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

FCVTPU (scalar, S to X)

Test 1: uops

Code:

  fcvtpu x0, s0
  mov x0, 1
  mov x1, 2

(no loop instructions)

1000 unrolls and 1 iteration

Retires: 2.000

Issues: 3.000

Integer unit issues: 1.001

Load/store unit issues: 0.000

SIMD/FP unit issues: 2.000

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)? int retires (ef)
20041039300110012000200011000200020002000100110001000
20041037300110012000200011000200020002000100110001000
20041037300110012000200011000200020002000100110001000
20041037300110012000200011000200020002000100110001000
20041037300110012000200011000200020002000100110001000
20041037300110012000200011000200020002000100110001000
20041037300110012000200011000200020002000100110001000
20041037300110012000200011000200020002000100110001000
20041037300110012000200011000200020002000100110001000
20041037300110012000200011000200020002000100110001000

Test 2: Latency 1->2 roundtrip

Code:

  fcvtpu x0, s0
  fmov d0, x0
  mov x0, 1
  mov x1, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 10.0032

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)schedule simd uop (54)schedule ldst uop (55)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 ldst uop inputs (80)map simd uop inputs (81)? int output thing (e9)? ldst retires (ed)? simd retires (ee)? int retires (ef)
3020410003240101101012000010000100200001000030015467862578206301002001000220004200100012000310001100001000010100
3020410003240101101012000010000100200001001630015479832580139301462001001920040200100012000310001100001000010100
3020410003240101101012000010000100200001000030015467962578206301002001000120003200100012000310001100001000010100
3020510009440109101042000310002102200281000030015468302578258301002001000120003200100012000310001100001000010100
3020410003240101101012000010000100200001000030015467962578206301002001000120003200100012000310001100001000010100
3020410003240101101012000010000100200001000030015467962578206301002001000120003200100012000310001100001000010100
3020410003240101101012000010000100200001000030015467962578206301002001000120003200100012000310001100001000010100
3020510007540107101022000310002100200281001930015482502580613301492001002020040200100012000310001100001000010100
3020410003240101101012000010000100200001000030015467962578206301002001000120003200100012000310001100001000010100
3020410003240101101012000010000100200001000030015467962578206301002001000120003200100012000310001100001000010100

1000 unrolls and 10 iterations

Result (median cycles for code): 10.0035

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)schedule simd uop (54)schedule ldst uop (55)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 ldst uop inputs (80)map simd uop inputs (81)? int output thing (e9)? ldst retires (ed)? simd retires (ee)? int retires (ef)
300251001194001710012200041000110200301001630154716225795413005620100172003620100012000310001100001000010010
300241000324001110011200001000010200001000030154634625782063001020100002000020100002000010001100001000010010
300251001134001710012200031000210200281001730154738525798793005720100192004120100002000010001100001000010010
300241000324001110011200001000010200001000030154634625782063001020100002000020100002000010001100001000010010
300241000324001110011200001000010200001000030154634625782063001020100012000320100022000410001100001000010010
300241000324001110011200001000010200001000030154634625782063001020100002000020100002000010001100001000010010
300241000324001110011200001000010200001000030154634625782063001020100002000020100002000010001100001000010010
300241000324001110011200001000010200001000030154634625782063001020100002000020100002000010001100001000010010
300241000324001110011200001000010200001000030154634625782063001020100002000020100002000010001100001000010010
300241000324001110011200001000010200001000030154634625782063001020100002000020100002000010001100001000010010

Test 3: throughput

Count: 8

Code:

  fcvtpu x0, s8
  fcvtpu x1, s8
  fcvtpu x2, s8
  fcvtpu x3, s8
  fcvtpu x4, s8
  fcvtpu x5, s8
  fcvtpu x6, s8
  fcvtpu x7, s8
  mov x8, 9

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 1.0005

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)
1602048004224010380101160002100160008300880041160108200160012200160012800018000080100
1602048003724010380101160002100160008300880041160108200160012200160068800158000080100
1602048003724010380101160002100160008300880041160108200160012200160012800018000080100
1602048003724010380101160002100160008300880041160108200160012200160012800018000080100
1602048003724010380101160002100160008300880041160108200160012200160012800018000080100
1602048003724010380101160002100160008300880041160108200160012200160012800018000080100
1602048003724010380101160002100160008300880041160108200160012200160012800018000080100
1602048003724010380101160002100160008300880041160108200160012200160012800018000080100
1602048003724010380101160002100160008300880041160108200160012200160068800158000080100
1602048003724010380101160002100160008300880041160108200160012200160012800018000080100

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 1.0005

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)
160024800632400138001116000210160008308800001600102016000020160000800018000080010
160024800372400118001116000010160000308800001600102016000020160000800018000080010
160024800372400118001116000010160000308800001600102016000020160000800018000080010
160024800372400118001116000010160000308800001600102016000020160000800018000080010
160024800372400118001116000010160000308800001600102016000020160000800018000080010
160024800372400118001116000010160000308800001600102016000020160000800018000080010
160024800372400118001116000010160000308800001600102016000020160000800018000080010
160024800372400118001116000010160000308800001600102016000020160068800158000080010
160024800372400118001116000010160000308800001600102016000020160000800018000080010
160024800372400118001116000010160000308800001600102016000020160000800018000080010