Patents by Inventor Ramesh B. Gunna
Ramesh B. Gunna 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: 11947457Abstract: A scalable cache coherency protocol for system including a plurality of coherent agents coupled to one or more memory controllers is described. The memory controller may implement a precise directory for cache blocks from the memory to which the memory controller is coupled. Multiple requests to a cache block may be outstanding, and snoops and completions for requests may include an expected cache state at the receiving agent, as indicated by a directory in the memory controller when the request was processed, to allow the receiving agent to detect race conditions. In an embodiment, the cache states may include a primary shared and a secondary shared state. The primary shared state may apply to a coherent agent that bears responsibility for transmitting a copy of the cache block to a requesting agent. In an embodiment, at least two types of snoops may be supported: snoop forward and snoop back.Type: GrantFiled: November 22, 2022Date of Patent: April 2, 2024Assignee: Apple Inc.Inventors: James Vash, Gaurav Garg, Brian P. Lilly, Ramesh B. Gunna, Steven R. Hutsell, Lital Levy-Rubin, Per H. Hammarlund, Harshavardhan Kaushikkar
-
Patent number: 11934313Abstract: A system including a plurality of processor cores, a plurality of graphics processing units, a plurality of peripheral circuits, and a plurality of memory controllers is configured to support scaling of the system using a unified memory architecture.Type: GrantFiled: August 22, 2022Date of Patent: March 19, 2024Assignee: Apple Inc.Inventors: Per H. Hammarlund, Lior Zimet, James Vash, Gaurav Garg, Sergio Kolor, Harshavardhan Kaushikkar, Ramesh B. Gunna, Steven R. Hutsell
-
Patent number: 11868258Abstract: A scalable cache coherency protocol for system including a plurality of coherent agents coupled to one or more memory controllers is described. The memory controller may implement a precise directory for cache blocks from the memory to which the memory controller is coupled. Multiple requests to a cache block may be outstanding, and snoops and completions for requests may include an expected cache state at the receiving agent, as indicated by a directory in the memory controller when the request was processed, to allow the receiving agent to detect race conditions. In an embodiment, the cache states may include a primary shared and a secondary shared state. The primary shared state may apply to a coherent agent that bears responsibility for transmitting a copy of the cache block to a requesting agent. In an embodiment, at least two types of snoops may be supported: snoop forward and snoop back.Type: GrantFiled: January 27, 2023Date of Patent: January 9, 2024Assignee: Apple Inc.Inventors: James Vash, Gaurav Garg, Brian P. Lilly, Ramesh B. Gunna, Steven R. Hutsell, Lital Levy-Rubin, Per H. Hammarlund, Harshavardhan Kaushikkar
-
Patent number: 11675409Abstract: An apparatus includes an execute circuit configured to execute a plurality of operations received from a queue, as well as a power estimator circuit, and a power sensing circuit. The power estimator circuit is configured to predict power consumption due to execution of a particular operation of the plurality of operations, and to withdraw, based on the predicted power consumption, a first amount of power credits from a power credit pool. The power sensing circuit is configured to monitor one or more characteristics of a power supply node coupled to the execute circuit to generate a power value, and to deposit a second amount of power credits into the power credit pool. The second amount of power credits may be based on the power value indicating that power consumed during the execution of the particular operation is less than the predicted power consumption.Type: GrantFiled: July 12, 2022Date of Patent: June 13, 2023Assignee: Apple Inc.Inventors: Matthias Knoth, Srikanth Balasubramanian, Venkatram Krishnaswamy, Ramesh B. Gunna
-
Publication number: 20230169003Abstract: A scalable cache coherency protocol for system including a plurality of coherent agents coupled to one or more memory controllers is described. The memory controller may implement a precise directory for cache blocks from the memory to which the memory controller is coupled. Multiple requests to a cache block may be outstanding, and snoops and completions for requests may include an expected cache state at the receiving agent, as indicated by a directory in the memory controller when the request was processed, to allow the receiving agent to detect race conditions. In an embodiment, the cache states may include a primary shared and a secondary shared state. The primary shared state may apply to a coherent agent that bears responsibility for transmitting a copy of the cache block to a requesting agent. In an embodiment, at least two types of snoops may be supported: snoop forward and snoop back.Type: ApplicationFiled: January 27, 2023Publication date: June 1, 2023Inventors: James Vash, Gaurav Garg, Brian P. Lilly, Ramesh B. Gunna, Steven R. Hutsell, Lital Levy-Rubin, Per H. Hammarlund, Harshavardhan Kaushikkar
-
Publication number: 20230083397Abstract: A scalable cache coherency protocol for system including a plurality of coherent agents coupled to one or more memory controllers is described. The memory controller may implement a precise directory for cache blocks from the memory to which the memory controller is coupled. Multiple requests to a cache block may be outstanding, and snoops and completions for requests may include an expected cache state at the receiving agent, as indicated by a directory in the memory controller when the request was processed, to allow the receiving agent to detect race conditions. In an embodiment, the cache states may include a primary shared and a secondary shared state. The primary shared state may apply to a coherent agent that bears responsibility for transmitting a copy of the cache block to a requesting agent. In an embodiment, at least two types of snoops may be supported: snoop forward and snoop back.Type: ApplicationFiled: November 22, 2022Publication date: March 16, 2023Inventors: James Vash, Gaurav Garg, Brian P. Lilly, Ramesh B. Gunna, Steven R. Hutsell, Lital Levy-Rubin, Per H. Hammarlund, Harshavardhan Kaushikkar
-
Publication number: 20230056044Abstract: A system including a plurality of processor cores, a plurality of graphics processing units, a plurality of peripheral circuits, and a plurality of memory controllers is configured to support scaling of the system using a unified memory architecture.Type: ApplicationFiled: August 22, 2022Publication date: February 23, 2023Inventors: Per H. Hammarlund, Lior Zimet, James Vash, Gaurav Garg, Sergio Kolor, Harshavardhan Kaushikkar, Ramesh B. Gunna, Steven R. Hutsell
-
Patent number: 11544193Abstract: A scalable cache coherency protocol for system including a plurality of coherent agents coupled to one or more memory controllers is described. The memory controller may implement a precise directory for cache blocks from the memory to which the memory controller is coupled. Multiple requests to a cache block may be outstanding, and snoops and completions for requests may include an expected cache state at the receiving agent, as indicated by a directory in the memory controller when the request was processed, to allow the receiving agent to detect race conditions. In an embodiment, the cache states may include a primary shared and a secondary shared state. The primary shared state may apply to a coherent agent that bears responsibility for transmitting a copy of the cache block to a requesting agent. In an embodiment, at least two types of snoops may be supported: snoop forward and snoop back.Type: GrantFiled: May 10, 2021Date of Patent: January 3, 2023Assignee: Apple Inc.Inventors: James Vash, Gaurav Garg, Brian P. Lilly, Ramesh B. Gunna, Steven R. Hutsell, Lital Levy-Rubin, Per H. Hammarlund, Harshavardhan Kaushikkar
-
Publication number: 20220342471Abstract: An apparatus includes an execute circuit configured to execute a plurality of operations received from a queue, as well as a power estimator circuit, and a power sensing circuit. The power estimator circuit is configured to predict power consumption due to execution of a particular operation of the plurality of operations, and to withdraw, based on the predicted power consumption, a first amount of power credits from a power credit pool. The power sensing circuit is configured to monitor one or more characteristics of a power supply node coupled to the execute circuit to generate a power value, and to deposit a second amount of power credits into the power credit pool. The second amount of power credits may be based on the power value indicating that power consumed during the execution of the particular operation is less than the predicted power consumption.Type: ApplicationFiled: July 12, 2022Publication date: October 27, 2022Inventors: Matthias Knoth, Srikanth Balasubramanian, Venkatram Krishnaswamy, Ramesh B. Gunna
-
Patent number: 11416056Abstract: An apparatus includes an execute circuit configured to execute a plurality of operations received from a queue, as well as a power estimator circuit, and a power sensing circuit. The power estimator circuit is configured to predict power consumption due to execution of a particular operation of the plurality of operations, and to withdraw, based on the predicted power consumption, a first amount of power credits from a power credit pool. The power sensing circuit is configured to monitor one or more characteristics of a power supply node coupled to the execute circuit to generate a power value, and to deposit a second amount of power credits into the power credit pool. The second amount of power credits may be based on the power value indicating that power consumed during the execution of the particular operation is less than the predicted power consumption.Type: GrantFiled: September 18, 2020Date of Patent: August 16, 2022Assignee: Apple Inc.Inventors: Matthias Knoth, Srikanth Balasubramanian, Venkatram Krishnaswamy, Ramesh B. Gunna
-
Patent number: 11347198Abstract: Systems, apparatuses, and methods for implementing an optimized adaptive thermal control mechanism for an integrated circuit (IC) are described. A control unit receives a digital input value which is representative of a temperature of an IC. The control unit compares the input value to at least two set points. A result of a first comparison determines whether an accumulator is incremented or decremented by a programmable gain value. A result of a second comparison determines whether the accumulator is primed with a preset ramp-up value. The preset ramp-up value is used since the accumulator can take several sensing cycles to reach the optimal control value while thermal gradients can become critical in only a few cycles. The output of the accumulator is provided to an actuator which adjusts parameter(s) to modulate the IC's temperature. The granularity and range of the accumulator matches the granularity and range of the actuator.Type: GrantFiled: September 4, 2020Date of Patent: May 31, 2022Assignee: Apple Inc.Inventors: Matthias Knoth, Ramesh B. Gunna, Srikanth Balasubramanian
-
Publication number: 20220091649Abstract: An apparatus includes an execute circuit configured to execute a plurality of operations received from a queue, as well as a power estimator circuit, and a power sensing circuit. The power estimator circuit is configured to predict power consumption due to execution of a particular operation of the plurality of operations, and to withdraw, based on the predicted power consumption, a first amount of power credits from a power credit pool. The power sensing circuit is configured to monitor one or more characteristics of a power supply node coupled to the execute circuit to generate a power value, and to deposit a second amount of power credits into the power credit pool. The second amount of power credits may be based on the power value indicating that power consumed during the execution of the particular operation is less than the predicted power consumption.Type: ApplicationFiled: September 18, 2020Publication date: March 24, 2022Inventors: Matthias Knoth, Srikanth Balasubramanian, Venkatram Krishnaswamy, Ramesh B. Gunna
-
Publication number: 20220083472Abstract: A scalable cache coherency protocol for system including a plurality of coherent agents coupled to one or more memory controllers is described. The memory controller may implement a precise directory for cache blocks from the memory to which the memory controller is coupled. Multiple requests to a cache block may be outstanding, and snoops and completions for requests may include an expected cache state at the receiving agent, as indicated by a directory in the memory controller when the request was processed, to allow the receiving agent to detect race conditions. In an embodiment, the cache states may include a primary shared and a secondary shared state. The primary shared state may apply to a coherent agent that bears responsibility for transmitting a copy of the cache block to a requesting agent. In an embodiment, at least two types of snoops may be supported: snoop forward and snoop back.Type: ApplicationFiled: May 10, 2021Publication date: March 17, 2022Inventors: James Vash, Gaurav Garg, Brian P. Lilly, Ramesh B. Gunna, Steven R. Hutsell, Lital Levy-Rubin, Per H. Hammarlund
-
Publication number: 20220075343Abstract: Systems, apparatuses, and methods for implementing an optimized adaptive thermal control mechanism for an integrated circuit (IC) are described. A control unit receives a digital input value which is representative of a temperature of an IC. The control unit compares the input value to at least two set points. A result of a first comparison determines whether an accumulator is incremented or decremented by a programmable gain value. A result of a second comparison determines whether the accumulator is primed with a preset ramp-up value. The preset ramp-up value is used since the accumulator can take several sensing cycles to reach the optimal control value while thermal gradients can become critical in only a few cycles. The output of the accumulator is provided to an actuator which adjusts parameter(s) to modulate the IC's temperature. The granularity and range of the accumulator matches the granularity and range of the actuator.Type: ApplicationFiled: September 4, 2020Publication date: March 10, 2022Inventors: Matthias Knoth, Ramesh B. Gunna, Srikanth Balasubramanian
-
Patent number: 10901484Abstract: A processor includes a mechanism for disabling a memory array of a branch prediction unit. The processor may include a next fetch prediction unit that may include a number of entries. Each entry may correspond to a next instruction fetch group and may store an indication of whether or not the corresponding the next fetch group includes a conditional branch instruction. In response to an indication that the next fetch group does not include a conditional branch instruction, the fetch prediction unit may be configured to disable, in a next instruction execution cycle, the memory array of the branch prediction unit.Type: GrantFiled: March 25, 2019Date of Patent: January 26, 2021Assignee: Apple Inc.Inventors: Conrado Blasco, Ronald P. Hall, Ramesh B. Gunna, Ian D. Kountanis, Shyam Sundar, André Seznec
-
Publication number: 20190286218Abstract: A processor includes a mechanism for disabling a memory array of a branch prediction unit. The processor may include a next fetch prediction unit that may include a number of entries. Each entry may correspond to a next instruction fetch group and may store an indication of whether or not the corresponding the next fetch group includes a conditional branch instruction. In response to an indication that the next fetch group does not include a conditional branch instruction, the fetch prediction unit may be configured to disable, in a next instruction execution cycle, the memory array of the branch prediction unit.Type: ApplicationFiled: March 25, 2019Publication date: September 19, 2019Inventors: Conrado Blasco, Ronald P. Hall, Ramesh B. Gunna, Ian D. Kountanis, Shyam Sundar, André Seznec
-
Patent number: 10416692Abstract: A method and apparatus for reducing capacitor noise in electronic systems is disclosed. A system includes at least one functional circuit block coupled to receive a variable supply voltage. The value of the supply voltage is controlled by a power management circuit. Changing a performance state of the functional circuit block includes increasing the supply voltage for higher performance, and reducing the supply voltage for reduced performance demands. The power management circuit, in changing to a higher performance state, increases the supply voltage at a first rate. A rate control circuit causes the power management circuit to reduce the supply voltage, when changing to a lower performance state, at a second rate that is less than the first rate.Type: GrantFiled: September 19, 2017Date of Patent: September 17, 2019Assignee: Apple Inc.Inventors: Jong-Suk Lee, Ramesh B. Gunna, Shih-Chieh Wen
-
Patent number: 10410688Abstract: An IC in which a power state of a circuit in one power domain is managed based at least in part on a power state of a circuit in another power domain is disclosed. In one embodiment, an IC includes first and second functional circuit blocks in first and second power domains, respectively. A third functional block shared by the first and second is also implemented in the first power domain. A power management unit may control power states of each of the first, second, and third functional circuit blocks. The power management circuit may, when the first functional circuit block is in a sleep state, set a power state of the third functional block in accordance with that of the second functional circuit block.Type: GrantFiled: December 3, 2018Date of Patent: September 10, 2019Assignee: Apple Inc.Inventors: Shih-Chieh Wen, Jong-Suk Lee, Ramesh B. Gunna
-
Patent number: 10401938Abstract: Systems, apparatuses, and methods for reaching power targets across different clock domains are described. In various embodiments, a first processor complex and a second processor complex operate while powered by a same, single power plane, but with respective clock domains. When a request is detected to change an operating mode of a particular core from one of the processor complexes to an operating mode which does not provide the worst-case power supply load on the single power plane, an amount of voltage margin to recover from the operational voltage is determined based on the second operating mode prior to granting the request and based on each other core in the complexes operating in respective current operating modes. An operational voltage less the determined voltage margin to recover is assigned to the processor complexes while different clock frequencies are assigned to the processor complexes.Type: GrantFiled: April 10, 2017Date of Patent: September 3, 2019Assignee: Apple Inc.Inventors: Jong-Suk Lee, Ramesh B. Gunna, Shih-Chieh R. Wen, John H. Mylius
-
Publication number: 20190221241Abstract: An IC in which a power state of a circuit in one power domain is managed based at least in part on a power state of a circuit in another power domain is disclosed. In one embodiment, an IC includes first and second functional circuit blocks in first and second power domains, respectively. A third functional block shared by the first and second is also implemented in the first power domain. A power management unit may control power states of each of the first, second, and third functional circuit blocks. The power management circuit may, when the first functional circuit block is in a sleep state, set a power state of the third functional block in accordance with that of the second functional circuit block.Type: ApplicationFiled: December 3, 2018Publication date: July 18, 2019Inventors: Shih-Chieh Wen, Jong-Suk Lee, Ramesh B. Gunna