Nvidia Corporation

Padmanabham Patki

Jue Wu

Chung-Hong Lai

Michael Alan Ditty

Marc R Delvaux

Smbat Tonoyan

Nisha Bhushan

Arpit Gupta

Kiran Kumar Dash

Andrew Elias

Laurent Dahan

Chiang Hsu

In various examples, a diagnostic circuit is connected to a target system to automatically trigger the target system to enter a diagnostic mode. The diagnostic circuit receives diagnostic data from the target system when the target system performs a diagnostic operation in the diagnostic mode.

Performing diagnostic operations on a target system

Systems, computer program products, and methods are described herein for automated data retrieval from an integrated circuit (IC). An example system receives an alert indicating a trigger event associated with the IC; extracts, using a scan island (e.g., a partition of the IC that is isolated for data retrieval), data from a plurality of scan chains and a plurality of random-access memories (RAMs) associated with the IC in response to receiving the alert; stores the data in an external non-volatile storage media; and reboots the IC upon storing the data in the external non-volatile storage media. In this way, embodiments of the present invention offer a scalable and secure method for real-time data extraction and processing in the event of an integrated circuit malfunction, improving diagnostics while ensuring cost-effectiveness and data security.

System for automated data retrieval from an integrated circuit for event analysis

In various examples, methods may include receiving first data transmitted from a second component and second data transmitted from the second component during a first time period. The first data may be transmitted via a first data lane and the second data may be transmitted via a second data lane. The method may include receiving de-skew symbols at an interval from the second component via the first data lane and via the second data lane during the first time period. The method may include compensating for a first skew introduced to a first propagation of the first data across the first data lane and a second skew introduced to a second propagation of the second data across the second data lane using the de-skew symbols received during the first time period.

Dynamic skew realignment over multiple transmission lanes in integrated computing platforms

Computing devices are now used for various purposes ranging from monitoring a refrigerator to driving automobiles. Protecting the data and logic within the chips of the computing devices is essential to ensure reliable operation. When a particular partition of a chip is powered-up but the logic of the partition is not reset, the logic will be in an unpredictable random state. To operate in a secure environment, it is necessary to start the operation of the logic from a known state and not a random state. To ensure the logic is operating in a secure environment, a digital reset detector circuit (DRDC) is provided that indicates if the logic was reset after power-up. The DRDC can ensure chips are secure from attacks involving reset deprivation upon power-up and help protect various secure and secret assets in a chip, including customer keys.

Protecting circuits from hacking using a digital reset detector

Growing community of inventors

Santa Clara, CA, United States of America

Padmanabham Patki