VANDNPS

VANDNPS inverts its first source and ANDs the result with the second: vandnps ymm0, ymm1, ymm2 computes NOT ymm1 AND ymm2. It is the VEX and EVEX form of ANDNPS and belongs to the family that VANDPS describes: bits only, no MXCSR, and an EVEX form that needs AVX512DQ. The inverted operand is the one that is easy to get wrong. In a case recorded on Zen 5, a first source of 00000001H and a second of FFFFFFFFH gave FFFFFFFEH; swapped, the result would have been 0. The intrinsic _mm256_andnot_ps(a, b) keeps the order, inverting a.

Clearing what a mask selects#

With a compare result as the first source, VANDNPS zeroes the elements where the compare held and keeps the others. GCC 14 at -O2 with -march=x86-64-v3 compiles a[i] = b[i] < c[i] ? 0.0f : d[i] to VCMPLTPS and vandnps ymm0, ymm0, ymm1. Clang 23 reverses the compare instead, to VCMPNLTPS, and loads d[i] under it with VMASKMOVPS, which zeroes the other elements by itself. Both give d[i] when b[i] or c[i] is a NaN, as the C expression does. With -0.0 as the first source, VANDNPS clears the sign bits: GCC uses it that way for the fabsf in -fabsf(x) (see VORPS).

A zeroing idiom#

VANDNPS of a register with itself gives zero whatever the register held, as vandnps xmm24, xmm3, xmm3 did on Zen 5 with the EVEX encoding. AMD’s Zen 5 guide lists that form among the zeroing idioms, next to VXORPS, which clear the destination without waiting for the source and take no cycles (Section 2.9.2).

Forms

Latency and throughput

Data flow

Where the elements of the sources go, element 0 on the right; results that depend on a control value are drawn for an example. How to read.

Operation

What the forms do, in pseudocode that our tests run against results recorded on real processors. How to read.

for i in range(VL // 32):
    if k1[i]:  # 1 without a write mask
        zmm1.d[i] = ~zmm2.d[i] & src3.d[i]
    elif zeroing:
        zmm1.d[i] = 0
    # else: zmm1.d[i] keeps its value
# m32bcst: src3.d[i] is the m32 for every i
# zmm1 is zeroed from bit VL up

Operands

Flags

None: XED records no flag that the forms test or change.

Exceptions

The exception class is XED's name for the set of exception conditions that a form follows. How to read.

VEX forms

avx-type-4

EVEX forms

AVX512-E4

Attributes

What else XED records about the forms. How to read.

EVEX forms

write masking

EVEX memory forms

embedded broadcast (tuple Full); memory fault suppression

Intrinsics

The C intrinsics of Clang 23.1.2 that compile to VANDNPS, each with the form it compiles to and the Clang options it needs. The forms come from compiling every intrinsic and decoding the result with Intel XED; all intrinsics A–Z explains the method.

CPUID and processors

Each form belongs to an XED ISA set. A form can be used when the processor reports every CPUID bit of one of its ISA set's alternatives.

More AVX instructions

• VADDPD
• VADDPS
• VADDSD
• VADDSS
• VADDSUBPD
• VADDSUBPS
• VANDNPD
• VANDPD
• VANDPS
• VBLENDPD
• VBLENDPS
• VBLENDVPD
• VBLENDVPS
• VBROADCASTF128
• VBROADCASTSD
• VBROADCASTSS
• VCMPPD
• VCMPPS
• VCMPSD
• VCMPSS
• VCOMISD
• VCOMISS
• VCVTDQ2PD
• VCVTDQ2PS
• VCVTPD2DQ
• VCVTPD2PS
• VCVTPS2DQ
• VCVTPS2PD
• VCVTSD2SI
• VCVTSD2SS
• VCVTSI2SD
• VCVTSI2SS
• VCVTSS2SD
• VCVTSS2SI
• VCVTTPD2DQ
• VCVTTPS2DQ
• VCVTTSD2SI
• VCVTTSS2SI
• VDIVPD
• VDIVPS

All 253 AVX instructions

More AVX10.1 instructions

• KXNORD
• KXNORQ
• KXNORW
• KXORB
• KXORD
• KXORQ
• KXORW
• VADDPD
• VADDPH
• VADDPS
• VADDSD
• VADDSH
• VADDSS
• VAESDEC
• VAESDECLAST
• VAESENC
• VAESENCLAST
• VALIGND
• VALIGNQ
• VANDNPD
• VANDPD
• VANDPS
• VBLENDMPD
• VBLENDMPS
• VBROADCASTF32X2
• VBROADCASTF32X4
• VBROADCASTF32X8
• VBROADCASTF64X2
• VBROADCASTF64X4
• VBROADCASTI32X2
• VBROADCASTI32X4
• VBROADCASTI32X8
• VBROADCASTI64X2
• VBROADCASTI64X4
• VBROADCASTSD
• VBROADCASTSS
• VCMPPD
• VCMPPH
• VCMPPS
• VCMPSD

All 686 AVX10.1 instructions