Uber Technologies, Inc.

Advanced Micro Devices Corporation

Altera Corporation

Qi Chen

Qing Xu

Miao Yu

Linfeng Shi

Casey Lawler

David E Huang

Adam J Wright

Peter Boyle

Binh Vo

Daniel Parrenas Shimizu

Kaiyu Ren

John DiCosola

Laiq Chughtai

Chi-Shung David Wang

Wan-Pin Hung

Seng Yew Lim

A network system dynamically determines a route, including start and end points, for vehicles in a transportation network. The transportation network receives a service request from a user of the transportation network including an origin location for the trip and a destination location for the trip. The transportation network then generates a waypoint plan for one or more vehicles, which includes the requested origin and destination in addition to any previously requested origins and destinations included in the vehicles current route. The network system then determines a directionality for each of the waypoints in the waypoint plan and retrieves candidate start and end points that have an associated directionality within a threshold angle of the directionality of each waypoint and are proximate to the waypoint. The network system evaluates each combination of retrieved candidate points to select a route for the vehicle.

Dynamically determining origin and destination locations for a network system

Embodiments of a dynamic leakage control circuit for use with graphics processor circuitry are described. The dynamic leakage control circuit selectively enables back biasing of the transistors comprising the graphics processor circuits during particular modes of operation. The back biasing levels are controlled by two separate power rails. A first power rail is coupled to an existing power supply and the second power rail is coupled to a separate adjustable voltage regulator. A separate voltage regulator may also be provided for the first power rail. A hardware-based state machine or software process is programmed to detect the occurrence of one or more modes of operation and adjust the voltage regulators for the first and second power rails to either enable or disable the back biasing state of the circuit, or alter the threshold voltage of the circuit within a specified voltage range.

Dynamic leakage control using selective back-biasing

Techniques are provided for isolating failed routing resources on a programmable circuit. Failing test patterns and the test logs are fed to a Statistical Failure Isolation (SFI) tool. The SFI tool extracts failing paths from the test patterns. A statistical analysis is performed on interconnect resources related to failing paths. The resources on the paths are then tallied to create a histogram of resources. These resources are then be fed into an Adaptive Failure Isolation (AFI) tool to auto-generate verification patterns. A tester uses the verification patterns to isolate failed interconnect resources.

Functional failure analysis techniques for programmable integrated circuits

Growing community of inventors

Sunnyvale, CA, United States of America

Qi Chen