Patents by Inventor Dave Roberts

Dave Roberts 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: 12259822
    Abstract: A method of GPU virtualization comprises allocating each virtual machine (or operating system running on a VM) an identifier by the hypervisor and then this identifier is used to tag every transaction deriving from a GPU workload operating within a given VM context (i.e. every GPU transaction on the system bus which interconnects the CPU, GPU and other peripherals). Additionally, dedicated portions of a memory resource (which may be GPU registers or RAM) are provided for each VM and whilst each VM can only see their allocated portion of the memory, a microprocessor within the GPU can see all of the memory. Access control is achieved using root memory management units which are configured by the hypervisor and which map guest physical addresses to actual memory addresses based on the identifier associated with the transaction.
    Type: Grant
    Filed: April 27, 2021
    Date of Patent: March 25, 2025
    Assignee: Imagination Technologies Limited
    Inventors: Dave Roberts, Mario Sopena Novales, John W. Howson
  • Patent number: 11853796
    Abstract: The operation of a GPU is controlled based on one or more deadlines by which one or more GPU tasks must be completed and estimates of the time required to complete the execution of a first GPU task (which is currently being executed) and the time required to execute one or more other GPU tasks (which are not currently being executed). Based on a comparison between the deadline(s) and the estimates, the operating parameters of the GPU may be changed.
    Type: Grant
    Filed: December 9, 2016
    Date of Patent: December 26, 2023
    Assignee: Imagination Technologies Limited
    Inventors: Dave Roberts, Jackson Dsouza
  • Patent number: 11847492
    Abstract: The operation of a GPU is controlled based on one or more deadlines by which one or more GPU tasks must be completed and estimates of the time required to complete the execution of a first GPU task (which is currently being executed) and the time required to execute one or more other GPU tasks (which are not currently being executed). Based on a comparison between the deadline(s) and the estimates, context switching may or may not be triggered.
    Type: Grant
    Filed: December 9, 2016
    Date of Patent: December 19, 2023
    Assignee: Imagination Technologies Limited
    Inventors: Dave Roberts, Jackson Dsouza
  • Publication number: 20230227296
    Abstract: A material handling lift including safety features for reducing the risk of injury to a user. The features may include one or more of: a roller shade for covering components of a mast, a scanner for detecting people in a protected area of the lift, a fan for preventing accumulation of dust on the scanner, spacers between a lift platform and a side guard panel, perforated side guard panels to allow viewing of a space under the lift, entrance guards to keep people away from the side and front of the lift, a power unit with dual redundant locking valves, and tamper-resistant fasteners for fastening a rear panel to a mast.
    Type: Application
    Filed: January 20, 2023
    Publication date: July 20, 2023
    Inventors: Kim Decato, Randall Dickinson, Dave Roberts
  • Publication number: 20210263857
    Abstract: A method of GPU virtualization comprises allocating each virtual machine (or operating system running on a VM) an identifier by the hypervisor and then this identifier is used to tag every transaction deriving from a GPU workload operating within a given VM context (i.e. every GPU transaction on the system bus which interconnects the CPU, GPU and other peripherals). Additionally, dedicated portions of a memory resource (which may be GPU registers or RAM) are provided for each VM and whilst each VM can only see their allocated portion of the memory, a microprocessor within the GPU can see all of the memory. Access control is achieved using root memory management units which are configured by the hypervisor and which map guest physical addresses to actual memory addresses based on the identifier associated with the transaction.
    Type: Application
    Filed: April 27, 2021
    Publication date: August 26, 2021
    Inventors: Dave Roberts, Mario Sopena Novales, John W. Howson
  • Patent number: 11016906
    Abstract: A method of GPU virtualization comprises allocating each virtual machine (or operating system running on a VM) an identifier by the hypervisor and then this identifier is used to tag every transaction deriving from a GPU workload operating within a given VM context (i.e. every GPU transaction on the system bus which interconnects the CPU, GPU and other peripherals). Additionally, dedicated portions of a memory resource (which may be GPU registers or RAM) are provided for each VM and whilst each VM can only see their allocated portion of the memory, a microprocessor within the GPU can see all of the memory. Access control is achieved using root memory management units which are configured by the hypervisor and which map guest physical addresses to actual memory addresses based on the identifier associated with the transaction.
    Type: Grant
    Filed: September 7, 2020
    Date of Patent: May 25, 2021
    Assignee: Imagination Technologies Limited
    Inventors: Dave Roberts, Mario Sopena Novales, John W. Howson
  • Publication number: 20200409863
    Abstract: A method of GPU virtualization comprises allocating each virtual machine (or operating system running on a VM) an identifier by the hypervisor and then this identifier is used to tag every transaction deriving from a GPU workload operating within a given VM context (i.e. every GPU transaction on the system bus which interconnects the CPU, GPU and other peripherals). Additionally, dedicated portions of a memory resource (which may be GPU registers or RAM) are provided for each VM and whilst each VM can only see their allocated portion of the memory, a microprocessor within the GPU can see all of the memory. Access control is achieved using root memory management units which are configured by the hypervisor and which map guest physical addresses to actual memory addresses based on the identifier associated with the transaction.
    Type: Application
    Filed: September 7, 2020
    Publication date: December 31, 2020
    Inventors: Dave Roberts, Mario Sopena Novales, John W. Howson
  • Publication number: 20200408306
    Abstract: A fugitive emissions packing set is provided. The packing set a top ring and a bottom ring made from a fluoropolymer that includes a filler that enhances the mechanical and/or dimensional stability of the ring, as compared to unfilled fluoropolymer. The intermediate rings are formed from an essentially pure fluoropolymer. The top, bottom, or intermediate rings may include a metal insert in any combination.
    Type: Application
    Filed: June 18, 2020
    Publication date: December 31, 2020
    Inventors: Wayne Evans, Dave Roberts, Brett Yoder, Joseph Young
  • Patent number: 10802985
    Abstract: A method of GPU virtualization comprises allocating each virtual machine (or operating system running on a VM) an identifier by the hypervisor and then this identifier is used to tag every transaction deriving from a GPU workload operating within a given VM context (i.e. every GPU transaction on the system bus which interconnects the CPU, GPU and other peripherals). Additionally, dedicated portions of a memory resource (which may be GPU registers or RAM) are provided for each VM and whilst each VM can only see their allocated portion of the memory, a microprocessor within the GPU can see all of the memory. Access control is achieved using root memory management units which are configured by the hypervisor and which map guest physical addresses to actual memory addresses based on the identifier associated with the transaction.
    Type: Grant
    Filed: June 14, 2019
    Date of Patent: October 13, 2020
    Assignee: Imagination Technologies Limited
    Inventors: Dave Roberts, Mario Sopena Novales, John W. Howson
  • Patent number: 10639190
    Abstract: A temperature-regulating sports wrap is provided, including a first web of fabric having elongate side edges and first and second end edges, a second web of fabric is secured to the first web of fabric. At least one pocket is defined between opposing surfaces of the first and second webs and opens along one of the elongate side edges. At least one first strip extends outwardly from the first side edges along an axis parallel to the elongate side edges, has one of a hook-and-loop fastener material. At least one second strip extends inwardly from the second side edges, is secured to a surface of one of the first and second webs. The at least one second strip has the opposite one of a hook-and-loop fastener material from the at least one first strip, the at least one second strip extends inwardly at a non-parallel angle from the second edge.
    Type: Grant
    Filed: March 6, 2017
    Date of Patent: May 5, 2020
    Assignee: MONTEREY BAY ASSOCIATES
    Inventor: Dave Roberts
  • Publication number: 20190294556
    Abstract: A method of GPU virtualization comprises allocating each virtual machine (or operating system running on a VM) an identifier by the hypervisor and then this identifier is used to tag every transaction deriving from a GPU workload operating within a given VM context (i.e. every GPU transaction on the system bus which interconnects the CPU, GPU and other peripherals). Additionally, dedicated portions of a memory resource (which may be GPU registers or RAM) are provided for each VM and whilst each VM can only see their allocated portion of the memory, a microprocessor within the GPU can see all of the memory. Access control is achieved using root memory management units which are configured by the hypervisor and which map guest physical addresses to actual memory addresses based on the identifier associated with the transaction.
    Type: Application
    Filed: June 14, 2019
    Publication date: September 26, 2019
    Inventors: Dave Roberts, Mario Sopena Novales, John W. Howson
  • Patent number: 10366012
    Abstract: A method of GPU virtualization comprises allocating each virtual machine (or operating system running on a VM) an identifier by the hypervisor and then this identifier is used to tag every transaction deriving from a GPU workload operating within a given VM context (i.e. every GPU transaction on the system bus which interconnects the CPU, GPU and other peripherals). Additionally, dedicated portions of a memory resource (which may be GPU registers or RAM) are provided for each VM and while each VM can only see their allocated portion of the memory, a microprocessor within the GPU can see all of the memory. Access control is achieved using root memory management units which are configured by the hypervisor and which map guest physical addresses to actual memory addresses based on the identifier associated with the transaction.
    Type: Grant
    Filed: December 2, 2016
    Date of Patent: July 30, 2019
    Assignee: Imagination Technologies Limited
    Inventors: Dave Roberts, Mario Sopena Novales, John W. Howson
  • Patent number: 9779402
    Abstract: Methods and devices are provided for use in detecting relay attacks between devices in a communications network. One method includes sending first data by a first device to a second device, and receiving, by the first device, a communication from the second device where the communication comprises second data generated at the second device and a time parameter related to the generation of the second data. The method also includes measuring a total transmission time at the first device between sending the first data and receiving the communication, and determining a further time parameter related to the generation of the second data based at least in part on the measured total transmission time. The method then further includes determining the presence of a relay attack between the first and second devices in dependence on a comparison of the time parameter and the further time parameter.
    Type: Grant
    Filed: September 14, 2016
    Date of Patent: October 3, 2017
    Assignee: MASTERCARD INTERNATIONAL INCORPORATED
    Inventors: Duncan Garrett, Dave Roberts, Patrik Smets
  • Publication number: 20170161204
    Abstract: A method of GPU virtualization comprises allocating each virtual machine (or operating system running on a VM) an identifier by the hypervisor and then this identifier is used to tag every transaction deriving from a GPU workload operating within a given VM context (i.e. every GPU transaction on the system bus which interconnects the CPU, GPU and other peripherals). Additionally, dedicated portions of a memory resource (which may be GPU registers or RAM) are provided for each VM and whilst each VM can only see their allocated portion of the memory, a microprocessor within the GPU can see all of the memory. Access control is achieved using root memory management units which are configured by the hypervisor and which map guest physical addresses to actual memory addresses based on the identifier associated with the transaction.
    Type: Application
    Filed: December 2, 2016
    Publication date: June 8, 2017
    Inventors: Dave Roberts, Mario Sopena Novales, John W. Howson
  • Publication number: 20170006048
    Abstract: Methods and devices are provided for use in detecting relay attacks between devices in a communications network. One method includes sending first data by a first device to a second device, and receiving, by the first device, a communication from the second device where the communication comprises second data generated at the second device and a time parameter related to the generation of the second data. The method also includes measuring a total transmission time at the first device between sending the first data and receiving the communication, and determining a further time parameter related to the generation of the second data based at least in part on the measured total transmission time. The method then further includes determining the presence of a relay attack between the first and second devices in dependence on a comparison of the time parameter and the further time parameter.
    Type: Application
    Filed: September 14, 2016
    Publication date: January 5, 2017
    Inventors: Duncan Garrett, Dave Roberts, Patrik Smets
  • Patent number: 9485092
    Abstract: A transaction device for establishing a shared secret with a point of interaction (POI) over a communications network to enable encrypted communications between the transaction device and the point of interaction, the device comprising: an input arranged to receive communications from the point of interaction; a processor arranged to generate a first communication according to a Diffie-Hellman protocol; an output arranged to send the first communication to the point of interaction; wherein the processor is arranged to apply a randomly generated blinding factor, r, when generating the first communication and wherein, in response to receiving a second communication from the point of interaction at the input, the second communication having been generated according to the Diffie-Hellman protocol, the processor is arranged to apply the randomly generated blinding factor and generate a shared secret according to the Diffie-Hellman protocol in dependence on data contained within the second communication.
    Type: Grant
    Filed: June 6, 2014
    Date of Patent: November 1, 2016
    Assignee: MASTERCARD INTERNATIONAL INCORPORATED
    Inventors: Patrik Smets, Patrick Mestre, Dave Roberts, Duncan Garret
  • Publication number: 20160080151
    Abstract: A system and method of authenticating a communication network comprising a first computing device, a second computing device and an intermediary computing device, wherein there is a first path between the first computing device and the intermediary computing device and a second path between the second computing device and the intermediary computing device. The method is executed at the intermediary computing device, and comprises receiving, from the first computing device, a first session key generated by the first computing device using a function, wherein an input to the function comprises an incremented variable; receiving, from the second computing device, data associated with a second session key generated by the second computing device using the function; determining that the first session key and the second session key are the same; and defining the communication network as authentic when the first session key and the second session key are the same.
    Type: Application
    Filed: September 10, 2015
    Publication date: March 17, 2016
    Inventors: Patrik Smets, Patrick Mestré, Dave Roberts, Duncan Garrett
  • Publication number: 20140365776
    Abstract: A transaction device for establishing a shared secret with a point of interaction (POI) over a communications network to enable encrypted communications between the transaction device and the point of interaction, the device comprising: an input arranged to receive communications from the point of interaction; a processor arranged to generate a first communication according to a Diffie-Hellman protocol; an output arranged to send the first communication to the point of interaction; wherein the processor is arranged to apply a randomly generated blinding factor, r, when generating the first communication and wherein, in response to receiving a second communication from the point of interaction at the input, the second communication having been generated according to the Diffie-Hellman protocol, the processor is arranged to apply the randomly generated blinding factor and generate a shared secret according to the Diffie-Hellman protocol in dependence on data contained within the second communication.
    Type: Application
    Filed: June 6, 2014
    Publication date: December 11, 2014
    Inventors: Patrik Smets, Patrick Mestré, Dave Roberts, Duncan Garrett
  • Patent number: 8909557
    Abstract: A system and method for generating an authentication token which is used by an issuer associated with a integrated circuit card to authenticate a transaction. A personal card reader receives data, including an authentication cryptogram, from the integrated circuit card. The personal card reader uses the data received from the integrated circuit card to select one of at least two default bitmaps stored in a memory portion of the personal card reader. The personal card reader uses the selected default bitmap and the authentication cryptogram to build the authentication token.
    Type: Grant
    Filed: September 9, 2009
    Date of Patent: December 9, 2014
    Assignee: MasterCard International Incorporated
    Inventors: Mehdi Collinge, Jean-Paul Edmond Rans, Paul Vanneste, John Beric, Dave Roberts, Thomas Tan, Fikret Ates, Jonathan Grossar
  • Patent number: 8864557
    Abstract: A multi-bird poultry shackle includes a stanchion having a connection point adapted to couple to a conveyor; a first shackle coupled to the stanchion, the first shackle having a first pair of leg loops adapted to hold the legs of a first bird, and a second pair of leg loops adapted to hold the legs of a second bird; a second shackle coupled to the stanchion below the first shackle, the second shackle having a third pair of leg loops adapted to hold the legs of a third bird, and a fourth pair of leg loops adapted to hold the legs of a fourth bird; and a third shackle coupled to the stanchion below the second shackle, the third shackle having a fifth pair of leg loops adapted to hold the legs of a fifth bird, and a sixth pair of leg loops adapted to hold the legs of a sixth bird. A method of conveying poultry carcasses is also described.
    Type: Grant
    Filed: February 7, 2014
    Date of Patent: October 21, 2014
    Assignee: Perdue Foods LLC
    Inventors: Dave Roberts, Lonnie Bailey