Patents by Inventor Keerthinarayan P. Heragu

Keerthinarayan P. Heragu 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).

  • Patent number: 8301431
    Abstract: A method of accelerating a Monte Carlo (MC) simulation for a system including a first component having a first input parameter and a second component having a second input parameter. The simulation model provided includes a first component model including a first model parameter corresponding to the first input parameter and a second component model having a second model parameter corresponding to the second input parameter. A first acceleration factor for the first component and a second acceleration factor for the second component are calculated based on at least the respective number of instances. A first scaled distribution is computed from the first distribution and a second scaled distribution is computed from the second distribution based on the respective acceleration factors. The MC simulation for the system is run, wherein values for the first model parameter value and second model parameter value are obtained based on the respective scaled distributions.
    Type: Grant
    Filed: December 30, 2008
    Date of Patent: October 30, 2012
    Assignee: Texas Instruments Incorporated
    Inventors: Keerthinarayan P. Heragu, Theodore W. Houston, Anand Seshadri, Hugh T. Mair
  • Patent number: 8266464
    Abstract: Embodiments of the present disclosure provide a power controller, a method of operating a power controller and a semiconductor memory system. In one embodiment, the power controller is for use with a memory and includes an access module configured to provide an active state of the memory to allow memory access. The power controller also includes a retain-till-access module configured to cycle a portion of the memory between the active state and a low leakage data retention state of the memory. The power controller further includes an expanded retain-till-access module configured to extend the active state of the memory for a specified period of time before returning the memory to the low leakage data retention state.
    Type: Grant
    Filed: October 22, 2007
    Date of Patent: September 11, 2012
    Assignee: Texas Instruments Incorporated
    Inventors: Michael P. Clinton, Keerthinarayan P. Heragu, Uming U. Ko
  • Patent number: 7932756
    Abstract: Various systems and methods for delaying a signal relative to another signal are disclosed. As one example, a delay lock loop circuit is disclosed that includes at least two delay stages. Each of the aforementioned delay stages include a plurality of selectable delay elements. Such selectable delay elements may be, but are not limited to, a plurality of single input buffers, and a plurality of multiple input logic gates. Further, a first of the delay stages is selectably driven by one of a first signal and a reference signal, and the stage provides a first stage output. A second of the delay stages is selectably driven by one of a second signal and the first stage output, and the stage provides a second stage output. The circuit further includes a mode signal that has at least two states.
    Type: Grant
    Filed: August 1, 2007
    Date of Patent: April 26, 2011
    Assignee: Texas Instruments Incorporated
    Inventors: Keerthinarayan P. Heragu, Padattil K. Nisha
  • Patent number: 7719332
    Abstract: Delay lock loop circuits are described, which may include two or more delay stages that each includes a plurality of selectable delay elements. A reference signal drives an input of the first delay stage, which provides a first output. The first output drives an input of the second delay stage, which provides a second output. The circuits further include a first selector register that is associated with the first delay stage. A value maintained in the first selector register determines a number of the selectable delay elements utilized in the first delay stage. The circuits further include a second selector register associated with the second delay stage. A value maintained in the second selector register determines a number of the selectable delay elements utilized in the second delay stage. Modification of the values maintained in the first and second selector registers are synchronized to the first and second outputs, respectively.
    Type: Grant
    Filed: August 1, 2007
    Date of Patent: May 18, 2010
    Assignee: Texas Instruments Incorporated
    Inventors: Keerthinarayan P. Heragu, Padattil K. Nisha
  • Patent number: 7573307
    Abstract: Various systems and methods for signal synchronization are disclosed. For example, some embodiments of the present invention provide methods for reduced area delay signal timing. Such methods include providing a delay lock loop circuit with a plurality of selectable delay elements. The methods further include operating the delay lock loop circuit in a first mode where a program number is established in relation to a reference frequency. The program number corresponds to a number of the plurality of selectable delay elements used to establish a first delay time at the reference frequency. The program number is multiplied by a multiplicand, and the product of the multiplication is used while operating the delay lock loop circuit in a second mode to select the number of delay elements utilized in delaying an input signal. In the second mode, an input signal is delayed by a second delay time that is approximately the first delay time multiplied by the multiplicand.
    Type: Grant
    Filed: August 1, 2007
    Date of Patent: August 11, 2009
    Assignee: Texas Instruments Incorporated
    Inventors: Keerthinarayan P. Heragu, Padattil K. Nisha
  • Publication number: 20090177451
    Abstract: A method of accelerating a Monte Carlo (MC) simulation for a system including a first component having a first input parameter and a second component having a second input parameter. The simulation model provided includes a first component model including a first model parameter corresponding to the first input parameter and a second component model having a second model parameter corresponding to the second input parameter. A first acceleration factor for the first component and a second acceleration factor for the second component are calculated based on at least the respective number of instances. A first scaled distribution is computed from the first distribution and a second scaled distribution is computed from the second distribution based on the respective acceleration factors. The MC simulation for the system is run, wherein values for the first model parameter value and second model parameter value are obtained based on the respective scaled distributions.
    Type: Application
    Filed: December 30, 2008
    Publication date: July 9, 2009
    Applicant: TEXAS INSTRUMENTS INCORPORATED
    Inventors: KEERTHINARAYAN P. HERAGU, THEODORE W. HOUSTON, ANAND SESHADRI, HUGH T. MAIR
  • Patent number: 7550993
    Abstract: Various embodiments of the present invention provide systems and methods for glitch reduced circuits. As one example, a glitch reduced, variable width driver circuit is disclosed. Such circuits include a data output, and at least two transistors that each includes a gate, a first leg and a second leg. The gate of the first transistor is electrically coupled to a first combined control signal, and the gate of the second transistor is electrically coupled to a second combined control signal. The first leg of the first transistor and the first leg of the second transistor are electrically coupled to a power source, and the second leg of the first transistor and the second leg of the second transistor are electrically coupled to an output signal. The circuits further include a control circuit that combines a first control signal with the data output to create the first combined control signal, and combines a second control signal with the data output to create the second combined control signal.
    Type: Grant
    Filed: September 21, 2007
    Date of Patent: June 23, 2009
    Assignee: Texas Instruments Incorporated
    Inventors: Keerthinarayan P. Heragu, Rajat Chauhan, Chintamani Keshav Bhaktavatson
  • Publication number: 20090051390
    Abstract: Various embodiments of the present invention provide systems and methods for glitch reduced circuits. As one example, a glitch reduced, variable width driver circuit is disclosed. Such circuits include a data output, and at least two transistors that each includes a gate, a first leg and a second leg. The gate of the first transistor is electrically coupled to a first combined control signal, and the gate of the second transistor is electrically coupled to a second combined control signal. The first leg of the first transistor and the first leg of the second transistor are electrically coupled to a power source, and the second leg of the first transistor and the second leg of the second transistor are electrically coupled to an output signal. The circuits further include a control circuit that combines a first control signal with the data output to create the first combined control signal, and combines a second control signal with the data output to create the second combined control signal.
    Type: Application
    Filed: September 21, 2007
    Publication date: February 26, 2009
    Inventors: Keerthinarayan P. Heragu, Rajat Chauhan, Chintamani Keshav Bhaktavatson
  • Publication number: 20090033388
    Abstract: Various systems and methods for signal synchronization are disclosed. For example, some embodiments of the present invention provide methods for reduced area delay signal timing. Such methods include providing a delay lock loop circuit with a plurality of selectable delay elements. The methods further include operating the delay lock loop circuit in a first mode where a program number is established in relation to a reference frequency. The program number corresponds to a number of the plurality of selectable delay elements used to establish a first delay time at the reference frequency. The program number is multiplied by a multiplicand, and the product of the multiplication is used while operating the delay lock loop circuit in a second mode to select the number of delay elements utilized in delaying an input signal. In the second mode, an input signal is delayed by a second delay time that is approximately the first delay time multiplied by the multiplicand.
    Type: Application
    Filed: August 1, 2007
    Publication date: February 5, 2009
    Inventors: Keerthinarayan P. Heragu, Padattil K. Nisha
  • Publication number: 20090033385
    Abstract: Various embodiments of the present invention provide delay lock loop circuits. Such delay lock loop circuits include two or more delay stages that each include a plurality of selectable delay elements. A reference signal drives an input of the first delay stage, and the first delay stage provides a first output. The first output drives an input of the second delay stage, and the second delay stage provides a second output. The circuits further include a first selector register that is associated with the first delay stage. A value maintained in the first selector register determines a number of the selectable delay elements utilized in the first delay stage. Modification of the value maintained in the first selector register is synchronized to the first output. The circuits further include a second selector register associated with the second delay stage. A value maintained in the second selector register determines a number of the selectable delay elements utilized in the second delay stage.
    Type: Application
    Filed: August 1, 2007
    Publication date: February 5, 2009
    Inventors: Keerthinarayan P. Heragu, Padattil K. Nisha
  • Publication number: 20090033386
    Abstract: Various systems and methods for delaying one signal in relation to another are disclosed. As one example, a delay lock loop circuit is discussed that includes at least a first delay stage and a second delay stage, each including a plurality of selectable delay elements. The delay stages are configured such that a gated reference signal drives an input of the first delay stage, and a first output from the first delay stage drives an input of the second delay stage. The circuits further include a unified selector register that is associated with both the first delay stage and the second delay stage. A value maintained in the unified selector register determines a number of the selectable delay elements utilized in the first delay stage and a number of the selectable delay elements utilized in the second delay stage. In operation, modification of the value maintained in the unified selector register is synchronized to the reference signal.
    Type: Application
    Filed: August 1, 2007
    Publication date: February 5, 2009
    Inventors: Padattil K. Nisha, Keerthinarayan P. Heragu
  • Publication number: 20090033387
    Abstract: Various systems and methods for delaying a signal relative to another signal are disclosed. As one example, a delay lock loop circuit is disclosed that includes at least two delay stages. Each of the aforementioned delay stages include a plurality of selectable delay elements. Such selectable delay elements may be, but are not limited to, a plurality of single input buffers, and a plurality of multiple input logic gates. Further, a first of the delay stages is selectably driven by one of a first signal and a reference signal, and the stage provides a first stage output. A second of the delay stages is selectably driven by one of a second signal and the first stage output, and the stage provides a second stage output. The circuit further includes a mode signal that has at least two states.
    Type: Application
    Filed: August 1, 2007
    Publication date: February 5, 2009
    Inventors: Keerthinarayan P. Heragu, Padattil K. Nisha
  • Publication number: 20080098244
    Abstract: Embodiments of the present disclosure provide a power controller, a method of operating a power controller and a semiconductor memory system. In one embodiment, the power controller is for use with a memory and includes an access module configured to provide an active state of the memory to allow memory access. The power controller also includes a retain-till-access module configured to cycle a portion of the memory between the active state and a low leakage data retention state of the memory. The power controller further includes an expanded retain-till-access module configured to extend the active state of the memory for a specified period of time before returning the memory to the low leakage data retention state.
    Type: Application
    Filed: October 22, 2007
    Publication date: April 24, 2008
    Applicant: Texas Instruments Incorporated
    Inventors: Michael P. Clinton, Keerthinarayan P. Heragu, Uming U. Ko
  • Patent number: 7216247
    Abstract: The disclosed invention provides methods and systems for writing and reading data in systems using multiple FIFO buffer elements. For each buffer element, a determination is made of when the rising edge of the read clock occurs during the second half of the write clock cycle. Responsive to this determination, the data written into the FIFO buffer element is shifted in order to reduce skew.
    Type: Grant
    Filed: August 5, 2004
    Date of Patent: May 8, 2007
    Assignee: Texas Instruments Incorporated
    Inventors: Keerthinarayan P. Heragu, Patrick Bosshart
  • Patent number: 6930953
    Abstract: A circuit takes a reference strobe signal as a first input, and a strobe ready signal generated from a memory that is strobed by the reference strobe signal as a second input. The circuit generates a self-timed strobe signal along with other control signals that are used to perform a subsequent memory access in a single clock cycle, thereby doubling memory access bandwidth.
    Type: Grant
    Filed: August 5, 2003
    Date of Patent: August 16, 2005
    Assignee: Texas Instruments Incorporated
    Inventors: Keerthinarayan P. Heragu, Mustafa Ulvi Erdogan, Robert James Landers
  • Publication number: 20040052153
    Abstract: A circuit takes a reference strobe signal as a first input, and a strobe ready signal generated from a memory that is strobed by the reference strobe signal as a second input. The circuit generates a self-timed strobe signal along with other control signals that are used to perform a subsequent memory access in a single clock cycle, thereby doubling memory access bandwidth.
    Type: Application
    Filed: August 5, 2003
    Publication date: March 18, 2004
    Inventors: Keerthinarayan P. Heragu, Mustafa Ulvi Erdogan, Robert James Landers
  • Patent number: 6553547
    Abstract: A method for generating charge sharing test vectors for a circuit generates a first test vector (120) and a second test vector (122). The method provides a test model (98) including a logic cell (10) and an auxiliary test circuit (100) where the auxiliary test circuit (100) includes a discharge AND gate (102) and a charge sharing AND gate (104). The method generates a first test vector (120) for the test model (98) having an input pattern to discharge nodes of the logic cell (10) and evaluate discharge AND gate (102) to a logic level 1. The method generates a second test vector (122) having an input pattern to evoke the worst charge sharing behavior for the logic cell (10) and evaluate charge sharing AND gate (104) to a logic level 1.
    Type: Grant
    Filed: November 15, 2000
    Date of Patent: April 22, 2003
    Assignee: Texas Instruments Incorporated
    Inventors: Keerthinarayan P. Heragu, Patrick W. Bosshart