VPOPCNTW

VPOPCNTW counts the set bits of each word of a vector and writes each count, 0 to 16, to the word in the same place: 8 words in an XMM register, 16 in a YMM and 32 in a ZMM one, under a write mask. It is the word form of VPOPCNTB, with which it came in AVX512_BITALG, and whose page explains the nibble-table sequence it replaces and the use of the counts; VPOPCNTD and VPOPCNTQ belong to AVX512_VPOPCNTDQ instead. Neither extension is part of x86-64-v4, while -march=znver5 enables both. A word count is POPCNT of a 16-bit value, element by element, without the flags.

Recorded on Zen 5#

• The edges. {evex} vpopcntw xmm17, xmm30 counted 16 for FFFFh, 15 for 7FFFh and 1 for both 0001h and 8000h.
• Zeroing. In vpopcntw xmm5{k3}{z}, xmmword ptr [...], with mask bits 2 and 4 clear, those words became 0, while the others got 15, 1, 0, 1, 1 and 1 for 7FFFh, 8000h, 0, 0010h, 0008h and 8000h.

In compiled code#

At -O2 over uint16_t arrays, a loop storing __builtin_popcount(a[i]) becomes VPOPCNTW straight from memory with -march=znver5, on ZMM registers in GCC 14 and Clang 23. With -march=x86-64-v4, which lacks BITALG, Clang 23 builds the byte counts with VPSHUFB table lookups and adds each word’s two bytes with a shift and VPADDB, and GCC 14 keeps the scalar POPCNT. _mm512_popcnt_epi16 is one VPOPCNTW, and needs BITALG enabled in both compilers.

AMD’s Zen 5 latency spreadsheet gives one macro-op on pipe FP0 or FP3 and 2 cycles, two per cycle, as for VPOPCNTB.

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 // 16):
    if k1[i]:
        zmm1.w[i] = popcount(src2.w[i])
    elif zeroing:
        zmm1.w[i] = 0
    # else: zmm1.w[i] keeps its value
# 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.

All forms

AVX512-E4

Attributes

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

All forms

write masking

Memory forms

memory fault suppression

Intrinsics

The C intrinsics of Clang 23.1.2 that compile to VPOPCNTW, 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 AVX10.1 instructions

• VPMOVW2M
• VPMOVWB
• VPMOVZXBD
• VPMOVZXBQ
• VPMOVZXBW
• VPMOVZXDQ
• VPMOVZXWD
• VPMOVZXWQ
• VPMULDQ
• VPMULHRSW
• VPMULHUW
• VPMULHW
• VPMULLD
• VPMULLQ
• VPMULLW
• VPMULTISHIFTQB
• VPMULUDQ
• VPOPCNTB
• VPOPCNTD
• VPOPCNTQ
• VPORD
• VPORQ
• VPROLD
• VPROLQ
• VPROLVD
• VPROLVQ
• VPRORD
• VPRORQ
• VPRORVD
• VPRORVQ
• VPSADBW
• VPSCATTERDD
• VPSCATTERDQ
• VPSCATTERQD
• VPSCATTERQQ
• VPSHLDD
• VPSHLDQ
• VPSHLDVD
• VPSHLDVQ
• VPSHLDVW

All 686 AVX10.1 instructions