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, 2D)

Test 1: uops

Code:

  suqadd v0.2d, v1.2d
  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)031e3f4e51inst 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
10043037230612548251000100010003983130301830373037241532895100010002000303730371110011000017173116112630100030383038303830383038
10043037231261254825100010001000398313030183037303724153289510001000200030373037111001100001273116112630100030383038303830383038
1004303723061254825100010001000398313030183037303724153289510001000200030373037111001100006673116112630100030383038303830383038
1004303723061254825100010001000398313030183037303724153289510001000200030373037111001100006973116112630100030383038303830383038
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110000973116112630100030383038303830383038
1004303723061254825100010001000398313030183037303724153289510001000200030373037111001100001273116112630100030383038303830383038
1004303722061254825100010001000398313030183037303724153289510001000200030373037111001100001273116112630100030383038303830383038
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038

Test 2: Latency 1->1

Code:

  suqadd v0.2d, v1.2d
  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)033f4e51inst 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)9fa0a1a6a8a9cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102043003722576329548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000007101161129634100001003003830038300383003830038
102043003722587829548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000007101161129634100001003003830038300383003830038
102043003722591429548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000547101161129634100001003003830038300383003830038
1020430037225151229548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000007101161129634100001003003830038300383003830038
102043003722540729548251010010010000100100005004277313130018300373003728265328745101002001000020020000300373003711102011009910010010000100000007101161129634100001003003830038300383003830038
1020430037225107329548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000007101161129634100001003003830038300383003830038
1020430037225149129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000007101161129634100001003003830038300383003830038
1020430037225137329548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000007101161129634100001003003830038300383003830038
102043003722584029548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000007101161129634100001003003830038300383003830038
1020430037224151629548251010010010000100100005004277313130018300373003728265328745101002001000020020000300373003711102011009910010010000100000007101161129634100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)030818191e1f3f4e51inst 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)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)9fa0a1a8acc2branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100243003722500000612954825100101010000101000050427731313001830037300372828703287671001020100002020000300373003711100211091010100001000130006402162229630010000103003830038300383003830038
100243003722500000612954825100101010000121000050427867003001830037300372828703287671001020100002020000300373003711100211091010100001000030550006402162229630010000103003830038300383003830038
10024300372250000040529548251001010100001010000504277313130018300373003728287032876710010201000020200003003730037111002110910101000010002960006402162229630010000103003830038300383003830038
10024300372250000072629548251001010100001010000504277313030018300373003728287032876710010201000020200003003730037111002110910101000010002460006402162229630010000103003830038300383003830038
1002430037225000006129548251001010100001010000504277313030018300373003728287032876710010201000020200003003730037111002110910101000010002960006402162229630010000103003830038300383003830038
1002430037225000008429548251001010100001010000504277313130018300373003728287032876710010201000020200003003730037111002110910101000010002660106402162229630010000103003830038300383003830038
100243003722500000612954825100101010000101000050427731303001830037300372828703287671001020100002020000300373003711100211091010100001000034006402162229630010000103003830038300383003830038
1002430037225000006129548251001010100001010000504277313130018300373003728287032876710010201000020200003003730037111002110910101000010008800006402162229630010000103003830038300383003830038
10024300372240001206129548251001010100001010000504277313130018300373003728287032876710010201000020200003003730037111002110910101000010002900006402162229630010000103003830038300383003830038
1002430037225000006129548251001010100001010000504277313030018300373003728287032876710010201000020200003003730037111002110910101000010002890006402162229630010000103003830038300383003830038

Test 3: Latency 1->2

Code:

  suqadd v0.2d, v0.2d
  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)0308090b18191e1f3f4e51inst 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)9fa0a1a6a8a9acc5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102043003722500000001103295472510100100100001001000050042771603001830037300372827172874110100200100082002001630037300371110201100991001001000010000000011171701600296450100001003003830038300383003830038
10204300372250000000187295382510100100100001001000050042771603001830037300372827172874110100200100002002000030037300375110201100991001001000010000000011172222422296290100001003003830038300383003830038
10204300372250100000164295472510100100100001001000050042771603001830037300372825262873310100200100002002132430037300371110201100991001001000010000000011172222422296290100001003003830038300383003830038
10204300372250000000204295472510100100100001001000050042771603001830037300372825262873310100200100002002000030037300371110201100991001001000010000000011172222422296290100001003003830038300383003830038
10204300372250000000204295472510100100100001001000050042771603001830037300372825262873310100200100002002000030037300371110201100991001001000010000000284311172222422296290100001003003830038300383003830038
10204300372250000000181295472510100100100001001000050042771603001830037300372827562873310100200100002002000030037300371110201100991001001000010000000011172222422296290100001003003830038300383003830038
10204300372250000000204295472510100100100001001000050042771603001830037300372825262873310100200100002002059030084300371110201100991001001000010000000011172222422296290100001003003830038300383003830038
10204300372250000000204295292510100100100001001000050042771603001830037300372825262873310100204100002002000030037300371110201100991001001000010000000011172222422296290100001003003830038300383003830085
1020430037225000000097295472510100100100001001000050042771603009030037300372825262873310100200100002002000030037300371110201100991001001000010000410011172222422296290100001003003830038300383003830038
10204300372250000060139295102510100100100001001000050042771603001830037300372825262873310100200100002002000030037300371110201100991001001000010000000011179522422296290100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)0307080b18191e1f3f4e51inst 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)9fa0a1a6a7a8a9acc2branch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100243003722500000001032954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000000000640316332962910000103003830038300383003830038
1002430037225000001080822954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000000000640316332962910000103003830038300383003830038
10024300372240000000842954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000000300640316332962910000103003830038300383003830038
10024300372250000000822954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000010001640316332962910000103003830038300383003830038
10024300372250000000842954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000000000642316342962910000103003830038300383003830038
1002430037225100001501262954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000000000640316332962910000103003830038300383003830038
100243003722500000001052954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000000000640316332962910000103003830038300383003830038
10024300372250000000842954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000000000640316332962910000103003830038300383003830038
1002430037225000000020402954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000000000640316332962910000103003830038300383003830038
10024300372250000000842954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000000000640316332962910000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  movi v0.16b, 0
  suqadd v0.2d, v8.2d
  movi v1.16b, 0
  suqadd v1.2d, v8.2d
  movi v2.16b, 0
  suqadd v2.2d, v8.2d
  movi v3.16b, 0
  suqadd v3.2d, v8.2d
  movi v4.16b, 0
  suqadd v4.2d, v8.2d
  movi v5.16b, 0
  suqadd v5.2d, v8.2d
  movi v6.16b, 0
  suqadd v6.2d, v8.2d
  movi v7.16b, 0
  suqadd v7.2d, v8.2d
  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)03080b18191e1f3a3f51inst 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)60696d6escheduler 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)9fa0a1a6a7a8a9acc2c5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
160204200891500000000712580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000000000011110119116002006201600001002006620066200662006620066
160204200651500000000292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000000000011110119016002006201600001002006620066200662006620066
160204200651500000088029258011610080016100800285006401961200452006520065612801282008002820016005620065200651116020110099100100160000100000010983011110119016002006201600001002006620066200662006620066
160204200651500000000292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000000000011110119016002006201600001002006620066200662006620066
160204200651510000000292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000000000011110119016002006201600001002006620066200662006620066
160204200651510000000292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000000000011110119016002006201600001002006620066200662006620066
1602042006515000001200292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000000000011110119016002006201600001002006620066200662006620066
160204200651500000000712580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000000000011110119016002006201600001002006620066200662006620066
160204200651500000000292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000000000011110119016002006201600001002006620066200662006620066
160204200651500000000292580116100800161008002850064188412004520065202256128012820080028200160056200652006511160201100991001001600001000000103011110119016002006201600001002006620066200662006620066

1000 unrolls and 10 iterations

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

retire (01)cycle (02)033a3f51inst 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)5f6061696d6escheduler 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)9fa6a8a9acbranch mispredict (cb)cfd0d2icache miss (d3)d5d6d9dadbddinst fetch restart (de)e0ea? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1600242006815001292580010108000010800005064000010102003120046200463228001020800002016000020046200461116002110910101600001000000100483212320433115202004315160000102004720047200472004720047
160024200501500452580010108000010800005064000011020027200462005032280010208000020160000200462004611160021109101016000010000001004413112024333119172004315160000102004720047200472004720047
1600242004615004525800101080000108000050640000111020027200462004632280010208000020160000200462004611160021109101016000010000001004313222322364121182004315160000102004720047200472004720051
160024200461500873258001010800001080000506400000002002720046200463228001020800002016000020046200501116002110910101600001000000100433112120313117172004715160000102005120047200472004720047
160024200461500177258001010800001080000506400001102002720046200503228001020800002016000020050200461116002110910101600001000000100353122020324119192004715160000102004720047200472004720047
1600242004615003482580010108000010800005064000011020027200462005032280010208000020160000200462005011160021109101016000010300001004113622120323118172004715160000102005120047200512005120047
1600242004615101292580010108000010800005064000010102002720046200463228001020800002016000020046200501116002110910101600001000000100393111624346117172004330160000102004720047200512005120047
160024200461500164258001010800001080000506400001102002720046200463228001020800002016000020046200461116002110910101600001000000100413521820366118172004315160000102004720047200512004720047
1600242004615001082580010108000010800005064000011020027200462004632280010208000020160000200462005011160021109101016000010000001004413512020356220172004730160000102005120047200512004720051
16002420050150068258001010800001080000506400001102002720050200463228001020800002016000020046200461116002110910101600001000000100413112020286118192004730160000102004720051200472004720047

Test 5: throughput

Count: 16

Code:

  suqadd v0.2d, v16.2d
  suqadd v1.2d, v16.2d
  suqadd v2.2d, v16.2d
  suqadd v3.2d, v16.2d
  suqadd v4.2d, v16.2d
  suqadd v5.2d, v16.2d
  suqadd v6.2d, v16.2d
  suqadd v7.2d, v16.2d
  suqadd v8.2d, v16.2d
  suqadd v9.2d, v16.2d
  suqadd v10.2d, v16.2d
  suqadd v11.2d, v16.2d
  suqadd v12.2d, v16.2d
  suqadd v13.2d, v16.2d
  suqadd v14.2d, v16.2d
  suqadd v15.2d, v16.2d
  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)60696b6d6edispatch 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)9facc5branch mispredict (cb)cdcfd5d6e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
16020440039299030251601081001600081001600205001280132140020040039400391997761999016012020016003220032006440039400391116020110099100100160000100011110118016400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132140020040039400391997761999016012020016003220032006440039400391116020110099100100160000100011110118016400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132040020040039400391997761999016012020016003220032006440039400391116020110099100100160000100011110118016400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132140020040039400391997761999016012020016003220032006440039400391116020110099100100160000100011110118016400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132140020040039400391997761999016012020016003220032006440039400391116020110099100100160000100011110118016400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132140020040039400391997761999016012020016003220032006440039400391116020110099100100160000100011110118016400361600001004004040040400404004040040
16020440039299030251601081001600081001600205001280132040020040039400391997761999016012020016003220032006440039400391116020110099100100160000100011110118016400361600001004004040040400404004040040
160204400393000179251601081001600081001600205001280132140020040039400391997761999016012020016003220032006440039400391116020110099100100160000100011110118016400361600001004004040040400404004040040
16020440039299030251601081001600081001600205001280132140020040039400391997761999016012020016003220032006440039400391116020110099100100160000100011110118016400361600001004004040040400404004040040
16020440039299030251601081001600081001600205001280132140020040039400391997761999016012020016003220032006440039400391116020110099100100160000100011110118016400361600001004004040040400404004040040

1000 unrolls and 10 iterations

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

retire (01)cycle (02)0308091e3f51inst 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)5f60696d6edispatch 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)9fa6a8acc2cfd0d2icache miss (d3)d5d6d9dadbddinst fetch restart (de)e0? int output thing (e9)eaeb? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1600244005030000046251600101016000010160000501280000114002040039400391999603200191600102016000020320000400394003911160021109101016000010000010022313214162111213400360155160000104004040040400404004040040
1600244003929900046251600201016000010160000501280000114002040039400391999603200191600102016000020320000400394003911160021109101016000010000010022312715162111414400360155160000104004040040400404004040040
16002440039300000882516001010160000101600005012800001140020400394003919996032001916001020160106203200004003940039111600211091010160000100093010022312716162111414400360155160000104004040040400404004040040
1600244003929900046251600101016000010160000501280000114002040039400391999603200191600102016000020320000400394003911160021109101016000010000010022313014162111414400360155160000104004040040400404004040040
1600244003929900046251600101016000010160000501280000114002040039400391999603200191600102016000020320000400394003911160021109101016000010000010022312612162111211400360155160000104004040040400404004040040
1600244003930000046251600101016000010160000501280000114002040039400391999603200191600102016000020320000400394003911160021109101016000010000010022312813162111314400360155160000104004040040400404004040088
1600244003930000046251600101016000010160000501280000114002040039400391999603200191600102016000020320000400394003911160021109101016000010000010022312612162111112400360155160000104004040040400404004040040
1600244003929900046251600101016000010160000501280000114002040039400391999603200191600102016000020320000400394003911160021109101016000010000010022313115162111413400360155160000104004040040400404004040040
1600244003930000046251600101016000010160000501280000114002040039400391999603200191600102016000020320000400394003911160021109101016000010000010022312512162111311400360155160000104004040040400404004040040
1600244003930000046251600101016000010160000501280000114002040039400391999603200191600102016000020320000400394003911160021109101016000010000010022312713162111214400360155160000104004040040400404004040040