Searched refs:vadds (Results 1 – 18 of 18) sorted by relevance
117 tik_instance.vadds(16, t_transpose_local_ub[cc7 * 64 + cc8 * 16],137 tik_instance.vadds(16, t_transpose_local_ub[cc7 * 64 + cc8 * 16],147 tik_instance.vadds(16, t_transpose_local_ub2[cc72 * 64 + cc82 * 16],166 tik_instance.vadds(16, t_transpose_local_ub[cc7 * 256 + cc8 * 16],186 tik_instance.vadds(16, t_transpose_local_ub[cc7 * 128 + cc8 * 16],206 tik_instance.vadds(16, t_transpose_local_ub[cc7 * 128 + cc8 * 16],225 tik_instance.vadds(16, t_transpose_local_ub[cc7 * 512 + cc8 * 16],245 tik_instance.vadds(16, t_transpose_local_ub[cc7 * 256 + cc8 * 16],255 tik_instance.vadds(16, t_transpose_local_ub2[cc7 * 256 + cc8 * 16],271 tik_instance.vadds(16, t_transpose_local_ub[cc7 * 256 + cc8 * 16],[all …]
73 batch_std = te.lang.cce.vsqrt(te.lang.cce.vadds(batch_var_biased, epsilon))90 y = te.lang.cce.vadds(x_input, 0.0)91 running_mean = te.lang.cce.vadds(mean, 0.0)92 running_std = te.lang.cce.vsqrt(te.lang.cce.vadds(variance, epsilon))
71 nudge_zp = te.lang.cce.floor(te.lang.cce.vadds(nudge_zp_, 0.5))83 nudge_input = te.lang.cce.floor(te.lang.cce.vadds(nudge_input_, 0.5))
70 nudge_zp = te.lang.cce.floor(te.lang.cce.vadds(nudge_zp_, 0.5))83 nudge_input = te.lang.cce.floor(te.lang.cce.vadds(nudge_input_, 0.5))
61 min_value_tensor = te.lang.cce.vadds(data_zero, min_value)101 nudge_zp = te.lang.cce.floor(te.lang.cce.vadds(nudge_zp_, 0.5))
60 min_value_tensor = te.lang.cce.vadds(data_zero, min_value)97 nudge_zp = te.lang.cce.floor(te.lang.cce.vadds(nudge_zp_, 0.5))
72 …tik_instance.vadds(64, t_1_0_local_ub[vec_i * 64], input_1_local_ub[vec_i * 64], 0, 64, 1, 1, 16, …134 tik_instance.vadds(64, t_1_0_local_ub[vec_i * 64], input_1_local_ub[vec_i * 64], 0,
67 … nudge_input = te.lang.cce.floor(te.lang.cce.vadds(te.lang.cce.vmul(input_x, quant_max_data), 0.5))
109 … nudge_input = te.lang.cce.floor(te.lang.cce.vadds(te.lang.cce.vmul(input_x, quant_max_data), 0.5))
110 … nudge_input = te.lang.cce.floor(te.lang.cce.vadds(te.lang.cce.vmul(input_x, quant_max_data), 0.5))
371 tik_inst.vadds(64, mat_ub_lg_lb_subs[0], mat_ub_lg_lb[0],
26 sum3 = te.lang.cce.vadds(sum2, tvm.const(const_bias, dtype=input1.dtype))
330 void vadds(const Operand *OpSd, const Operand *OpSn, const Operand *OpSm,
675 Asm->vadds(Dest, getSrc(0), getSrc(1), CondARM32::AL); in emitIAS()
2370 void AssemblerARM32::vadds(const Operand *OpSd, const Operand *OpSn, in vadds() function in Ice::ARM32::AssemblerARM32
660 void vadds(SRegister sd, SRegister sn, SRegister sm, Condition cond = AL);
926 void Assembler::vadds(SRegister sd, SRegister sn, SRegister sm,