Apple Microarchitecture Research by Dougall Johnson

M1/A14 P-core (Firestorm): Overview | Base Instructions | SIMD and FP Instructions
M1/A14 E-core (Icestorm):  Overview | Base Instructions | SIMD and FP Instructions

MOVI (vector, 4S, lsl)

Test 1: uops

Code:

  movi v0.4s, #3, lsl #8

(no loop instructions)

1000 unrolls and 1 iteration

Retires: 1.000

Issues: 1.000

Integer unit issues: 0.001

Load/store unit issues: 0.000

SIMD/FP unit issues: 1.000

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)schedule simd uop (54)dispatch simd uop (57)ldst uops in schedulers (5b)dispatch uop (78)map simd uop (7e)? int output thing (e9)? simd retires (ee)
1004584100111000100040001000100011000
1004534100111000100040001000100011000
1004534100111000100040001000100011000
1004534100111000100040001000100011000
1004534100111000100040001000100011000
1004534100111000100040001000100011000
1004534100111000100040001000100011000
1004534100111000100040001000100011000
1004534100111000100040001000100011000
1004534100111000100040001000100011000

Test 2: throughput

Count: 8

Code:

  movi v0.4s, #3, lsl #8
  movi v1.4s, #3, lsl #8
  movi v2.4s, #3, lsl #8
  movi v3.4s, #3, lsl #8
  movi v4.4s, #3, lsl #8
  movi v5.4s, #3, lsl #8
  movi v6.4s, #3, lsl #8
  movi v7.4s, #3, lsl #8

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.5004

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)schedule simd uop (54)dispatch int uop (56)dispatch simd uop (57)dispatch ldst uop (58)int uops in schedulers (59)simd uops in schedulers (5a)ldst uops in schedulers (5b)dispatch uop (78)map int uop (7c)map ldst uop (7d)map simd uop (7e)map int uop inputs (7f)? int output thing (e9)? simd retires (ee)? int retires (ef)
80204400448010710180006100800100300032004480110200080014200180000100
80204400358010910180008100800120300032004480110200080014200180000100
80204400348010510180004100800080300032003680108200080012200180000100
80204400348010510180004100800080300032003680108200080012200180000100
80204400348010510180004100800080300032003680108200080012200180000100
80204400348010510180004100800080300032003680108200080012200180000100
80204400348010510180004100800080300032003680108200080012200180000100
80204400348010510180004100800080300032003680108200080012200180000100
80204400348010510180004100800080300032003680108200080012200180000100
80204400348010510180004100800080300032003680108200080012200180000100

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.5004

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)schedule simd uop (54)dispatch int uop (56)dispatch simd uop (57)int uops in schedulers (59)ldst uops in schedulers (5b)dispatch uop (78)map int uop (7c)map simd uop (7e)map int uop inputs (7f)? int output thing (e9)? simd retires (ee)? int retires (ef)
800244016780027218000620800106832005280032208001820118000010
800244003480021218000020800006532000080020208000020118000010
800244003480021218000020800006532000080020208000020118000010
800244003480021218000020800006532000080020208000020118000010
800244003480021218000020800006532000080020208000020118000010
800244003480021218000020800006532000080020208000020118000010
800244003480021218000020800006532000080020208000020118000010
800244003480021218000020800006532000080020208000020118000010
800244003480021218000020800006632020880074208006020118000010
800244010280027218000620800106532000080020208000020118000010