17884755. CONCURRENT COMPUTE CONTEXT simplified abstract (Intel Corporation)

From WikiPatents
Jump to navigation Jump to search

CONCURRENT COMPUTE CONTEXT

Organization Name

Intel Corporation

Inventor(s)

Joydeep Ray of Folsom CA (US)

Vasanth Ranganathan of El Dorado Hills CA (US)

James Valerio of North Plains OR (US)

Jeffery S. Boles of Folsom CA (US)

Hema Chand Nalluri of Bengaluru (IN)

Aditya Navale of Folsom CA (US)

Ben J. Ashbaugh of Folsom CA (US)

Michal Mrozek of Gdansk (PL)

Murali Ramadoss of Folsom CA (US)

Hong Jiang of Los Altos CA (US)

Ankur Shah of Folsom CA (US)

CONCURRENT COMPUTE CONTEXT - A simplified explanation of the abstract

This abstract first appeared for US patent application 17884755 titled 'CONCURRENT COMPUTE CONTEXT

Simplified Explanation

- System of concurrent compute queues for scheduling compute contexts on graphics processor hardware - Apparatus includes system interface and general-purpose graphics processor - Graphics processor has multiple hardware resources partitioned into isolated partitions - Each partition includes command streamers and circuitry for scheduling compute workloads - Workloads are submitted to command queues associated with command streamers

Potential Applications

- High-performance computing - Artificial intelligence and machine learning - Scientific simulations and modeling - Virtual reality and augmented reality applications

Problems Solved

- Efficient scheduling of compute workloads on graphics processor hardware - Simultaneous execution of a large number of compute contexts - Partitioning of hardware resources for isolation and improved performance

Benefits

- Increased performance and throughput for compute-intensive tasks - Scalability for handling a large number of compute contexts simultaneously - Flexibility in partitioning hardware resources for different workloads - Enhanced capabilities for graphics processing and general-purpose computing


Original Abstract Submitted

Embodiments described herein provide a system of concurrent compute queues that enable the scheduling of a large number of compute contexts simultaneously on graphics processor hardware. One embodiment provides an apparatus comprising a system interface and a general-purpose graphics processor coupled with the system interface. The general-purpose graphics processor comprises a plurality of graphics processor hardware resources configured to be partitioned into a plurality of isolated partitions, each of the plurality of isolated partitions including a first command streamer, a second command streamer, and circuitry configured to schedule general-purpose graphics compute workloads submitted to a first plurality of command queues associated with the first command streamer and a second plurality of command queues associated with the second command streamer.