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

SUB (shifted immediate, 64-bit)

Test 1: uops

Code:

  sub x0, x0, #3, lsl #12
  mov x0, 1
  mov x1, 2

(no loop instructions)

1000 unrolls and 1 iteration

Retires: 1.000

Issues: 1.000

Integer unit issues: 1.001

Load/store unit issues: 0.000

SIMD/FP unit issues: 0.000

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)dispatch int uop (56)int uops in schedulers (59)dispatch uop (78)map int uop (7c)map int uop inputs (7f)? int output thing (e9)? int retires (ef)
100410301001100110002560310001000100010011000
100410301001100110002560310001000100010011000
100410301001100110002560310001000100010011000
100410301001100110002560310001000100010011000
100410301001100110002560310001000100010011000
100410301001100110002560310001000100010011000
100410301001100110002560310001000100010011000
100410301001100110002560310001000100010011000
100410301001100110002560310001000100010011000
100410301001100110002560310001000100010011000

Test 2: Latency 1->2

Code:

  sub x0, x0, #3, lsl #12
  mov x0, 1
  mov x1, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 1.0030

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)dispatch int uop (56)int uops in schedulers (59)dispatch uop (78)map int uop (7c)map int uop inputs (7f)? int output thing (e9)? int retires (ef)
10204100301010110101101072593291010710214102601001510100
10204100301010110101101072594151010710214102121000110100
10204100301010110101101072595391010710212102121000110100
10204100301010110101101072595391010710212102121000110100
10204100301010110101101072595391010710212102121000110100
10204100301010110101101072595391010710212102121000110100
10204100301010110101101072595391010710212102121000110100
10204100301010110101101072595391010710212102121000110100
10204100301010110101101072595391010710212102121000110100
10204100301010110101101072595391010710212102121000110100

1000 unrolls and 10 iterations

Result (median cycles for code): 1.0030

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)schedule ldst uop (55)dispatch int uop (56)int uops in schedulers (59)dispatch uop (78)map int uop (7c)map int uop inputs (7f)? int output thing (e9)? int retires (ef)
100241003010021100210100282595281003010038100341001110010
100241003010021100210100202595911002010020100201001110010
100241003010021100210100202595911002010020100201001110010
100241003010021100210100202595911002010020100201001110010
100241003010021100210100202595911002010020100201001110010
100241003010021100210100202595911002010020100201001110010
100241003010021100210100202595911002010020100201001110010
100241003010021100210100202595911002010020100201001110010
100241003010021100210100202595911002010020100201001110010
100241003010021100210100202595911002010020100201001110010

Test 3: throughput

Count: 8

Code:

  sub x0, x8, #3, lsl #12
  sub x1, x8, #3, lsl #12
  sub x2, x8, #3, lsl #12
  sub x3, x8, #3, lsl #12
  sub x4, x8, #3, lsl #12
  sub x5, x8, #3, lsl #12
  sub x6, x8, #3, lsl #12
  sub x7, x8, #3, lsl #12
  mov x8, 9

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.3342

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)dispatch int uop (56)int uops in schedulers (59)dispatch uop (78)map int uop (7c)map int uop inputs (7f)? int output thing (e9)? int retires (ef)
80204268748011580115801202403608012080222802228001580100
80204267678011580115801202403608012080224802248001580100
80204267378011580115801202403608012080224802248001580100
80204267378011580115801202403608012080224802248001580100
80204267378011580115801202403608012080224802248001580100
80204267378011580115801202403608012080224802248001580100
80204267378011580115801202405048017180276802248001580100
80204267378011580115801202403608012080224802248001580100
80204267378011580115801202403608012080224802248001580100
80204267378011580115801202403608012080224802248001580100

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.3339

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)dispatch int uop (56)int uops in schedulers (59)dispatch uop (78)map int uop (7c)map int uop inputs (7f)? int output thing (e9)? int retires (ef)
80024280738003880038800432668118004280044800468002880010
80024267588002180021800202774048002080020800208001180010
80024267138002180021800202774048002080020800208001180010
80024267138002180021800202774048002080020800208001180010
80024267138002180021800202774048002080020800208001180010
80024267138002180021800202774048002080020800208001180010
80024267138002180021800202774048002080020800208001180010
80024267138002180021800202774048002080020800208001180010
80024267138002180021800202774048002080020800208001180010
80024267138002180021800202774048002080020800208001180010