RDTSCP
Summary:
"Read Time-Stamp Counter and Processor ID"
Reference:
https://www.felixcloutier.com/x86/RDTSCP.html
Extension:
RDTSCP
Category:
SYSTEM
ISA-Set:
RDTSCP
CPL:
3
iform:
RDTSCP
iclass:
RDTSCP
ASM:
RDTSCP
Operands
Operand 1 (w, suppressed): Register (EAX)
Operand 2 (w, suppressed): Register (EDX)
Operand 3 (w, suppressed): Register (ECX)
Operand 4 (r, suppressed): Register (TSC)
Operand 5 (r, suppressed): Register (TSCAUX)
Available performance data
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
Ivy Bridge
Sandy Bridge
Westmere
Nehalem
Tremont
Goldmont Plus
Goldmont
Airmont
AMD Zen 4
AMD Zen 3
AMD Zen 2
AMD Zen+
Alder Lake-P
Measurements
Throughput
Computed from the port usage: 5.25
Measured (loop):
33.00
Measured (unrolled):
33.00
Number of μops
Executed: 20
Retire slots: 23
Decoded (MITE): 2
Microcode Sequencer (MS): 21
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
2*p0+2*p01+1*p015+2*p056+7*p06+4*p1+3*p5
Alder Lake-E
Measurements
Throughput
Measured (loop):
41.75
Measured (unrolled):
41.75
Number of μops
Executed: 17
Microcode Sequencer (MS): 16
Requires the complex decoder
Rocket Lake
Measurements
Throughput
Computed from the port usage: 5.00
Measured (loop):
35.00
Measured (unrolled):
36.17
Number of μops
Executed: 17
Retire slots: 20
Decoded (MITE): 2
Microcode Sequencer (MS): 18
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
2*p0+1*p01+1*p05+6*p06+5*p1+1*p15+1*p5
Tiger Lake
Measurements
Throughput
Computed from the port usage: 5.00
Measured (loop):
35.90
Measured (unrolled):
36.00
Number of μops
Executed: 17
Retire slots: 20
Decoded (MITE): 2
Microcode Sequencer (MS): 18
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
2*p0+1*p01+6*p06+5*p1+2*p5
Ice Lake
Measurements
Throughput
Computed from the port usage: 5.00
Measured (loop):
35.90
Measured (unrolled):
36.00
Number of μops
Executed: 17
Retire slots: 20
Decoded (MITE): 2
Microcode Sequencer (MS): 18
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
2*p0+1*p01+1*p0156+5*p06+5*p1+2*p5
Cascade Lake
Measurements
Throughput
Computed from the port usage: 6.00
Measured (loop):
32.06
Measured (unrolled):
32.00
Number of μops
Executed: 20
Retire slots: 20
Decoded (MITE): 0
Microcode Sequencer (MS): 20
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
3*p0+7*p06+6*p1+4*p5
Cannon Lake
Measurements
Throughput
Computed from the port usage: 5.00
Measured (loop):
32.25
Measured (unrolled):
32.00
Number of μops
Executed: 16
Retire slots: 18
Decoded (MITE): 4
Microcode Sequencer (MS): 14
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
2*p0+1*p0156+1*p05+4*p06+5*p1+1*p15+2*p5
Skylake-X
Measurements
Throughput
Computed from the port usage: 6.00
Measured (loop):
32.06
Measured (unrolled):
32.25
Number of μops
Executed: 20
Retire slots: 20
Decoded (MITE): 0
Microcode Sequencer (MS): 20
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
3*p0+7*p06+6*p1+4*p5
IACA 2.3
Throughput
Computed from the port usage: 5.50
IACA:
5.71
Number of μops:
22
Port usage:
20*p0156+2*p5
IACA 3.0
Throughput
Computed from the port usage: 5.50
IACA:
5.37
Number of μops:
22
Port usage:
20*p0156+2*p5
Coffee Lake
Measurements
Throughput
Computed from the port usage: 6.00
Measured (loop):
32.06
Measured (unrolled):
32.25
Number of μops
Executed: 20
Retire slots: 20
Decoded (MITE): 0
Microcode Sequencer (MS): 20
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
3*p0+7*p06+6*p1+4*p5
Kaby Lake
Measurements
Throughput
Computed from the port usage: 6.00
Measured (loop):
32.06
Measured (unrolled):
32.25
Number of μops
Executed: 20
Retire slots: 20
Decoded (MITE): 0
Microcode Sequencer (MS): 20
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
3*p0+7*p06+6*p1+4*p5
Skylake
Measurements
Throughput
Computed from the port usage: 6.00
Measured (loop):
32.06
Measured (unrolled):
32.25
Number of μops
Executed: 20
Retire slots: 20
Decoded (MITE): 0
Microcode Sequencer (MS): 20
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
3*p0+7*p06+6*p1+4*p5
IACA 2.3
Throughput
Computed from the port usage: 5.50
IACA:
5.71
Number of μops:
22
Port usage:
20*p0156+2*p5
IACA 3.0
Throughput
Computed from the port usage: 5.50
IACA:
5.37
Number of μops:
22
Port usage:
20*p0156+2*p5
Broadwell
Measurements
Throughput
Computed from the port usage: 5.50
Measured (loop):
30.00
Measured (unrolled):
30.00
Number of μops
Executed: 21
Retire slots: 21
Decoded (MITE): 0
Microcode Sequencer (MS): 21
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
4*p0+7*p06+5*p1+1*p15+4*p5
IACA 2.2
Throughput
Computed from the port usage: 5.50
IACA:
5.75 (with the -no_interiteration flag: 5.55)
Number of μops:
22
Port usage:
20*p0156+2*p5
IACA 2.3
Throughput
Computed from the port usage: 5.50
IACA:
5.71
Number of μops:
22
Port usage:
20*p0156+2*p5
IACA 3.0
Throughput
Computed from the port usage: 5.50
IACA:
5.37
Number of μops:
22
Port usage:
20*p0156+2*p5
Haswell
Measurements
Throughput
Computed from the port usage: 5.50
Measured (loop):
30.00
Measured (unrolled):
30.00
Number of μops
Executed: 21
Retire slots: 21
Decoded (MITE): 0
Microcode Sequencer (MS): 21
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
4*p0+7*p06+5*p1+1*p15+4*p5
IACA 2.2
Throughput
Computed from the port usage: 5.50
IACA:
5.75 (with the -no_interiteration flag: 5.55)
Number of μops:
22
Port usage:
20*p0156+2*p5
IACA 2.3
Throughput
Computed from the port usage: 5.50
IACA:
5.71
Number of μops:
22
Port usage:
20*p0156+2*p5
IACA 3.0
Throughput
Computed from the port usage: 5.50
IACA:
5.37
Number of μops:
22
Port usage:
20*p0156+2*p5
Ivy Bridge
Measurements
Throughput
Computed from the port usage: 10.00
Measured (loop):
34.50
Measured (unrolled):
34.50
Number of μops
Executed: 23
Retire slots: 23
Decoded (MITE): 0
Microcode Sequencer (MS): 23
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
5*p0+1*p015+7*p1+10*p5
Sandy Bridge
Measurements
Throughput
Computed from the port usage: 10.00
Measured (loop):
36.00
Measured (unrolled):
36.00
Number of μops
Executed: 23
Retire slots: 23
Decoded (MITE): 0
Microcode Sequencer (MS): 23
Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
Port usage:
5*p0+1*p01+7*p1+10*p5
Westmere
Measurements
Throughput
Computed from the port usage: 8.00
Measured (loop):
30.02
Measured (unrolled):
30.00
Number of μops
Executed: 23
Retire slots: 23
Microcode Sequencer (MS): 48
Requires the complex decoder
Port usage:
6*p0+1*p015+8*p1+8*p5
IACA 2.1
Latency:
9
Throughput
Computed from the port usage: 2.00
IACA:
2.00 (with the -no_interiteration flag: 2.00)
Number of μops:
4
Port usage:
2*p1+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:
4
Port usage:
2*p1+2*p5
Nehalem
Measurements
Throughput
Computed from the port usage: 10.00
Measured (loop):
31.50
Measured (unrolled):
31.50
Number of μops
Executed: 25
Retire slots: 25
Microcode Sequencer (MS): 59
Requires the complex decoder
Port usage:
6*p0+9*p1+10*p5
IACA 2.1
Latency:
9
Throughput
Computed from the port usage: 2.00
IACA:
2.00 (with the -no_interiteration flag: 2.00)
Number of μops:
4
Port usage:
2*p1+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:
4
Port usage:
2*p1+2*p5
Tremont
Measurements
Throughput
Measured (loop):
39.25
Measured (unrolled):
39.25
Number of μops
Executed: 23
Microcode Sequencer (MS): 23
Requires the complex decoder
Goldmont Plus
Measurements
Throughput
Measured (loop):
34.50
Measured (unrolled):
34.50
Number of μops
Executed: 22
Microcode Sequencer (MS): 22
Requires the complex decoder
Goldmont
Measurements
Throughput
Measured (loop):
31.00
Measured (unrolled):
31.00
Number of μops
Executed: 20
Microcode Sequencer (MS): 20
Requires the complex decoder
Airmont
Measurements
Throughput
Measured (loop):
34.50
Measured (unrolled):
34.50
Number of μops
Executed: 13
Microcode Sequencer (MS): 13
Requires the complex decoder
AMD Zen 4
Measurements
Throughput
Measured (loop):
65.00
Measured (unrolled):
65.00
Number of μops
Executed: 46
AMD Zen 3
Measurements
Throughput
Measured (loop):
69.00
Measured (unrolled):
69.00
Number of μops
Executed: 44
Documentation
Number of μops: ucode
AMD Zen 2
Measurements
Throughput
Measured (loop):
68.00
Measured (unrolled):
68.00
Number of μops
Executed: 37
Documentation
Number of μops: ucode
AMD Zen+
Measurements
Throughput
Measured (loop):
65.00
Measured (unrolled):
65.00
Number of μops
Executed: 37
Documentation
Number of μops: ucode