Ticket #3 (new defect)

Opened 4 years ago

Last modified 4 years ago

Unhelpful error when using Double on pre-1.3 CUDA devices

Reported by: chak Owned by:
Priority: minor Milestone:
Component: CUDA backend Version: 0.8.1.0
Keywords: Cc:

Description

Only CUDA devices with compute capability 1.3 and up contain hardware support for Double. Currently, Accelerate will nevertheless generate code for Doubles for these devices, which leads to a failure further down the pipeline. This is not very user-friendly.

This raises the general questions of how to handle device capabilities elegantly.

Change History

Changed 4 years ago by chak

  • owner chak deleted
  • component changed from Accelerate language to CUDA backend

Changed 4 years ago by chak

  • version changed from 0.7.1.0 to 0.8.0.0

Changed 4 years ago by chak

  • version changed from 0.8.0.0 to 0.8.1.0
Note: See TracTickets for help on using tickets.