[Libre-soc-bugs] [Bug 230] Video opcode development and discussion

bugzilla-daemon at libre-soc.org bugzilla-daemon at libre-soc.org
Mon Jan 18 15:22:28 GMT 2021


--- Comment #68 from Luke Kenneth Casson Leighton <lkcl at lkcl.net> ---
i added a summary at the top of the wiki page, fascinatingly it is pretty

also i noted that FP estimates can be done by changing the meaning of "single
precision" v3.0B to be "half of bitwidth precision".  then when elwidth=32
using single^H^H^H^H^H half precision operations these are effectively FP
"estimates" because FP16 accuracy operations are stored in FP32 format.

You are receiving this mail because:
You are on the CC list for the bug.

More information about the libre-soc-bugs mailing list