Apple M1 Microarchitecture Research by Dougall Johnson

Firestorm: Overview | Base Instructions | SIMD and FP Instructions
Icestorm:  Overview | Base Instructions | SIMD and FP Instructions

SUQADD (vector, 2S)

Test 1: uops

Code:

  suqadd v0.2s, v1.2s
  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: 0.000

Load/store unit issues: 0.000

SIMD/FP unit issues: 1.000

retire (01)cycle (02)03081e3f4e51inst issue (52)~issue fp/simd (54)~dispatch fp/simd (57)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op fp/simd (7e)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst neon or fp (9a)a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
10043037220015625482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110000373116112630100030383038303830383038
1004303723008025482510001000100039831303018303730372415328951000100020003037303711100110000673116112630100030383038303830383038
1004303723006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303722006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110000973116112630100030383038303830383038
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110004373116112630100030383038303830383038

Test 2: Latency 1->1

Code:

  suqadd v0.2s, v1.2s
  movi v0.16b, 1
  movi v1.16b, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03080b181e3f4e51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa0a6a8acbranch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102043003722500006129548251010010010000100100005004277313300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
1020430037225000012429548251010010010000100100005004277313300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
1020430037225000017029548251010010010000100100005004277313300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
1020430037225000048029548251010010010000100100005004277313300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
1020430037225000031729548251010010010000100100005004277313300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
10204300372250010612954825101001001000011610000500427731330018300373008428265328745101002001000020020000300373003711102011009910010010000100000123071011611296340100001003003830038300383003830038
1020430037225000017029548251010010010000100100005004277313300183003730037282653287451041620010000200200003003730037111020110099100100100001000010071011611296340100001003003830038300383003830038
102043003722501036129548251010010010000100100005004277313300183003730037282653287451010020410000200203303003730037111020110099100100100001002200071011611296340100001003003830038300383003830038
1020430037225000014529548251010010010000100100005004277313300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
1020430037225000014529548251010010010000100100005004277313300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003008530038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)030818191e1f3a3f4e51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa0a1a6a7a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024300372250000008429548251001010100001010000504277313030018301313022628298252883710712201000020213023027230228611002110910101000010000000006402163229630010000103003830038300383003830038
10024300372250541500612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038
10024300372250009001812954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038
1002430037225004660264061295482510010101000010100005042773130300183003730037282873628898113542010000202000030037300371110021109101010000102001001112006402162229630010000103003830038300383007430038
100243003722500000073729548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100000001113506402162229630010000103003830038300383003830038
1002430037225000000612954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010400000007884982229630010000103003830038300383013230038
100243003722500053161606129548251001010100001010000504277313130018300373003728287328767109122211140242198030369300848110021109101010000100000121972806402513429882010000103003830038300383003830038
1002430037225000000612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038
1002430037225000000612954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038
1002430037225000000612953925100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038

Test 3: Latency 1->2

Code:

  suqadd v0.2s, v0.2s
  movi v0.16b, 1

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)0318191e1f3a3f4e51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa1a6a7a8acc5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372250000014729547251010010010000100100005004277160300183003730037282716287401010020010008200200163003730037111020110099100100100001000000011171701600296460100001003003830038300383003830038
1020430037224000008429547251010010010000100100005004277160300183003730037282716287411010020010008200200163003730037111020110099100100100001000000011171801600296450100001003003830038300383003830085
10204300372250000014929547251010010010000100100005004277160300183003730037282716287401010020010008200200163003730037111020110099100100100001000000011171801600296450100001003003830038300383003830038
10204300372240090014929547251010010010000100100005004277160300183003730037282717287411010020010008200200163003730037111020110099100100100001000000011171801600296450100001003003830038300383003830038
1020430037225000008229547251010010010000100100005004277160300183003730037282716287411010020010008200200163003730037111020110099100100100001000000011171701600296460100001003003830038300383003830038
1020430037224000006129547251010010010000100100005004277160300183003730037282716287401010020010008200200163003730037111020110099100100100001000000011171701600296450100001003003830038300383003830038
10204300372250000010729547251010010010000100100005004277160300183003730037282716287411010020010008200200163003730037111020110099100100100001000000011171801600296460100001003003830038300383003830038
1020430037225000006129547251010010010000100100005004277160300183003730037282717287411010020010008200200163003730037111020110099100100100001000000011171701600296460100001003003830038300383003830038
1020430037225000008429547251010010010000100100005004277160300183003730037282716287401010020010008200200163003730037111020110099100100100001000000011171701600296460100001003003830038300383003830038
10204300372250000010729547251010010010000100100005004277160300183003730037282716287401010020010008200200163003730037111020110099100100100001000000011171701600296450100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)031e3f4e51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)6061696b6d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa0a6a8a9accfd0d5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024300372600103295472510010101000010100005042771600530018030037300372828632876710010201000020200003003730037111002110910101000010001036400516652962910000103003830038300383003830038
10024300372590103295472510010101000010100005042771600030018030037300372828632876710010201000020200003003730037111002110910101000010001006405616652962910000103003830038300383003830038
1002430037260084295472510010101000010100005042771601030018030037300372828632876710010201000020200003003730037111002110910101000010000006406616652962910000103003830038300383003830038
10024300372600231295472510010101000010100005042771600630018030037300372828632876710010201000020200003003730037111002110910101000010000006400516562962910000103003830038300383003830038
10024300372410258295472510010101000010100005042771601030018030070300372828632876710010201000020200003003730037111002110910101000010000006406616562962910000103003830038300383003830038
1002430037241061295472510010101000010100005042771601630018030037300372828632876710010201000020200003003730037111002110910101000010000006406616652962910000103003830085301333003830038
10024300372410243295472510010101000010100005042771601630018030037300372828632876710010201000020200003003730037111002110910101000010000006406516552962910000103003830038300383003830038
10024300372420166295472510010101000010100005042771601030018030037300372828632876710010201000020200003003730037111002110910101000010001036406516562962910000103003830038300383003830038
10024300372330256295472510010101000010100005042771601630018030037300372828632876710010201000020200003003730037111002110910101000010000006400516652962910000103003830038300383003830038
10024300372330216295472510010101000010100005042771600030018030037300372828632876710010201000020200003003730037111002110910101000010000006400616652962910000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  movi v0.16b, 0
  suqadd v0.2s, v8.2s
  movi v1.16b, 0
  suqadd v1.2s, v8.2s
  movi v2.16b, 0
  suqadd v2.2s, v8.2s
  movi v3.16b, 0
  suqadd v3.2s, v8.2s
  movi v4.16b, 0
  suqadd v4.2s, v8.2s
  movi v5.16b, 0
  suqadd v5.2s, v8.2s
  movi v6.16b, 0
  suqadd v6.2s, v8.2s
  movi v7.16b, 0
  suqadd v7.2s, v8.2s
  movi v8.16b, 9

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.2508

retire (01)cycle (02)030a0b181e1f3f51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)696d6escheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa6a8acc2c5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1602042009115000000292580116100800161008002850064019620045200652006561280128200800282001600562006520065111602011009910010016000010000001111011901600200621600001002006620066200662006620066
16020420065150000006752580116100800161008002850064019620045200652006561280128200800282001600562006520065111602011009910010016000010000301111011901600200621600001002006620066200662006620066
16020420065151000001172580116100800161008002850064019620045200652006561280128200800282001600562006520065111602011009910010016000010000001111011901600200621600001002006620066200662013420066
160204200651500000011302580116100800161008002850064019620045200652006561280128200800282001600562006520065111602011009910010016000010000001111011901600200621600001002006620066200662006620066
1602042006515000000292580116100800161008002850064019620045200652006561280128200800282001600562006520065211602011009910010016000010000001111011901600200621600001002006620066200662006620066
16020420065150000005152580116100800161008002850064019620045200652006561280128200800282001600562006520065111602011009910010016000010000001111011901600200621600001002006620066200662006620066
16020420065151000001362580116100800161008002850064019620045200652006561280128200800282001600562006520065111602011009910010016000010000001111011901600200621600001002006620066200662006620066
1602042006515000000942580116100800161008002850064019620045200652006561280128200800282001600562006520065111602011009910010016000010000001111011901600200621600001002006620066200662006620066
16020420065150000001132580116100800161008002850064019620045200652006561280128200800282001600562006520065111602011009910010016000010000001111011901600200621600001002006620066200662006620066
16020420065150000008012580116100800161008002850064019620045200652006561280128200800282001600562006520065111602011009910010016000010000001111011901600200621600001002006620066200662006620066

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2506

retire (01)cycle (02)0318193a3f51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)5f60696d6escheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa8c2cfd0d2icache miss (d3)d5d6d9dadbddinst fetch restart (de)e0ea? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
16002420057150001292258001010800001080000506400001120027200462004632280010208000020160000200462004611160021109101016000010001003762120244125102004315160000102004720047200472004720051
16002420050150000198258001010800001080000506400000120031200502005032280010208000020160000200462005011160021109101016000010001003331110202116132004315160000102005120051200472005120047
160024200501500004882580010108000010800005064000011200272004620046322800102080000201600002004620046111600211091010160000100410041622524429592004315160000102005120051200512005120051
1600242005015000087258001010800001080000506400001120027200462004632280010208000020160000200462004611160021109101016000010201002862210242221052004715160000102004720047200472004720047
1600242004615000015425800101080000108000050640000012003120050200503228001020800002016000020050200501116002110910101600001000100306229244226142004730160000102005120051200512005120051
1600242005015000021125800101080000108000050640000012002720050200463228001020800002016000020050200501116002110910101600001000100293217202117122004315160000102004720047200472004720047
1600242004615000095258001010800001080000506400000120031200462004632280010208000020160000200502005011160021109101016000010001003532212244121492004730160000102004720051200512005120051
160024200501500001712580010108000010800005064000001200272005020046322800102080000201600002004620046111600211091010160000100010028311520211892004715160000102004720047200472004720047
16002420050150000452580010108000010800005064000011200312004620046322800102080000201600002005020046111600211091010160000100010030612520211992004315160000102004720047200512005120047
160024200461500004525800101080000108000050640000112002720046200463228001020800002016000020046200461116002110910101600001000100293119204215122004315160000102005120051200472004720047

Test 5: throughput

Count: 16

Code:

  suqadd v0.2s, v16.2s
  suqadd v1.2s, v16.2s
  suqadd v2.2s, v16.2s
  suqadd v3.2s, v16.2s
  suqadd v4.2s, v16.2s
  suqadd v5.2s, v16.2s
  suqadd v6.2s, v16.2s
  suqadd v7.2s, v16.2s
  suqadd v8.2s, v16.2s
  suqadd v9.2s, v16.2s
  suqadd v10.2s, v16.2s
  suqadd v11.2s, v16.2s
  suqadd v12.2s, v16.2s
  suqadd v13.2s, v16.2s
  suqadd v14.2s, v16.2s
  suqadd v15.2s, v16.2s
  movi v16.16b, 17

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.2502

retire (01)cycle (02)031e3f51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa6a8acc5branch mispredict (cb)cdcfd6inst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
160204400393000302516010810016039910016002050012801321400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000011110118160400361600001004004040040400404004040040
160204400393000302516010810016000810016002050012801321400204003940039199776199901601202001600322003200644003940039111602011009910010016000010023011110118160400361600001004004040040400404004040040
160204400393000302516010810016000810016002050012801321400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000011110118160400361600001004004040040400404004040040
160204400393000302516010810016000810016002050012801321400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000011110118160400361600001004004040040400404004040040
1602044003929901252516010810016000810016002050012801321400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000011110118160400361600001004004040040400404004040040
160204400393000302516010810016000810016002050012801321400204003940039199776199901601202021600322003200644003940039111602011009910010016000010001011110118160400361600001004004040040400404004040040
1602044003929902012516010810016000810016002050012801321400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000011110118160400361600001004004040040400404004040040
160204400393000302516010810016000810016002050012801321400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000011110118160400361600001004004040040400404004040040
160204400393000302516010810016000810016002050012801321400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000011110118160400361600001004004040040400404004040040
160204400393000302516010810016000810016002050012801321400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000011110118161400361600001004004040040400404004040040

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2502

retire (01)cycle (02)0307080a0b181e1f3f51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)5f6061696d6edispatch stall (70)72scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa1a6a8a9accfd0d2icache miss (d3)d5d6d9dadbddinst fetch restart (de)e0eaeb? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
160024400403001000000522516001010160000101600005012800001154002040039400391999603200191600102016000020320000400394003911160021109101016000010000001002283136162129440036206160000104004040040400404004040040
1600244003930000010004625160010101600001016000050128000011540020400394003919996027200191600102016000020320000400394003911160021109101016000010000001002283114162117640036206160000104004040040400404004040040
160024400393000000000462516001010160000101600005012800000154002040039400391999603200191600102016000020320000400394003911160021109101016000010000001002283151621161340036206160000104004040040400404004040040
16002440039300000000046251600101016000010160000501280000115400204003940039199960320019160010201600002032000040039400391116002110910101600001000000100248118162117540036206160000104004040040400404004040040
160024400393000000000462516001010160000101600005012800001154002040039400391999603200191600102016000020320838400394003911160021109101016000010000001009983161621114940036206160000104004040040400404004040040
16002440039300000000046251600101016000010160000501280000115400204003940039199960320019160010201600002032000040039400391116002110910101600001000000100228317162117640036206160000104004040040400404004040040
16002440039300010000046251600101016000010160000501280000115400204003940039199960320019160010201600002032000040039400391116002110910101600001000000100228217162116640036206160000104004040040400404004040040
1600244003930000000004625160010101600001016000050128000011540020400394003919996032001916001020160000203200004003940039111600211091010160000100000010022831121621171140036206160000104004040040400404004040040
1600244003930000000004625160010101600001016000050128000011540020400394003919996032001916001020160000203200004003940039111600211091010160000100010010022831131621113840036206160000104004040040400404004040040
160024400393000000000462516001010160000101600005012800001154002040039400391999603200191600102016000020320000400394003911160021109101016000010000001002211315162116640036206160000104004040040400404004040040