Lines Matching full:precision

35 	  that supports only single-precision operations with 16 double-word
46 	  that supports half- and single-precision operations with 16
58 	  that supports half- and single-precision operations (including fused
69 	  that supports single- and double-precision operations with 16
81 	  that supports half-, single- and double-precision operations with 16
94 	  that supports half-, single- and double-precision operations
108 	  that supports half-, single-, double-precision operations (including
122 	  that supports half-, single- and double-precision operations
136 	  that supports half-, single-, double-precision operations (including
149 	  half-precision operations (half-precision extension).
155 	  single-precision operations.
162 	  double-precision operations.
180 	  double-precision (32 single-precision) floating-point registers.
186 	  double-precision (64 single-precision) floating-point registers.