[ARROW-1055] [C++] GPU support library development

• Type: New Feature
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.12.0
• Component/s: C++
• External issue URL: https://github.com/apache/arrow/issues/16647
• Assignee: Wes McKinney
• Reporter: Wes McKinney
• Created: 19/May/17 18:30
• Updated: 11/Jan/23 07:12
• Resolved: 06/Feb/19 03:47

Description

Umbrella issue to cover GPU buffer memory management, IPC loading, mixed-device schema/record batch management, etc.

Issue Links

• incorporates: ARROW-1364 [C++] IPC reader and writer specialized for GPU device memory
• incorporates: ARROW-1387 [C++] Set up GPU leaf library build toolchain
• incorporates: ARROW-1392 [C++] Implement reader and writer IO interfaces for GPU buffers
• incorporates: ARROW-1394 [Plasma] Add optional extension for allocating memory on GPUs
• incorporates: ARROW-1393 [C++] Simplified CUDA IPC writer and reader for communicating a CPU + GPU payload to another process
• incorporates: ARROW-1423 [C++] Create non-owned CudaContext from context handle provided by thirdparty user
• incorporates: ARROW-2447 [C++] Create a device abstraction
• incorporates: ARROW-3653 [Python/C++] Support data copying between different GPU devices
• incorporates: ARROW-4486 [Python][CUDA] pyarrow.cuda.Context.foreign_buffer should have a `base=None` argument
• is related to: ARROW-1385 [C++] Add Buffer implementation and helper functions for POSIX shared memory