_mm512_rsqrt_phADD TO COMPARE ADDED TO COMPARE
Intel 64-bit (64 bits)/ AVX512
View official documentation
View Félix Cloutier's documentation
Purpose:
Compute the approximate reciprocal square root of packed half-precision (16-bit) floating-point elements in "a", and store the results in "dst". The maximum relative error for this approximation is less than 1.5*2^-12.
Result:
__m512h
Example:
#include <immintrin.h>
#include <stdio.h>
int main() {
__m512h a = _mm512_setr_ph(1.0, 4.0, 9.0, 16.0, 25.0, 36.0, 49.0, 64.0, 81.0, 100.0, 121.0, 144.0, 169.0, 196.0, 225.0, 256.0, 1.0, 4.0, 9.0, 16.0, 25.0, 36.0, 49.0, 64.0, 81.0, 100.0, 121.0, 144.0, 169.0, 196.0, 225.0, 256.0);
__m512h result = _mm512_rsqrt_ph(a);
_Float16* res = (_Float16*)&result;
for (int i = 0; i < 32; i++)
printf("%f\n", (float)res[i]);
}Prototypes
Assembly Instruction:
VRSQRTPH
Usage:
__m512h result =
_mm512_rsqrt_ph(
__m512h a
)
Performance Metrics:
📊 Unlock Performance Insights
Get access to detailed performance metrics including latency, throughput, and CPU-specific benchmarks for this intrinsic.
SIMD Intrinsics Summary
| SIMD Engines: | 6 |
| C Intrinsics: | 10444 |
| NEON: | 4353 |
| AVX2: | 405 |
| AVX512: | 4717 |
| SSE4.2: | 598 |
| VSX: | 192 |
| IBM-Z: | 179 |