Please read this before reporting a bug:
https://wiki.archlinux.org/title/Bug_reporting_guidelines
Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.
REPEAT: Do NOT report bugs for outdated packages!
https://wiki.archlinux.org/title/Bug_reporting_guidelines
Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.
REPEAT: Do NOT report bugs for outdated packages!
FS#53995 - tensorflow should support more cuda capacity
Attached to Project:
Community Packages
Opened by Yichao Zhou (zhou13) - Tuesday, 09 May 2017, 22:10 GMT
Last edited by Sven-Hendrik Haase (Svenstaro) - Wednesday, 10 May 2017, 08:14 GMT
Opened by Yichao Zhou (zhou13) - Tuesday, 09 May 2017, 22:10 GMT
Last edited by Sven-Hendrik Haase (Svenstaro) - Wednesday, 10 May 2017, 08:14 GMT
|
DetailsMy GPU only support cuda capacity 3.0, which is not included in current tensorflow build.
|
This task depends upon
Closed by Sven-Hendrik Haase (Svenstaro)
Wednesday, 10 May 2017, 08:14 GMT
Reason for closing: Fixed
Wednesday, 10 May 2017, 08:14 GMT
Reason for closing: Fixed
Comment by Yichao Zhou (zhou13) -
Tuesday, 09 May 2017, 22:13 GMT
cuda capacity -> cuda compute capability
Comment by Yichao Zhou (zhou13) -
Tuesday, 09 May 2017, 22:15 GMT
From https://github.com/tensorflow/tensorflow/issues/25 , it seems that 3.0 is the minimum supported compute capability.