vcmp_xxx
Function
vcmp_[eq|ge|gt|le|lt|ne] instruction abstraction. The following six instructions have the same performance.
vcmp_eq: (x == y), where x is equal to y by element-wise comparison.
vcmp_ge: (x >= y), where x is greater than or equal to y by element-wise comparison.
vcmp_gt: (x > y), where x is greater than y by element-wise comparison.
vcmp_le: (x <= y), where x is less than or equal to y by element-wise comparison.
vcmp_lt: (x < y), where x is smaller than y by element-wise comparison.
vcmp_ne: (x != y), where x is not equal to y by element-wise comparison.
Prototype
1 | class vcmp(x, y) |
Parameters
Parameter |
Input/Output |
Data Type |
Description |
|---|---|---|---|
x |
Input |
Tensor variable |
Input x-vector tensor. FP16 and FP32 are supported. |
y |
Output |
Tensor variable |
Output y-vector tensor. FP16 and FP32 are supported. |
Constraints
The tensors of all input and output data of vector instructions are in the UB space, and their shapes must be the same.
Example
1 2 3 4 5 6 | from mskpp import vcmp, Tensor ub_x, ub_y = Tensor("UB"), Tensor("UB") gm_x, gm_y = Tensor("GM"), Tensor("GM") ub_x.load(gm_x) ub_y.load(gm_y) out = vcmp(ub_x, ub_y)() |