Apple M1 Microarchitecture Research by Dougall Johnson

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

FABD (vector, 4H)

Test 1: uops

Code:

  fabd v0.4h, v0.4h, v1.4h
  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
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372206125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110001073116112630100030383038303830383038
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110002973116112630100030383038303830383038
100430372206125482510001000100039831303018303730372415328951000100020003037303711100110000673116112630100030383038303830383038
100430372206125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372206125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723061254825100010001000398313030183037303724153289510001000200030373037111001100001273116112630100030383038303830383038
100430372206125482510001000100039831303018303730372415328951000100020003037303711100110001073116112630100030383038303830383038

Test 2: Latency 1->2

Code:

  fabd v0.4h, v0.4h, v1.4h
  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)03080b1e3f4e51inst 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)9fa0a1a6a8a9accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102043003722500061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722500061295482510100100100001121000050042786701301983032430037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722500061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722500061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722500061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722500089295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
1020430037225003361295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722500061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000100007101161129634100001003003830038300383003830038
102043003722500061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722500961295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)0308090b18191e1f3a3f4e51inst 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)9fa0a1a6a7a8a9acc2c5cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024300372250000012006129548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100000000006402162229630010000103003830038300383003830038
1002430037225000000006129548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100000000006402162229630010000103003830038300383003830038
100243003722500034162006129548251001010100001010000664277313030018300373003728287328767100102010161202000030084300371110021109101010000100000000006402162229630010000103003830038300383003830038
1002430037225000000006129548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100000000006402162229630010000103003830038300383003830038
10024300372250000000044129548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100000000006402162229630010000103003830038300383003830038
1002430037225000000006129548251001010100001010000504277797030018300373003728287328786100102010000202000030037300371110021109101010000100000000006402162229630010000103003830038300383003830038
1002430037224000000006129548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100000000006402162229630010000103003830038300383003830038
1002430037225000000006129548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100000000006402162229630010000103003830038300383003830038
1002430037225000000006129548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100000000006402162229630010000103003830038300383003830038
1002430037225000000006129548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100000100006402162229630010000103013230038300383003830038

Test 3: Latency 1->3

Code:

  fabd v0.4h, v1.4h, v0.4h
  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)030918191e1f3f4e51inst 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)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)9fa0a1a8acc5cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372250007206129548251010010010000100100005004277313130018030037300372826503287451010020010000200200003003730037111020110099100100100001000000007421161129634100001003003830038300383003830038
10204300852250002106129548251010010010000100100005004277313030090030037300372827803287451010020010000200200003003730037111020110099100100100001002210007101161129634100001003003830038300383003830170
1020430037225100006129548251010010010000100102985974277313030018030086300852826503287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722500000760295482510100100100001001000050042773130300180300373003728265017287631010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
1020430037224000006129548251010010010000100100005004277313130018030037300372826503287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
1020430037225000006129548251010010010000100100005004277313030018030037300372826503287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
1020430037225000006129548251010010010000100100005004277313030018030037300372826503287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
1020430037224000006129548251010010010000100100005004277313030018030037300372826503287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
1020430037225000006129548251010010010000100100005004277313030018030037300372826503287451010020010000200200003003730037111020110099100100100001000000107101161129634100001003003830038300383003830038
1020430037225000006129548251010010010000100100005004277313030018030037300372826503287451010020010000200200003003730037111020110099100100100001000000117101161129634100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)031e3a3f4e51inst 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)9fa8cfd0d5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037225006129548251001010100001010000504277313103006403003730037282873287671001020100002020000300373003711100211091010100001006400316332963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313003001803003730037282873287671001020100002020000300373003711100211091010100001006400316332963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313003001803003730037282873287671001020100002020000300373003711100211091010100001006400316342963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313003001803003730037282873287671001020100002020000300373003711100211091010100001006400316332963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313003001803003730037282877287671001020100002020000300373003711100211091010100001006400316332963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313003001803003730037282873287671001020100002020000300373003711100211091010100001006400316332963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313083001803003730037282873287671001020100002020000300373003711100211091010100001006400316332963010000103003830038300383003830038
1002430037224006129548251001010100001010000504277313003001803003730037282873287671001020100002020000300373003711100211091010100001006400316332963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313003001803003730037282873287671001020100002020000300373003711100211091010100001006400341332963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313003001803003730037282873287671001020100002020000300373003711100211091010100001006400316332963010000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  fabd v0.4h, v8.4h, v9.4h
  fabd v1.4h, v8.4h, v9.4h
  fabd v2.4h, v8.4h, v9.4h
  fabd v3.4h, v8.4h, v9.4h
  fabd v4.4h, v8.4h, v9.4h
  fabd v5.4h, v8.4h, v9.4h
  fabd v6.4h, v8.4h, v9.4h
  fabd v7.4h, v8.4h, v9.4h
  movi v8.16b, 9
  movi v9.16b, 10

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.2505

retire (01)cycle (02)03080b1e3f51inst 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)9fa1a6a8a9accdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8020420039150002741258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000000051102161120036800001002004020040200402004020040
802042003915000041258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000000051101161120036800001002004020040200402004020040
802042003915000041258010010080000100800005006400001200202003920039997339997801002008000020016000020039200391180201100991001008000010000000051101161120036800001002004020040200402004020040
802042003915000041258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000000051101161120036800001002004020040200402004020040
802042003915000041258010010080000100800005006400001200202003920039997339997801002008000020016000020039200391180201100991001008000010000000051101161120036800001002004020040200402004020040
80204200391500027341258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000000051101161120036800001002004020040200402004020040
802042003915000041258010010080000100800005006400001200202003920039997339997801002008000020016000020039200391180201100991001008000010001000051101161120036800001002004020040200402004020040
802042003915000041258010010080000100800005006400001200202003920039997339997801002008000020016000020039200391180201100991001008000010000000051101161120036800001002004020040200402004020040
802042003915100041258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000000051101161120036800001002004020040200402004020040
8020420039150002141258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000000051101161120036800001002004020040200402004020040

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2505

retire (01)cycle (02)03090a18191e1f3a3f51inst 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)9faccfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420048150000000040258001010800001080000506400001200202003920039999631001980010208000020160000200392003911800211091010800001005020191610162003680000102004020040200402004020040
8002420039150000000040258001010800001080000506400001200202003920039999631001980010208000020160000200392003911800211091010800001035020161610192003680000102004020040200402004020040
8002420039150000000040258001010800001080000506400001200202003920039999631001980010208000020160000200392003911800211091010800001005020101616162003680000102004020040200402004020040
8002420039150000000040258001010800001080000506400001200202003920039999631001980010208000020160000200392003911800211091010800001005020161616162003680000102004020040200402004020040
80024200391500000210040258001010800001080000506400001200202003920039999631001980010208000020160000200392003911800211091010800001005020101616102003680000102004020040200402004020040
8002420039150000000040258001010800001080000506400001200202003920039999631001980010208000020160000200392003911800211091010800001005020191610162003680000102004020040200402004020040
8002420039150000000040258001010800001080000506400001200202003920039999631001980010208000020160000200392003911800211091010800001005020161610202003680000102004020040200402004020040
8002420039150000000040258001010800001080000506400001200202003920039999631001980010208000020160000200392003911800211091010800001005020191610202003680000102004020040200402004020040
800242003915000000004025800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100502016169152003680000102004020040200402004020040
8002420039150000000040258001010800001080000506400001200202003920039999631001980010208000020160000200392003911800211091010800001005020161616162003680000102004020040200402004020088