Patents Assigned to NVIDIA Corp.
  • DETECTION AND MITIGATION OF UNSTABLE CELLS IN UNCLONABLE CELL ARRAY
    Publication number: 20220271951
    Abstract: A circuit includes a set of multiple bit generating cells. One or more adjustable current sources is coupled to introduce perturbations into outputs of the bit generating cells. Based on the perturbations, the outputs of a subset less than all of the bit generating cells are selected, and applied as a control.
    Type: Application
    Filed: February 24, 2021
    Publication date: August 25, 2022
    Applicant: NVIDIA Corp.
    Inventors: Sudhir Shrikantha Kudva, Nikola Nedovic, Carl Thomas Gray
  • MACHINE LEARNING TRAINING IN LOGARITHMIC NUMBER SYSTEM
    Publication number: 20220261650
    Abstract: An end-to-end low-precision training system based on a multi-base logarithmic number system and a multiplicative weight update algorithm. The multi-base logarithmic number system is applied to update weights of the neural network, with different bases of the multi-base logarithmic number system utilized between calculation of weight updates, calculation of feed-forward signals, and calculation of feedback signals. The LNS expresses a high dynamic range and computational energy efficiency, making it advantageous for on-board training in energy-constrained edge devices.
    Type: Application
    Filed: June 11, 2021
    Publication date: August 18, 2022
    Applicant: NVIDIA Corp.
    Inventors: Jiawei Zhao, Steve Haihang Dai, Rangharajan Venkatesan, Ming-Yu Liu, William James Dally, Anima Anandkumar
  • ERROR DETECTION PIN ENCODING SCHEME TO AVOID MAXIMUM TRANSITIONS AND FURTHER IMPROVE SIGNAL INTEGRITY ON HIGH SPEED GRAPHIC MEMORY INTERFACES
    Publication number: 20220262447
    Abstract: First symbols are generated on a plurality of data channels by applying PAM-N encoding on a first subset of bits of a data burst, the first symbols generated without maximum transitions; second symbols are generated on at least one optionally-activated additional data channel, the second symbols generated by applying the PAM-N encoding on a second subset of bits of the data burst, the second symbols generated without maximum transitions; and third symbols are generated on a channel for communicating error correction bits for the first bits and second bits, the third symbols generated by applying hybrid PAM-N encoding on the error correction bits and a third subset of bits of the data burst, the hybrid PAM-N encoding comprising an interleaving of symbols with N voltage levels and symbols with less than N voltage levels.
    Type: Application
    Filed: February 11, 2022
    Publication date: August 18, 2022
    Applicant: NVIDIA Corp.
    Inventors: Sunil Sudhakaran, Gautam Bhatia, Robert Bloemer
  • Detection and mitigation of unstable cells in unclonable cell array
    Patent number: 11411563
    Abstract: A circuit includes a set of multiple bit generating cells. One or more adjustable current sources is coupled to introduce perturbations into outputs of the bit generating cells. Based on the perturbations, the outputs of a subset less than all of the bit generating cells are selected, and applied as a control.
    Type: Grant
    Filed: February 24, 2021
    Date of Patent: August 9, 2022
    Assignee: NVIDIA Corp.
    Inventors: Sudhir Shrikantha Kudva, Nikola Nedovic, Carl Thomas Gray
  • System and methods for hardware-software cooperative pipeline error detection
    Patent number: 11409597
    Abstract: An error reporting system utilizes a parity checker to receive data results from execution of an original instruction and a parity bit for the data. A decoder receives an error correcting code (ECC) for data resulting from execution of a shadow instruction of the original instruction, and data error correction is initiated on the original instruction result on condition of a mismatch between the parity bit and the original instruction result, and the decoder asserting a correctable error in the original instruction result.
    Type: Grant
    Filed: March 6, 2020
    Date of Patent: August 9, 2022
    Assignee: NVIDIA Corp.
    Inventors: Michael Sullivan, Siva Hari, Brian Zimmer, Timothy Tsai, Stephen W. Keckler
  • Tensor-based driving scenario characterization
    Patent number: 11390301
    Abstract: Techniques to characterize driving scenarios for autonomous vehicles characterize a path in a driving scenario according to metrics such as narrowness and effort. The scenarios may be characterized using a tree-based or tensor-based approach.
    Type: Grant
    Filed: June 10, 2020
    Date of Patent: July 19, 2022
    Assignee: NVIDIA Corp.
    Inventors: Siva Kumar Sastry Hari, Iuri Frosio, Zahra Ghodsi, Anima Anandkumar, Timothy Tsai, Stephen W. Keckler
  • Scalable light-weight protocols for wire-speed packet ordering
    Patent number: 11363339
    Abstract: A communication method between a source device and a target device utilizes speculative connection setup between the source device and the target device, target-device-side packet ordering, and fine-grained ordering to remove packet dependencies.
    Type: Grant
    Filed: July 20, 2020
    Date of Patent: June 14, 2022
    Assignee: NVIDIA Corp.
    Inventors: Hans Eberle, Larry Robert Dennison
  • TECHNIQUES TO IMPROVE CURRENT REGULATOR CAPABILITY TO PROTECT THE SECURED CIRCUIT FROM POWER SIDE CHANNEL ATTACK
    Publication number: 20220149728
    Abstract: This disclosure relates to current flattening circuits for an electrical load. The current flattening circuits incorporate randomize various parameters to add noise onto the supply current. This added noise may act to reduce the signal to noise ratio in the supply current, increasing the difficulty of identifying a computational artifact signal from power rail noise.
    Type: Application
    Filed: January 20, 2022
    Publication date: May 12, 2022
    Applicant: NVIDIA Corp.
    Inventors: Sudhir Shrikantha Kudva, Nikola Nedovic, Sanquan Song
  • Full adder cell with improved power efficiency
    Patent number: 11294631
    Abstract: An adder circuit that includes an operand input and a second operand input to an XNOR cell. The XNOR cell is configured to provide the operand input and the second operand input to both a NAND gate and a first OAI cell. A second OAI cell transforms the output of the XNOR cell into a carry out signal.
    Type: Grant
    Filed: October 23, 2019
    Date of Patent: April 5, 2022
    Assignee: NVIDIA Corp.
    Inventors: Ilyas Elkin, Ge Yang, Xi Zhang
  • SCALABLE LIGHT-WEIGHT PROTOCOLS FOR WIRE-SPEED PACKET ORDERING
    Publication number: 20220095017
    Abstract: A communication method between a source device and a target device utilizes speculative connection setup between the source device and the target device, target-device-side packet ordering, and fine-grained ordering to remove packet dependencies.
    Type: Application
    Filed: December 1, 2021
    Publication date: March 24, 2022
    Applicant: NVIDIA Corp.
    Inventors: Hans Eberle, Larry Robert Dennison
  • Techniques to improve current regulator capability to protect the secured circuit from power side channel attack
    Patent number: 11283349
    Abstract: This disclosure relates to current flattening circuits for an electrical load. The current flattening circuits incorporate randomize various parameters to add noise onto the supply current. This added noise may act to reduce the signal to noise ratio in the supply current, increasing the difficulty of identifying a computational artifact signal from power rail noise.
    Type: Grant
    Filed: April 23, 2020
    Date of Patent: March 22, 2022
    Assignee: NVIDIA Corp.
    Inventors: Sudhir Shrikantha Kudva, Nikola Nedovic, Sanquan Song
  • Efficient Neural Network Accelerator Dataflows
    Publication number: 20220076110
    Abstract: A distributed deep neural net (DNN) utilizing a distributed, tile-based architecture includes multiple chips, each with a central processing element, a global memory buffer, and a plurality of additional processing elements. Each additional processing element includes a weight buffer, an activation buffer, and vector multiply-accumulate units to combine, in parallel, the weight values and the activation values using stationary data flows.
    Type: Application
    Filed: November 19, 2021
    Publication date: March 10, 2022
    Applicant: NVIDIA Corp.
    Inventors: Yakun Shao, Rangharajan Venkatesan, Miaorong Wang, Daniel Smith, William James Dally, Joel Emer, Stephen W. Keckler, Brucek Khailany
  • Efficient neural network accelerator dataflows
    Patent number: 11270197
    Abstract: A distributed deep neural net (DNN) utilizing a distributed, tile-based architecture includes multiple chips, each with a central processing element, a global memory buffer, and a plurality of additional processing elements. Each additional processing element includes a weight buffer, an activation buffer, and vector multiply-accumulate units to combine, in parallel, the weight values and the activation values using stationary data flows.
    Type: Grant
    Filed: November 4, 2019
    Date of Patent: March 8, 2022
    Assignee: NVIDIA Corp.
    Inventors: Yakun Shao, Rangharajan Venkatesan, Miaorong Wang, Daniel Smith, William James Dally, Joel Emer, Stephen W. Keckler, Brucek Khailany
  • EFFICIENT SOFTMAX COMPUTATION
    Publication number: 20220067513
    Abstract: Solutions improving efficiency of Softmax computation applied for efficient deep learning inference in transformers and other neural networks. The solutions utilize a reduced-precision implementation of various operations in Softmax, replacing ex with 2x to reduce instruction overhead associated with computing ex, and replacing floating point max computation with integer max computation. Further described is a scalable implementation that decomposes Softmax into UnNormalized Softmax and Normalization operations.
    Type: Application
    Filed: December 4, 2020
    Publication date: March 3, 2022
    Applicant: NVIDIA Corp.
    Inventors: Jacob Robert Stevens, Rangharajan Venkatesan, Steve Haihang Dai, Brucek Khailany
  • FOVEATION AND SPATIAL HASHING IN LAYER-BASED COMPUTER-GENERATED HOLOGRAMS
    Publication number: 20220026715
    Abstract: The computational scaling challenges of holographic displays are mitigated by techniques for generating holograms that introduce foveation into a wave front recording planes approach to hologram generation. Spatial hashing is applied to organize the points or polygons of a display object into keys and values.
    Type: Application
    Filed: July 23, 2020
    Publication date: January 27, 2022
    Applicant: NVIDIA Corp.
    Inventors: Jui-Hsien Wang, Ward Lopes, Rachel Anastasia Brown, Peter Shirley
  • TECHNIQUES FOR DIVERGENT THREAD GROUP EXECUTION SCHEDULING
    Publication number: 20220027194
    Abstract: Warp sharding techniques to switch execution between divergent shards on instructions that trigger a long stall, thereby interleaving execution between diverged threads within a warp instead of across warps. The technique may be applied to mitigate pipeline stalls in applications with low warp occupancy and high divergence. Warp data cache locality may also be improved by concentrating memory accesses within a warp rather than spreading them across warps.
    Type: Application
    Filed: February 24, 2021
    Publication date: January 27, 2022
    Applicant: NVIDIA Corp.
    Inventors: Sana Damani, Mark Stephenson, Ram Rangan, Daniel Robert Johnson, Rishkul Kulkarni
  • STANDARD CELL LAYOUT GENERATION WITH APPLIED ARTIFICIAL INTELLIGENCE
    Publication number: 20220027546
    Abstract: A genetic algorithm is utilized to generate routing candidates to which a reinforcement learning model is applied to correct the design rule constraint violations incrementally. A design rule checker provides feedback on the violations to the reinforcement learning model and the model learns how to fix the violations. A layout device placer based upon a simulated annealing method may also be utilized.
    Type: Application
    Filed: April 14, 2021
    Publication date: January 27, 2022
    Applicant: NVIDIA Corp.
    Inventors: Haoxing Ren, Matthew Rudolph Fojtik
  • Clock data recovery convergence using signed timing injection
    Patent number: 11212073
    Abstract: A system for data and clock recovery includes a timing error detector, a phase detector, and a phase increment injector. The phase increment injector may be used to determine an increment to affect an output of the phase detector or a clocking element. A sign of the increment is determined from a sign or direction of an accumulated version of a clock and data recovery gradient value.
    Type: Grant
    Filed: February 26, 2020
    Date of Patent: December 28, 2021
    Assignee: NVIDIA Corp.
    Inventors: Pervez Mirza Aziz, Vishnu Balan, Viswanath Annampedu
  • ADVERSARIAL SCENARIOS FOR SAFETY TESTING OF AUTONOMOUS VEHICLES
    Publication number: 20210389769
    Abstract: Techniques to generate driving scenarios for autonomous vehicles characterize a path in a driving scenario according to metrics such as narrowness and effort. Nodes of the path are assigned a time for action to avoid collision from the node. The generated scenarios may be simulated in a computer.
    Type: Application
    Filed: June 10, 2020
    Publication date: December 16, 2021
    Applicant: NVIDIA Corp.
    Inventors: Siva Kumar Sastry Hari, Iuri Frosio, Zahra Ghodsi, Anima Anandkumar, Timothy Tsai, Stephen W. Keckler, Alejandro Troccoli
  • TENSOR-BASED DRIVING SCENARIO CHARACTERIZATION
    Publication number: 20210387643
    Abstract: Techniques to characterize driving scenarios for autonomous vehicles characterize a path in a driving scenario according to metrics such as narrowness and effort. The scenarios may be characterized using a tree-based or tensor-based approach.
    Type: Application
    Filed: June 10, 2020
    Publication date: December 16, 2021
    Applicant: NVIDIA Corp.
    Inventors: Siva Kumar Sastry Hari, Iuri Frosio, Zahra Ghodsi, Anima Anandkumar, Timothy Tsai, Stephen W. Keckler