VPHADDD (YMM, YMM, YMM)
Summary:
"Packed Horizontal Add"
Reference:
https://www.felixcloutier.com/x86/phaddw:phaddd
Extension:
AVX2
Category:
AVX2
ISA-Set:
AVX2
CPL:
3
iform:
VPHADDD_YMMqq_YMMqq_YMMqq
iclass:
VPHADDD
ASM:
VPHADDD
Operands
Operand 1 (w): Register (YMM0, YMM1, YMM2, YMM3, YMM4, YMM5, YMM6, YMM7, YMM8, YMM9, YMM10, YMM11, YMM12, YMM13, YMM14, YMM15)
Operand 2 (r): Register (YMM0, YMM1, YMM2, YMM3, YMM4, YMM5, YMM6, YMM7, YMM8, YMM9, YMM10, YMM11, YMM12, YMM13, YMM14, YMM15)
Operand 3 (r): Register (YMM0, YMM1, YMM2, YMM3, YMM4, YMM5, YMM6, YMM7, YMM8, YMM9, YMM10, YMM11, YMM12, YMM13, YMM14, YMM15)
Available performance data
Arrow Lake-P
Arrow Lake-E
Meteor Lake-P
Meteor Lake-E
Emerald Rapids
Alder Lake-P
Alder Lake-E
Rocket Lake
Tiger Lake
Ice Lake
Cascade Lake
Cannon Lake
Skylake-X
Coffee Lake
Kaby Lake
Skylake
Broadwell
Haswell
AMD Zen 5
AMD Zen 4
AMD Zen 3
AMD Zen 2
AMD Zen+
Arrow Lake-P
Measurements
Latencies
Latency operand 2 → 1:
2
Latency operand 3 → 1:
2
Throughput
Measured (loop):
1.00
Measured (unrolled):
1.00
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (5 other instructions can be decoded with simple decoders in the same cycle)
Port usage:
3*UNKNOWN
Arrow Lake-E
Measurements
Latencies
Latency operand 2 → 1:
2
Latency operand 3 → 1:
2
Throughput
Measured (loop):
2.05
Measured (unrolled):
2.75
Number of μops
Executed: 8
Microcode Sequencer (MS): 8
Requires the complex decoder
Meteor Lake-P
Measurements
Latencies
Latency operand 2 → 1:
2
Latency operand 3 → 1:
2
Throughput
Computed from the port usage: 1.00
Measured (loop):
1.05
Measured (unrolled):
1.05
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (3 other instructions can be decoded with simple decoders in the same cycle)
Port usage:
1*p015+2*p15
Meteor Lake-E
Measurements
Latencies
Latency operand 2 → 1:
3
Latency operand 3 → 1:
3
Throughput
Measured (loop):
3.10
Measured (unrolled):
3.50
Number of μops
Executed: 6
Microcode Sequencer (MS): 6
Requires the complex decoder
Emerald Rapids
Measurements
Latencies
Latency operand 2 → 1:
2
Latency operand 3 → 1:
2
Throughput
Computed from the port usage: 1.00
Measured (loop):
1.05
Measured (unrolled):
1.05
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (3 other instructions can be decoded with simple decoders in the same cycle)
Port usage:
1*p015+2*p15
Alder Lake-P
Measurements
Latencies
Latency operand 2 → 1:
2
Latency operand 3 → 1:
2
Throughput
Computed from the port usage: 1.00
Measured (loop):
1.05
Measured (unrolled):
1.04
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (3 other instructions can be decoded with simple decoders in the same cycle)
Port usage:
1*p015+2*p15
Alder Lake-E
Measurements
Latencies
Latency operand 2 → 1:
3
Latency operand 3 → 1:
3
Throughput
Measured (loop):
3.10
Measured (unrolled):
3.56
Number of μops
Executed: 6
Microcode Sequencer (MS): 6
Requires the complex decoder
Rocket Lake
Measurements
Latencies
Latency operand 2 → 1:
2
Latency operand 3 → 1:
2
Throughput
Computed from the port usage: 1.00
Measured (loop):
1.07
Measured (unrolled):
1.07
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (2 other instructions can be decoded with simple decoders in the same cycle)
Port usage:
1*p015+2*p15
Tiger Lake
Measurements
Latencies
Latency operand 2 → 1:
2
Latency operand 3 → 1:
2
Throughput
Computed from the port usage: 1.00
Measured (loop):
1.07
Measured (unrolled):
1.07
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (2 other instructions can be decoded with simple decoders in the same cycle)
Port usage:
1*p015+2*p15
Ice Lake
Measurements
Latencies
Latency operand 2 → 1:
2
Latency operand 3 → 1:
2
Throughput
Computed from the port usage: 1.00
Measured (loop):
1.07
Measured (unrolled):
1.07
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (2 other instructions can be decoded with simple decoders in the same cycle)
Port usage:
1*p015+2*p15
Documentation
Throughput: 1.0
Cascade Lake
Measurements
Latencies
Latency operand 2 → 1:
3
Latency operand 3 → 1:
3
Throughput
Computed from the port usage: 2.00
Measured (loop):
2.00
Measured (unrolled):
2.00
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (2 other instructions can be decoded with simple decoders in the same cycle)
Port usage:
1*p015+2*p5
Cannon Lake
Measurements
Latencies
Latency operand 2 → 1:
3
Latency operand 3 → 1:
3
Throughput
Computed from the port usage: 2.00
Measured (loop):
2.00
Measured (unrolled):
2.00
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (2 other instructions can be decoded with simple decoders in the same cycle)
Port usage:
1*p015+2*p5
Skylake-X
Measurements
Latencies
Latency operand 2 → 1:
3
Latency operand 3 → 1:
3
Throughput
Computed from the port usage: 2.00
Measured (loop):
2.00
Measured (unrolled):
2.00
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (2 other instructions can be decoded with simple decoders in the same cycle)
Port usage:
1*p015+2*p5
IACA 2.3
Throughput
Computed from the port usage: 2.00
IACA:
2.00
Number of μops:
3
Port usage:
1*p015+2*p5
IACA 3.0
Throughput
Computed from the port usage: 2.00
IACA:
1.92
Number of μops:
3
Port usage:
1*p015+2*p5
Coffee Lake
Measurements
Latencies
Latency operand 2 → 1:
3
Latency operand 3 → 1:
3
Throughput
Computed from the port usage: 2.00
Measured (loop):
2.00
Measured (unrolled):
2.00
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (2 other instructions can be decoded with simple decoders in the same cycle)
Port usage:
1*p015+2*p5
Kaby Lake
Measurements
Latencies
Latency operand 2 → 1:
3
Latency operand 3 → 1:
3
Throughput
Computed from the port usage: 2.00
Measured (loop):
2.00
Measured (unrolled):
2.00
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (2 other instructions can be decoded with simple decoders in the same cycle)
Port usage:
1*p015+2*p5
Skylake
Measurements
Latencies
Latency operand 2 → 1:
3
Latency operand 3 → 1:
3
Throughput
Computed from the port usage: 2.00
Measured (loop):
2.00
Measured (unrolled):
2.00
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (2 other instructions can be decoded with simple decoders in the same cycle)
Port usage:
1*p015+2*p5
IACA 2.3
Throughput
Computed from the port usage: 2.00
IACA:
2.00
Number of μops:
3
Port usage:
1*p015+2*p5
IACA 3.0
Throughput
Computed from the port usage: 2.00
IACA:
1.92
Number of μops:
3
Port usage:
1*p015+2*p5
Broadwell
Measurements
Latencies
Latency operand 2 → 1:
3
Latency operand 3 → 1:
3
Throughput
Computed from the port usage: 2.00
Measured (loop):
2.00
Measured (unrolled):
2.00
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
1*p15+2*p5
IACA 2.2
Throughput
Computed from the port usage: 2.00
IACA:
2.00 (with the -no_interiteration flag: 2.00)
Number of μops:
3
Port usage:
1*p15+2*p5
IACA 2.3
Throughput
Computed from the port usage: 2.00
IACA:
2.00
Number of μops:
3
Port usage:
1*p15+2*p5
IACA 3.0
Throughput
Computed from the port usage: 2.00
IACA:
1.94
Number of μops:
3
Port usage:
1*p15+2*p5
Haswell
Measurements
Latencies
Latency operand 2 → 1:
3
Latency operand 3 → 1:
3
Throughput
Computed from the port usage: 2.00
Measured (loop):
2.00
Measured (unrolled):
2.00
Number of μops
Executed: 3
Retire slots: 3
Decoded (MITE): 3
Microcode Sequencer (MS): 0
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
1*p15+2*p5
IACA 2.1
Latency:
3
Throughput
Computed from the port usage: 2.00
IACA:
2.00 (with the -no_interiteration flag: 2.00)
Number of μops:
3
Port usage:
1*p15+2*p5
IACA 2.2
Throughput
Computed from the port usage: 2.00
IACA:
2.00 (with the -no_interiteration flag: 2.00)
Number of μops:
3
Port usage:
1*p15+2*p5
IACA 2.3
Throughput
Computed from the port usage: 2.00
IACA:
2.00
Number of μops:
3
Port usage:
1*p15+2*p5
IACA 3.0
Throughput
Computed from the port usage: 2.00
IACA:
1.92
Number of μops:
3
Port usage:
1*p15+2*p5
AMD Zen 5
Measurements
Latencies
Latency operand 2 → 1:
3
Latency operand 3 → 1:
3
Throughput
Computed from the port usage: 0.75
Measured (loop):
2.00
Measured (unrolled):
2.00
Number of μops
Executed: 3
Port usage:
1*FP0123+1*FP12+1*FP23
Documentation
Latency: 2
Throughput: 2.00
Number of μops: 3
Port usage: FP0/1/2/3, FP0/1/2/3, FP0/1/2/3
AMD Zen 4
Measurements
Latencies
Latency operand 2 → 1:
2
Latency operand 3 → 1:
2
Throughput
Computed from the port usage: 0.75
Measured (loop):
2.00
Measured (unrolled):
2.00
Number of μops
Executed: 3
Port usage:
1*FP0123+1*FP123+1*FP23
Documentation
Number of μops: ucode
AMD Zen 3
Measurements
Latencies
Latency operand 2 → 1:
2
Latency operand 3 → 1:
2
Throughput
Computed from the port usage: 1.00
Measured (loop):
2.00
Measured (unrolled):
2.00
Number of μops
Executed: 3
Port usage:
1*FP0123+1*FP1+1*FP12
Documentation
Number of μops: ucode
AMD Zen 2
Measurements
Latencies
Latency operand 2 → 1:
2
Latency operand 3 → 1:
2
Throughput
Measured (loop):
2.00
Measured (unrolled):
2.00
Number of μops
Executed: 3
Documentation
Number of μops: ucode
AMD Zen+
Measurements
Latencies
Latency operand 2 → 1:
3
Latency operand 3 → 1:
3
Throughput
Measured (loop):
3.00
Measured (unrolled):
3.00
Number of μops
Executed: 8
Documentation
Latency: 3
Throughput: 3.00