CVTSS2SI—Convert Scalar Single-Precision Floating-Point Value to Doubleword Integer

INSTRUCTION SET REFERENCE, A-L

3-262 Vol. 2A

CVTSS2SI—Convert Scalar Single-Precision Floating-Point Value to Doubleword Integer

Instruction Operand Encoding

Description

Converts a single-precision floating-point value in the source operand (the second operand) to a signed doubleword 
integer (or signed quadword integer if operand size is 64 bits) in the destination operand (the first operand). The 
source operand can be an XMM register or a memory location. The destination operand is a general-purpose 
register. When the source operand is an XMM register, the single-precision floating-point value is contained in the 
low doubleword of the register.
When a conversion is inexact, the value returned is rounded according to the rounding control bits in the MXCSR 
register or the embedded rounding control bits. 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 indefinite integer value 
(2

w-1

, where w represents the number of bits in the destination format) is returned.

Legacy SSE instructions: In 64-bit mode, Use of the REX.W prefix promotes the instruction to produce 64-bit data. 
See the summary chart at the beginning of this section for encoding data and limits.
VEX.W1 and EVEX.W1 versions: promotes the instruction to produce 64-bit data in 64-bit mode.
Note: VEX.vvvv and EVEX.vvvv are reserved and must be 1111b, otherwise instructions will #UD.
Software should ensure VCVTSS2SI is encoded with VEX.L=0. Encoding VCVTSS2SI with VEX.L=1 may encounter 
unpredictable behavior across different processor generations.

Opcode/

Instruction

Op / 

En

64/32 

bit Mode 

Support

CPUID 

Feature 

Flag

Description

F3 0F 2D /r

CVTSS2SI r32, xmm1/m32

RM

V/V

SSE

Convert one single-precision floating-point value from 

xmm1/m32 to one signed doubleword integer in r32.

F3 REX.W 0F 2D /r

CVTSS2SI r64, xmm1/m32

RM

V/N.E.

SSE

Convert one single-precision floating-point value from 

xmm1/m32 to one signed quadword integer in r64.

VEX.128.F3.0F.W0 2D /r

VCVTSS2SI r32, xmm1/m32

RM

V/V

AVX

Convert one single-precision floating-point value from 

xmm1/m32 to one signed doubleword integer in r32.

VEX.128.F3.0F.W1 2D /r

VCVTSS2SI r64, xmm1/m32

RM

V/N.E.

1

AVX

Convert one single-precision floating-point value from 

xmm1/m32 to one signed quadword integer in r64.

EVEX.LIG.F3.0F.W0 2D /r

VCVTSS2SI r32, xmm1/m32{er}

T1F

V/V

AVX512F

Convert one single-precision floating-point value from 

xmm1/m32 to one signed doubleword integer in r32.

EVEX.LIG.F3.0F.W1 2D /r

VCVTSS2SI r64, xmm1/m32{er}

T1F

V/N.E.

1

NOTES:

1. VEX.W1/EVEX.W1 in non-64 bit is ignored; the instructions behaves as if the W0 version is used.

AVX512F

Convert one single-precision floating-point value from 

xmm1/m32 to one signed quadword integer in r64.

Op/En

Operand 1

Operand 2

Operand 3

Operand 4

RM

ModRM:reg (w)

ModRM:r/m (r)

NA

NA

T1F

ModRM:reg (w)

ModRM:r/m (r)

NA

NA