ia32-64/x86/vcvttph2udq.html
2025-07-08 02:23:29 -03:00

115 lines
5.6 KiB
HTML
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

<!DOCTYPE html>
<html xmlns="http://www.w3.org/1999/xhtml" xmlns:svg="http://www.w3.org/2000/svg" xmlns:x86="http://www.felixcloutier.com/x86"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8"><link rel="stylesheet" type="text/css" href="style.css"></link><title>VCVTTPH2UDQ
— Convert with Truncation Packed FP16 Values to Unsigned DoublewordIntegers</title></head><body><header><nav><ul><li><a href='index.html'>Index</a></li><li>December 2023</li></ul></nav></header><h1>VCVTTPH2UDQ
— Convert with Truncation Packed FP16 Values to Unsigned DoublewordIntegers</h1>
<table>
<tr>
<th> Instruction En Bit Mode Flag
Support Instruction En Bit Mode Flag
Support 64/32 CPUID Feature Instruction En Bit Mode Flag CPUID Feature Instruction En Bit Mode Flag Op/ 64/32 CPUID Feature Instruction En Bit Mode Flag 64/32 CPUID Feature Instruction En Bit Mode Flag CPUID Feature Instruction En Bit Mode Flag Op/ 64/32 CPUID Feature </th>
<th></th>
<th>Support</th>
<th></th>
<th>Description</th></tr>
<tr>
<td>EVEX.128.NP.MAP5.W0 78 /r VCVTTPH2UDQ xmm1{k1}{z}, xmm2/m64/m16bcst</td>
<td>A</td>
<td>V/V</td>
<td>AVX512-FP16 AVX512VL</td>
<td>Convert four packed FP16 values in xmm2/m64/m16bcst to four unsigned doubleword integers, and store the result in xmm1 using truncation subject to writemask k1.</td></tr>
<tr>
<td>EVEX.256.NP.MAP5.W0 78 /r VCVTTPH2UDQ ymm1{k1}{z}, xmm2/m128/m16bcst</td>
<td>A</td>
<td>V/V</td>
<td>AVX512-FP16 AVX512VL</td>
<td>Convert eight packed FP16 values in xmm2/m128/m16bcst to eight unsigned doubleword integers, and store the result in ymm1 using truncation subject to writemask k1.</td></tr>
<tr>
<td>EVEX.512.NP.MAP5.W0 78 /r VCVTTPH2UDQ zmm1{k1}{z}, ymm2/m256/m16bcst {sae}</td>
<td>A</td>
<td>V/V</td>
<td>AVX512-FP16</td>
<td>Convert sixteen packed FP16 values in ymm2/m256/m16bcst to sixteen unsigned doubleword integers, and store the result in zmm1 using truncation subject to writemask k1.</td></tr></table>
<h2 id="instruction-operand-encoding">Instruction Operand Encoding<a class="anchor" href="#instruction-operand-encoding">
</a></h2>
<table>
<tr>
<th>Op/En</th>
<th>Tuple</th>
<th>Operand 1</th>
<th>Operand 2</th>
<th>Operand 3</th>
<th>Operand 4</th></tr>
<tr>
<td>A</td>
<td>Half</td>
<td>ModRM:reg (w)</td>
<td>ModRM:r/m (r)</td>
<td>N/A</td>
<td>N/A</td></tr></table>
<h3 id="description">Description<a class="anchor" href="#description">
</a></h3>
<p>This instruction converts packed FP16 values in the source operand to unsigned doubleword integers in the destination operand.</p>
<p>When a conversion is inexact, a truncated (round toward zero) value is returned. If a converted result cannot be represented in the destination format, the floating-point invalid exception is raised, and if this exception is masked, the integer indefinite value is returned.</p>
<p>The destination elements are updated according to the writemask.</p>
<h3 id="operation">Operation<a class="anchor" href="#operation">
</a></h3>
<h4 id="vcvttph2udq-dest--src">VCVTTPH2UDQ dest, src<a class="anchor" href="#vcvttph2udq-dest--src">
</a></h4>
<pre>VL = 128, 256 or 512
KL := VL / 32
FOR j := 0 TO KL-1:
IF k1[j] OR *no writemask*:
IF *SRC is memory* and EVEX.b = 1:
tsrc := SRC.fp16[0]
ELSE
tsrc := SRC.fp16[j]
DEST.dword[j] := Convert_fp16_to_unsigned_integer32_truncate(tsrc)
ELSE IF *zeroing*:
DEST.dword[j] := 0
// else dest.dword[j] remains unchanged
DEST[MAXVL-1:VL] := 0
</pre>
<h3 id="intel-c-c++-compiler-intrinsic-equivalent">Intel C/C++ Compiler Intrinsic Equivalent<a class="anchor" href="#intel-c-c++-compiler-intrinsic-equivalent">
</a></h3>
<pre>VCVTTPH2UDQ __m512i _mm512_cvtt_roundph_epu32 (__m256h a, int sae);
</pre>
<pre>VCVTTPH2UDQ __m512i _mm512_mask_cvtt_roundph_epu32 (__m512i src, __mmask16 k, __m256h a, int sae);
</pre>
<pre>VCVTTPH2UDQ __m512i _mm512_maskz_cvtt_roundph_epu32 (__mmask16 k, __m256h a, int sae);
</pre>
<pre>VCVTTPH2UDQ __m128i _mm_cvttph_epu32 (__m128h a);
</pre>
<pre>VCVTTPH2UDQ __m128i _mm_mask_cvttph_epu32 (__m128i src, __mmask8 k, __m128h a);
</pre>
<pre>VCVTTPH2UDQ __m128i _mm_maskz_cvttph_epu32 (__mmask8 k, __m128h a);
</pre>
<pre>VCVTTPH2UDQ __m256i _mm256_cvttph_epu32 (__m128h a);
</pre>
<pre>VCVTTPH2UDQ __m256i _mm256_mask_cvttph_epu32 (__m256i src, __mmask8 k, __m128h a);
</pre>
<pre>VCVTTPH2UDQ __m256i _mm256_maskz_cvttph_epu32 (__mmask8 k, __m128h a);
</pre>
<pre>VCVTTPH2UDQ __m512i _mm512_cvttph_epu32 (__m256h a);
</pre>
<pre>VCVTTPH2UDQ __m512i _mm512_mask_cvttph_epu32 (__m512i src, __mmask16 k, __m256h a);
</pre>
<pre>VCVTTPH2UDQ __m512i _mm512_maskz_cvttph_epu32 (__mmask16 k, __m256h a);
</pre>
<h3 class="exceptions" id="simd-floating-point-exceptions">SIMD Floating-Point Exceptions<a class="anchor" href="#simd-floating-point-exceptions">
</a></h3>
<p>Invalid, Precision.</p>
<h3 class="exceptions" id="other-exceptions">Other Exceptions<a class="anchor" href="#other-exceptions">
</a></h3>
<p>EVEX-encoded instructions, see <span class="not-imported">Table 2-46</span>, “Type E2 Class Exception Conditions.”</p><footer><p>
This UNOFFICIAL, mechanically-separated, non-verified reference is provided for convenience, but it may be
inc<span style="opacity: 0.2">omp</span>lete or b<sub>r</sub>oke<sub>n</sub> in various obvious or non-obvious
ways. Refer to <a href="https://software.intel.com/en-us/download/intel-64-and-ia-32-architectures-sdm-combined-volumes-1-2a-2b-2c-2d-3a-3b-3c-3d-and-4">Intel® 64 and IA-32 Architectures Software Developers Manual</a> for anything serious.
</p></footer></body></html>