthy has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).

  • Publication number: 20200126100
    Abstract: Techniques are described for machine learning-based generation of target segments is leveraged in a digital medium environment. A segment targeting system generates training data to train a machine learning model to predict strength of correlation between a set of users and a defined demographic. Further, a machine learning model is trained with visit statistics for the users to predict the likelihood that the users will visit a particular digital content platform. Those users with the highest predicted correlation with the defined demographic and the highest likelihood to visit the digital content platform can be selected and placed within a target segment, and digital content targeted to the defined demographic can be delivered to users in the target segment.
    Type: Application
    Filed: October 23, 2018
    Publication date: April 23, 2020
    Applicant: Adobe Inc.
    Inventors: Praveen Kumar Goyal, Piyush Gupta, Nikaash Puri, Balaji Krishnamurthy, Arun Kumar, Atul Kumar Shrivastava
  • Patent number: 10580518
    Abstract: A system and method for performing similarity searching is disclosed wherein programmable logic devices such as field programmable gate arrays (FPGAs) can be used to implement Bloom filters for identifying possible matches between a query and data. The Bloom filters can be implemented in a parallel architecture where the different parallel Bloom filters share access to the same memory units. Further, a hash table may be generated to map a set of strings to keys. In other examples, the hash table may be used to map a set of substrings to a position in a larger string.
    Type: Grant
    Filed: January 11, 2017
    Date of Patent: March 3, 2020
    Assignee: WASHINGTON UNIVERSITY
    Inventors: Jeremy Daniel Buhler, Roger Dean Chamberlain, Mark Allen Franklin, Kwame Gyang, Arpith Chacko Jacob, Praveen Krishnamurthy, Joseph Marion Lancaster
  • Patent number: 10402323
    Abstract: In one embodiment of the present invention a cache unit organizes data stored in an attached memory to optimize accesses to compressed data. In operation, the cache unit introduces a layer of indirection between a physical address associated with a memory access request and groups of blocks in the attached memory. The layer of indirection—virtual tiles—enables the cache unit to selectively store compressed data that would conventionally be stored in separate physical tiles included in a group of blocks in a single physical tile. Because the cache unit stores compressed data associated with multiple physical tiles in a single physical tile and, more specifically, in adjacent locations within the single physical tile, the cache unit coalesces the compressed data into contiguous blocks. Subsequently, upon performing a read operation, the cache unit may retrieve the compressed data conventionally associated with separate physical tiles in a single read operation.
    Type: Grant
    Filed: October 28, 2015
    Date of Patent: September 3, 2019
    Assignee: NVIDIA CORPORATION
    Inventors: Praveen Krishnamurthy, Peter B. Holmquist, Wishwesh Gandhi, Timothy Purcell, Karan Mehra, Lacky Shah
  • Patent number: 10338820
    Abstract: A system architecture conserves memory bandwidth by including compression utility to process data transfers from the cache into external memory. The cache decompresses transfers from external memory and transfers full format data to naive clients that lack decompression capability and directly transfers compressed data to savvy clients that include decompression capability. An improved compression algorithm includes software that computes the difference between the current data word and each of a number of prior data words. Software selects the prior data word with the smallest difference as the nearest match and encodes the bit width of the difference to this data word. Software then encodes the difference between the current stride and the closest previous stride. Software combines the stride, bit width, and difference to yield final encoded data word. Software may encode the stride of one data word as a value relative to the stride of a previous data word.
    Type: Grant
    Filed: June 7, 2016
    Date of Patent: July 2, 2019
    Assignee: NVIDIA CORPORATION
    Inventors: Rouslan Dimitrov, Jeff Pool, Praveen Krishnamurthy, Chris Amsinck, Karan Mehra, Scott Cutler
  • Patent number: 9934145
    Abstract: In one embodiment of the present invention a cache unit organizes data stored in an attached memory to optimize accesses to compressed data. In operation, the cache unit introduces a layer of indirection between a physical address associated with a memory access request and groups of blocks in the attached memory. The layer of indirection—virtual tiles—enables the cache unit to selectively store compressed data that would conventionally be stored in separate physical tiles included in a group of blocks in a single physical tile. Because the cache unit stores compressed data associated with multiple physical tiles in a single physical tile and, more specifically, in adjacent locations within the single physical tile, the cache unit coalesces the compressed data into contiguous blocks. Subsequently, upon performing a read operation, the cache unit may retrieve the compressed data conventionally associated with separate physical tiles in a single read operation.
    Type: Grant
    Filed: October 28, 2015
    Date of Patent: April 3, 2018
    Assignee: NVIDIA Corporation
    Inventors: Praveen Krishnamurthy, Peter B. Holmquist, Wishwesh Gandhi, Timothy Purcell, Karan Mehra, Lacky Shah
  • Publication number: 20170351429
    Abstract: A system architecture conserves memory bandwidth by including compression utility to process data transfers from the cache into external memory. The cache decompresses transfers from external memory and transfers full format data to naive clients that lack decompression capability and directly transfers compressed data to savvy clients that include decompression capability. An improved compression algorithm includes software that computes the difference between the current data word and each of a number of prior data words. Software selects the prior data word with the smallest difference as the nearest match and encodes the bit width of the difference to this data word. Software then encodes the difference between the current stride and the closest previous stride. Software combines the stride, bit width, and difference to yield final encoded data word. Software may encode the stride of one data word as a value relative to the stride of a previous data word.
    Type: Application
    Filed: June 7, 2016
    Publication date: December 7, 2017
    Inventors: Rouslan DIMITROV, Jeff POOL, Praveen KRISHNAMURTHY, Chris AMSINCK, Karan MEHRA, Scott CUTLER
  • Publication number: 20170124255
    Abstract: A system and method for performing similarity searching is disclosed wherein programmable logic devices such as field programmable gate arrays (FPGAs) can be used to implement Bloom filters for identifying possible matches between a query and data. The Bloom filters can be implemented in a parallel architecture where the different parallel Bloom filters share access to the same memory units. Further, a hash table may be generated to map a set of strings to keys. In other examples, the hash table may be used to map a set of substrings to a position in a larger string.
    Type: Application
    Filed: January 11, 2017
    Publication date: May 4, 2017
    Inventors: Jeremy Daniel Buhler, Roger Dean Chamberlain, Mark Allen Franklin, Kwame Gyang, Arpith Chacko Jacob, Praveen Krishnamurthy, Joseph Marion Lancaster
  • Publication number: 20170123977
    Abstract: In one embodiment of the present invention a cache unit organizes data stored in an attached memory to optimize accesses to compressed data. In operation, the cache unit introduces a layer of indirection between a physical address associated with a memory access request and groups of blocks in the attached memory. The layer of indirection—virtual tiles—enables the cache unit to selectively store compressed data that would conventionally be stored in separate physical tiles included in a group of blocks in a single physical tile. Because the cache unit stores compressed data associated with multiple physical tiles in a single physical tile and, more specifically, in adjacent locations within the single physical tile, the cache unit coalesces the compressed data into contiguous blocks. Subsequently, upon performing a read operation, the cache unit may retrieve the compressed data conventionally associated with separate physical tiles in a single read operation.
    Type: Application
    Filed: October 28, 2015
    Publication date: May 4, 2017
    Inventors: Praveen KRISHNAMURTHY, Peter B. HOLMQUIST, Wishwesh GANDHI, Timothy PURCELL, Karan MEHRA, Lacky SHAH
  • Publication number: 20170123978
    Abstract: In one embodiment of the present invention a cache unit organizes data stored in an attached memory to optimize accesses to compressed data. In operation, the cache unit introduces a layer of indirection between a physical address associated with a memory access request and groups of blocks in the attached memory. The layer of indirection—virtual tiles—enables the cache unit to selectively store compressed data that would conventionally be stored in separate physical tiles included in a group of blocks in a single physical tile. Because the cache unit stores compressed data associated with multiple physical tiles in a single physical tile and, more specifically, in adjacent locations within the single physical tile, the cache unit coalesces the compressed data into contiguous blocks. Subsequently, upon performing a read operation, the cache unit may retrieve the compressed data conventionally associated with separate physical tiles in a single read operation.
    Type: Application
    Filed: October 28, 2015
    Publication date: May 4, 2017
    Inventors: Praveen KRISHNAMURTHY, Peter B. HOLMQUIST, Wishwesh GANDHI, Timothy PURCELL, Karan MEHRA, Lacky SHAH
  • Patent number: 9547680
    Abstract: A system and method for performing similarity searching is disclosed wherein programmable logic devices such as field programmable gate arrays (FPGAs) can be used to implement Bloom filters for identifying possible matches between a query and data. The Bloom filters can be implemented in a parallel architecture where the different parallel Bloom filters share access to the same memory units.
    Type: Grant
    Filed: August 19, 2013
    Date of Patent: January 17, 2017
    Assignee: Washington University
    Inventors: Jeremy Daniel Buhler, Roger Dean Chamberlain, Mark Allen Franklin, Kwame Gyang, Arpith Chacko Jacob, Praveen Krishnamurthy, Joseph Marion Lancaster
  • Patent number: 9287875
    Abstract: Circuits and methods for controlling electrical coupling by a load switch are disclosed. In an embodiment, the load switch includes a pass element, level-shift circuit and low-resistance active path. The pass element is configured to be coupled to a power supply and a load, and is configured to electrically couple the power supply with the load during ON-state and electrically decouple the power supply from the load during OFF-state. The level-shift circuit includes a first transistor and pull-up resistor and is configured to generate a level-shifted signal in response to an enable signal, and enable the ON-state and the OFF-state of the pass element based on first and second voltages of the level-shifted signal. The low-resistance active path is coupled in parallel with the pull-up resistor for shunting the pull-up resistor in the OFF-state by providing a path for a leakage current of the first transistor in the OFF-state.
    Type: Grant
    Filed: May 7, 2014
    Date of Patent: March 15, 2016
    Assignee: TEXAS INSTRUMENTS INCORPORATED
    Inventors: Alok Kumar, Fenish Padinjaroot Prakasan, Rajkumar Jayaraman, Kalyan Cherukuri, Praveen Krishnamurthy
  • Publication number: 20150326226
    Abstract: Circuits and methods for controlling electrical coupling by a load switch are disclosed. In an embodiment, the load switch includes a pass element, level-shift circuit and low-resistance active path. The pass element is configured to be coupled to a power supply and a load, and is configured to electrically couple the power supply with the load during ON-state and electrically decouple the power supply from the load during OFF-state. The level-shift circuit includes a first transistor and pull-up resistor and is configured to generate a level-shifted signal in response to an enable signal, and enable the ON-state and the OFF-state of the pass element based on first and second voltages of the level-shifted signal. The low-resistance active path is coupled in parallel with the pull-up resistor for shunting the pull-up resistor in the OFF-state by providing a path for a leakage current of the first transistor in the OFF-state.
    Type: Application
    Filed: May 7, 2014
    Publication date: November 12, 2015
    Applicant: TEXAS INSTRUMENTS INCORPORATED
    Inventors: Alok Kumar, Fenish Padinjaroot Prakasan, Rajkumar Jayaraman, Kalyan Cherukuri, Praveen Krishnamurthy
  • Publication number: 20140067830
    Abstract: A system and method for performing similarity searching is disclosed wherein programmable logic devices such as field programmable gate arrays (FPGAs) can be used to implement Bloom filters for identifying possible matches between a query and data. The Bloom filters can be implemented in a parallel architecture where the different parallel Bloom filters share access to the same memory units.
    Type: Application
    Filed: August 19, 2013
    Publication date: March 6, 2014
    Inventors: Jeremy Daniel Buhler, Roger Dean Chamberlain, Mark Allen Franklin, Kwame Gyang, Arpith Chacko Jacob, Praveen Krishnamurthy, Joseph Marion Lancaster
  • Patent number: 8619978
    Abstract: A method and a system allow accessing several of a user's controlled access accounts by presenting the credentials of only one of the accounts. The method may include (a) storing the credentials for each of the user's accounts; (b) receiving from the user credentials corresponding to any of the user's accounts; (c) presenting the received credentials to access the corresponding account; and (d) upon successful access of the corresponding account, using the stored credentials to access one or more of the user's accounts without requiring the user to present the corresponding credentials. For each of the user's accounts, the credentials are stored encrypted, using a randomly generated key, common to all the encrypted credentials. In addition, the randomly generated key is encrypted using the credentials of each of the accounts. In that manner, plain-text copies of neither the random key nor the credentials of the accounts need to be stored.
    Type: Grant
    Filed: December 21, 2007
    Date of Patent: December 31, 2013
    Assignee: PageBites, Inc.
    Inventors: Ralph Harik, Georges Harik, Praveen Krishnamurthy
  • Patent number: 8515682
    Abstract: A system and method for performing similarity searching is disclosed. This includes a programmable logic device configured to include a pipeline that comprises a matching stage, the matching stage being configured to receive a data stream comprising a plurality of possible matches between a plurality of data strings and a plurality of substrings of a query string. The pipeline may further include an ungapped extension prefilter stage located downstream from the matching stage, the prefilter stage being configured to shift through pattern matches between the data strings and the plurality of substrings of a query string and provide a score so that only pattern matches that exceed a user defined score will pass downstream from the prefilter stage. The matching stage may include at least one Bloom filter.
    Type: Grant
    Filed: March 11, 2011
    Date of Patent: August 20, 2013
    Assignee: Washington University
    Inventors: Jeremy Daniel Buhler, Roger Dean Chamberlain, Mark Allen Franklin, Kwame Gyang, Arpith Chacko Jacob, Praveen Krishnamurthy, Joseph Marion Lancaster
  • Publication number: 20110231446
    Abstract: A system and method for performing similarity searching is disclosed. This includes a programmable logic device configured to include a pipeline that comprises a matching stage, the matching stage being configured to receive a data stream comprising a plurality of possible matches between a plurality of data strings and a plurality of substrings of a query string. The pipeline may further include an ungapped extension prefilter stage located downstream from the matching stage, the prefilter stage being configured to shift through pattern matches between the data strings and the plurality of substrings of a query string and provide a score so that only pattern matches that exceed a user defined score will pass downstream from the prefilter stage. The matching stage may include at least one Bloom filter.
    Type: Application
    Filed: March 11, 2011
    Publication date: September 22, 2011
    Applicant: Washington University
    Inventors: Jeremy Daniel Buhler, Roger Dean Chamberlain, Mark Allen Franklin, Kwame Gyang, Arpith Chacko Jacob, Praveen Krishnamurthy, Joseph Marion Lancaster
  • Patent number: 7958192
    Abstract: A method automatically and efficiently propagates information using previously established social methods of communication (e.g., email, instant messaging (IM), social network message posting, and short message services). In conjunction with the method, a system collects and disseminates information on a social network. The system includes: (a) a server for providing a user interface (e.g., a user facing web site) which allows a user to exchange information with one or more external services, wherein the information exchange includes receiving a message from the user that the user intends to disseminate through the external services; (b) a message module which prepares the message received from the user for dissemination; and (c) a communication module coupled to the user interface, the message module and the external services to mediate the information exchange between the user and the external services.
    Type: Grant
    Filed: April 25, 2007
    Date of Patent: June 7, 2011
    Inventors: Ralph Harik, Georges Harik, Praveen Krishnamurthy
  • Patent number: 7917299
    Abstract: An apparatus and method for performing similarity searching on a data stream with respect to a query string are disclosed, where the data stream comprises a plurality of data substrings, and where the query string comprises a plurality of query substrings. A programmable logic device is used to filter the data stream to find a plurality of possible matches between the data substrings and a plurality of the query substrings, wherein the data substrings and the query substrings comprise a plurality of characters. From these possible matches, a determination can be made as to a similarity between the query string and at least a portion of the data stream.
    Type: Grant
    Filed: February 22, 2006
    Date of Patent: March 29, 2011
    Assignee: Washington University
    Inventors: Jeremy Daniel Buhler, Roger Dean Chamberlain, Mark Allen Franklin, Kwame Gyang, Arpith Chacko Jacob, Praveen Krishnamurthy, Joseph Marion Lancaster
  • Patent number: 7840511
    Abstract: A medical concept is learned about or inferred from a medical transcript. A probabilistic model is trained from medical transcripts. For example, the problem is treated as a graphical model. Discrimitive or generative learning is used to train the probabilistic model. A mutual information criterion can be employed to identify a discrete set of words or phrases to be used in the probabilistic model. The model is based on the types of medical transcripts, focusing on this source of data to output the most probable state of a patient in the medical field or domain. The learned model may be used to infer a state of a medical concept for a patient.
    Type: Grant
    Filed: September 5, 2007
    Date of Patent: November 23, 2010
    Assignee: Siemens Medical Solutions USA, Inc.
    Inventors: Romer E. Rosales, Praveen Krishnamurthy, R. Bharat Rao, Harald Steck
  • Publication number: 20100098081
    Abstract: Methods and systems for performing parallel membership queries to Bloom filters for Longest Prefix Matching, where address prefix memberships are determined in sets of prefixes sorted by prefix length. Hash tables corresponding to each prefix length are probed from the longest to the shortest match in the vector, terminating when a match is found or all of the lengths are searched. The performance, as determined by the number of dependent memory accesses per lookup, is held constant for longer address lengths or additional unique address prefix lengths in the forwarding table given that memory resources scale linearly with the number of prefixes in the forwarding table. For less than 2 Mb of embedded RAM and a commodity SRAM, the present technique achieves average performance of one hash probe per lookup and a worst case of two hash probes and one array access per lookup.
    Type: Application
    Filed: September 24, 2009
    Publication date: April 22, 2010
    Inventors: Sarang Dharmapurikar, Praveen Krishnamurthy, David Edward Taylor