What devices don't support double precision?

6 views
Skip to first unread message

Aaron Hsu

unread,
Nov 18, 2025, 10:16:47 AMNov 18
to ArrayFire Users
Is there a published list of which devices don't support double precision floats? I'm seeing this error lately when I didn't see this as much in the past, and I'd like to know in advance whether a device is going to have support for this without running any code. I know about the hasDouble...() functionality, but is there anything where I can see the whole list? 


John Melonakos

unread,
Dec 9, 2025, 4:51:36 PM (12 days ago) Dec 9
to ArrayFire Users
Which backend are you using? Which hardware? I'm not aware of a place to find that info at a glance, other than searching each manufacturer's website.
-John

Aaron Hsu

unread,
Dec 10, 2025, 12:31:53 AM (11 days ago) Dec 10
to ArrayFire Users
I am mostly seeing this on Intel devices trying to run OpenCL or OneAPI backends, that is, backends that don't involve CUDA but do involve OpenCL or OneAPI. 
--
You received this message because you are subscribed to the Google Groups "ArrayFire Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to arrayfire-use...@googlegroups.com.

Reply all
Reply to author
Forward
0 new messages