We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent d1d3e17 commit eb18b6fCopy full SHA for eb18b6f
models/B200TCRN.m
@@ -59,7 +59,7 @@
59
end
60
61
elseif ismember(informat, {'tf32', 'tensorfloat32'})
62
- def_params.fma=4;
+ def_params.fma=8;
63
elseif ismember(informat, {'fp8-e5m2','fp8-e4m3','e5m2','e4m3'})
64
% FMA size is 16, but interleaved pattern is used to join two
65
% 16-element vectors.
0 commit comments