VPCMPEQQ

VPCMPEQQ compares quadwords for equality: the VEX form writes all ones or zero per quadword into a vector, as SSE4.1’s PCMPEQQ does, whose page covers the extension it belongs to and the code compilers build without it, and the EVEX form one bit per quadword into a mask register, as VPCMPEQD’s page describes. VEX.128 needs AVX and VEX.256 AVX2; the EVEX forms need AVX-512F (with AVX512VL below 512 bits) or AVX10.1, take a write mask, and, unlike the byte and word compares, a second source of one quadword broadcast from memory. On Zen 5:

• {evex} vpcmpeqq k1, ymm24, ymm3 set K1 to 0Ah: elements 1 and 3 were equal, and the bits above the four results were cleared.
• vpcmpeqq k2{k7}, zmm5, zmm5, a register against itself, copied the low byte of K7, E3h, into K2: the write mask clears the bits it leaves out rather than keeping them.
• vpcmpeqq ymm12, ymm6, ymm6 set all of YMM12, which AMD’s Zen 5 guide lists among the ones idioms that do not wait for the old value (Section 2.9.2).

In our builds#

In our builds at -O2 over fixed-size int64_t arrays with -march=x86-64-v3, b[i] == c[i] ? -1 : 0 is one VPCMPEQQ per vector in GCC 14 and Clang 23, and b[i] == c[i] ? 0 : b[i] adds a VPANDN. Counting matches, n += b[i] == k, subtracts each result with VPSUBQ: with 64-bit counters the -1 of a match cannot wrap as it would in a word counter.

With -march=x86-64-v4, GCC compares zmm registers into K1 and expands the mask with a zero-masked VMOVDQA64 of all ones, or of b[i] for the select, which it tests as VPCMPNEQQ, an alias of VPCMPQ. Clang keeps the VEX form on ymm registers for the plain result and the count, and uses VPCMPNEQQ with a masked move for the select.

AMD’s latency spreadsheet for Zen 5 gives the VEX row 2 cycles, two per cycle on pipes FP0 and FP3, as for PCMPEQQ, and the EVEX rows, which write a mask, 3, 4 and 6 cycles for xmm, ymm and zmm sources, two per cycle.

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 // 64):
    a = ymm2.q[i]
    b = src3.q[i]
    ymm1.q[i] = -1 if a == b else 0
# ymm1 is zeroed from bit VL up

for i in range(VL // 64):
    k1[i] = k2[i] and zmm2.q[i] == src3.q[i]
# m64bcst: src3.q[i] is the m64 for every i
# the other bits of k1 are cleared

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 VPCMPEQQ, 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

• VPACKUSDW
• VPACKUSWB
• VPADDB
• VPADDD
• VPADDQ
• VPADDSB
• VPADDSW
• VPADDUSB
• VPADDUSW
• VPADDW
• VPALIGNR
• VPAND
• VPANDN
• VPAVGB
• VPAVGW
• VPBLENDVB
• VPBLENDW
• VPCLMULQDQ
• VPCMPEQB
• VPCMPEQD
• VPCMPEQW
• VPCMPESTRI
• VPCMPESTRM
• VPCMPGTB
• VPCMPGTD
• VPCMPGTQ
• VPCMPGTW
• VPCMPISTRI
• VPCMPISTRM
• VPERM2F128
• VPERMILPD
• VPERMILPS
• VPEXTRB
• VPEXTRD
• VPEXTRQ
• VPEXTRW
• VPHADDD
• VPHADDSW
• VPHADDW
• VPHMINPOSUW

All 253 AVX instructions

More AVX2 instructions

• VPADDQ
• VPADDSB
• VPADDSW
• VPADDUSB
• VPADDUSW
• VPADDW
• VPALIGNR
• VPAND
• VPANDN
• VPAVGB
• VPAVGW
• VPBLENDD
• VPBLENDVB
• VPBLENDW
• VPBROADCASTB
• VPBROADCASTD
• VPBROADCASTQ
• VPBROADCASTW
• VPCMPEQB
• VPCMPEQD
• VPCMPEQW
• VPCMPGTB
• VPCMPGTD
• VPCMPGTQ
• VPCMPGTW
• VPERM2I128
• VPERMD
• VPERMPD
• VPERMPS
• VPERMQ
• VPGATHERDD
• VPGATHERDQ
• VPGATHERQD
• VPGATHERQQ
• VPHADDD
• VPHADDSW
• VPHADDW
• VPHSUBD
• VPHSUBSW
• VPHSUBW

All 138 AVX2 instructions

More AVX10.1 instructions

• VPANDND
• VPANDNQ
• VPANDQ
• VPAVGB
• VPAVGW
• VPBLENDMB
• VPBLENDMD
• VPBLENDMQ
• VPBLENDMW
• VPBROADCASTB
• VPBROADCASTD
• VPBROADCASTMB2Q
• VPBROADCASTMW2D
• VPBROADCASTQ
• VPBROADCASTW
• VPCLMULQDQ
• VPCMPB
• VPCMPD
• VPCMPEQB
• VPCMPEQD
• VPCMPEQW
• VPCMPGTB
• VPCMPGTD
• VPCMPGTQ
• VPCMPGTW
• VPCMPQ
• VPCMPUB
• VPCMPUD
• VPCMPUQ
• VPCMPUW
• VPCMPW
• VPCOMPRESSB
• VPCOMPRESSD
• VPCOMPRESSQ
• VPCOMPRESSW
• VPCONFLICTD
• VPCONFLICTQ
• VPDPBUSD
• VPDPBUSDS
• VPDPWSSD

All 686 AVX10.1 instructions