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

UADALP (vector, 4H)

Test 1: uops

Code:

  uadalp v0.4h, v1.8b
  movi v0.16b, 1
  movi v1.16b, 2

(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)map simd uop inputs (81)? int output thing (e9)? simd retires (ee)
1004303310011100010007590510001000200011000
1004303310011100010007590510001000200011000
1004303310011100010007590510001000200011000
1004303310011100010007590510001000200011000
1004303310011100010007590510001000200011000
1004303310011100010007590510001000200011000
1004303310011100010007590510001000200011000
1004303310011100010007590510001000200011000
1004303310011100010007590510001000200011000
1004303310011100010007590510001000200011000

Test 2: Latency 1->1

Code:

  uadalp v0.4h, v1.8b
  movi v0.16b, 1
  movi v1.16b, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 3.0033

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)map simd uop inputs (81)? int output thing (e9)? simd retires (ee)? int retires (ef)
1020430033101011011000010010000300768905101002001000620020008110000100
1020430033101011011000010010000300768905101002001000620020008110000100
1020430033101011011000010010000300768905101002001000420020008110000100
1020430033101011011000010010000300768905101002001000420020008110000100
1020430033101011011000010010000300768905101002001000420020008110000100
1020430033101011011000010010000300768905101002001000420020008110000100
1020430033101011011000010010000300768905101002001000420020008110000100
1020430033101011011000010010000300768905101002001000420020008110000100
1020530066101091031000610210031300768905101002001000420020008110000100
1020430033101011011000010010000300768905101002001000420020008110000100

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0033

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)map simd uop inputs (81)? int output thing (e9)? simd retires (ee)? int retires (ef)
10024300331002121100002010000707689051002020100062020000111000010
10024300331002121100002010000707689051002020100002020000111000010
10024300331002121100002010000707689051002020100002020000111000010
10024300331002121100002010000707689051002020100002020000111000010
10024300331002121100002010000707689051002020100002020000111000010
10024300331002121100002010000707689051002020100002020000111000010
10024300331002121100002010000707689051002020100002020000111000010
10024300331002121100002010000707689051002020100002020000111000010
10024300331002121100002010000707689051002020100002020000111000010
10024300331002121100002010000707689051002020100002020000111000010

Test 3: Latency 1->2

Code:

  uadalp v0.4h, v0.8b
  movi v0.16b, 1

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 3.0033

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)map simd uop inputs (81)? int output thing (e9)? simd retires (ee)? int retires (ef)
1020430033101011011000010010000300768905101002001000420020008110000100
1020430033101011011000010010000300768905101002001000420020008110000100
1020430033101011011000010010000300768905101002001000620020008110000100
1020430033101011011000010010000300768905101002001000420020008110000100
1020430033101011011000010010000300768905101002001000420020008110000100
1020430033101011011000010010000300768905101002001000420020088110000100
1020430033101011011000010010000300768905101002001000420020008110000100
1020430033101011011000010010000300768905101002001000420020008110000100
1020430033101011011000010010000300768905101002001000420020008110000100
1020430033101011011000010010000300768905101002001000420020008110000100

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0033

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)map simd uop inputs (81)? int output thing (e9)? simd retires (ee)? int retires (ef)
10024300331002121100002010000140170772193247735321356227981720100822020008111000010
10024300331002121100002010000070076890510020200100002020000111000010
10024300331002121100002010000070076890510020200100002020000111000010
10024300331002121100002010000070076890510020200100002020000111000010
10024300331002121100002010000070076890510020200100002020000111000010
10024300331002121100002010000070076890510020200100002020000111000010
10024300331002121100002010000070076890510020200100002020000111000010
10024300331002121100002010000070076890510020200100002020000111000010
10024300331002121100002010000070076890510020200100002020000111000010
10024300331002121100002010000070076890510020200100002020000111000010

Test 4: throughput

Count: 8

Code:

  movi v0.16b, 0
  uadalp v0.4h, v8.8b
  movi v1.16b, 0
  uadalp v1.4h, v8.8b
  movi v2.16b, 0
  uadalp v2.4h, v8.8b
  movi v3.16b, 0
  uadalp v3.4h, v8.8b
  movi v4.16b, 0
  uadalp v4.4h, v8.8b
  movi v5.16b, 0
  uadalp v5.4h, v8.8b
  movi v6.16b, 0
  uadalp v6.4h, v8.8b
  movi v7.16b, 0
  uadalp v7.4h, v8.8b
  movi v8.16b, 9

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.5011

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)map ldst uop inputs (80)map simd uop inputs (81)? int output thing (e9)? ldst retires (ed)? simd retires (ee)? int retires (ef)
160204405188010910180008100800123003200528011220080012200016002410160000100
160204401088011010180009100800133003200528011220080012200016002410160000100
160204400868010910180008100800123003200568011320080013200016002410160000100
160204400868010910180008100800123003200528011220080012200016002410160000100
160204400868010910180008100800123003200528011220080012200016002410160000100
160204400868010910180008100800123003200528011220080012200016002410160000100
160204400868010910180008100800123003200528011220080012200016010010160000100
160204400868010910180008100800123003200568011320080013200016009610160000100
160204400988011010180009100800133003200528011220080012200016002610160000100
160204400868010910180008100800123003200528011220080012200016002410160000100

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.5053

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)map simd uop inputs (81)? int output thing (e9)? simd retires (ee)? int retires (ef)
1600244371680019118000810800123032005680023208001320160000116000010
1600244126480011118000010800003032000080010208000020160000116000010
1600254092280054118004310800473032000080010208000020160000116000010
1600244045980011118000010800003032000080010208000020160000116000010
1600244041480011118000010800003032000080010208000020160000116000010
1600244042780011118000010800003032000080010208000020160000116000010
1600244038880011118000010800003032000080010208000020160000116000010
1600244041680011118000010800003032000080010208000020160000116000010
1600244041680011118000010800003032000080010208000020160000116000010
1600244042780011118000010800003032000080010208000020160000116000010

Test 5: throughput

Count: 16

Code:

  uadalp v0.4h, v16.8b
  uadalp v1.4h, v16.8b
  uadalp v2.4h, v16.8b
  uadalp v3.4h, v16.8b
  uadalp v4.4h, v16.8b
  uadalp v5.4h, v16.8b
  uadalp v6.4h, v16.8b
  uadalp v7.4h, v16.8b
  uadalp v8.4h, v16.8b
  uadalp v9.4h, v16.8b
  uadalp v10.4h, v16.8b
  uadalp v11.4h, v16.8b
  uadalp v12.4h, v16.8b
  uadalp v13.4h, v16.8b
  uadalp v14.4h, v16.8b
  uadalp v15.4h, v16.8b
  movi v16.16b, 17

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.5002

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)map ldst uop inputs (80)map simd uop inputs (81)? int output thing (e9)? ldst retires (ed)? simd retires (ee)? int retires (ef)
16020480076160105101160004100160008300640036160108200160012200032002410160000100
16020480035160107101160006100160010300640036160108200160012200032002410160000100
16020480035160105101160004100160008300640036160108200160012200032002410160000100
16020480035160107101160006100160010300640044160110200160014200032002410160000100
16020480035160105101160004100160008300640036160108200160012200032012410160000100
16020480035160105101160004100160008300640036160108200160012200032002410160000100
16020480035160105101160004100160008300640036160108200160012200032002410160000100
16020480035160105101160004100160008300640036160108200160012200032002410160000100
16020580070160143101160042100160054300640204160153200160061200032002810160000100
16020480035160105101160004100160008300640036160108200160012200032002410160000100

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.5002

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)map simd uop inputs (81)? int output thing (e9)? simd retires (ee)? int retires (ef)
160024801191600151116000410160008306400001600102016000020320000116000010
160024800511600111116000010160000306400001600102016000020320000116000010
160024800351600111116000010160000306400001600102016000020320000116000010
160024800351600111116000010160000306400001600102016000020320000116000010
160024800351600111116000010160000306400001600102016000020320000116000010
160024800351600111116000010160000306402001600622016006220320000116000010
160024800351600111116000010160000306400001600102016000020320000116000010
160024800351600111116000010160000306400001600102016000020320000116000010
160025800701600511116004010160052306400361600182016001220320000116000010
160024800411600111116000010160000306400001600102016000020320000116000010