Apple M1 Microarchitecture Research by Dougall Johnson

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

SQRDMULH (by element, S)

Test 1: uops

Code:

  sqrdmulh s0, s0, v1.s[1]
  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)03070a0b1e3a3f4e51inst issue (52)~issue fp/simd (54)~dispatch fp/simd (57)huge thing fp/simd (5b)696d6edispatch 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
1004303723110021812548251000100010003983133018303730372415328951000100020003037303711100110000077416442630100030383038303830383038
100430372311002672548251000100010003983133018303730372415328951000100020003037303711100110000077416442630100030383038303830383038
100430372211002672548251000100010003983133018303730372415328951000100020003037303711100110000077416442630100030383038303830383038
100430372311002672548251000100010003983133018303730372415328951000100020003037303711100110000077416442630100030383038303830383038
100430372211002672539251000100010003983133018303730372415328951000100020003037303711100110000077416442630100030383038303830383038
100430372311002672548251000100010003983133018303730372415328951000100020003037303711100110000077416442630100030383038303830383038
100430372311002672548251000100010003983133018303730372415328951000100020003037303711100110000077416442630100030383038303830383038
1004303723110022572548251000100010003983133018303730372415328951000100020003037303711100110000077416442630100030383038303830383038
100430372311002672548251000100010003983133018303730372415328951000100020003037303711100110000077416442630100030383038303830383038
100430372211002672548251000100010003983133018303730372415328951000100020003037303711100110000077416442630100030383038303830383038

Test 2: Latency 1->2

Code:

  sqrdmulh s0, s0, v1.s[1]
  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)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)9fa0a1a6a7a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372250000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000000071011611296340100001003003830038300383003830038
102043003722500000000103295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000000071011611296340100001003003830038300383003830038
10204300372250000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000000071011611296340100001003003830038300383003830038
10204300372250000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000000071011611296340100001003003830038300383003830038
10204300372250000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000000071011611296340100001003003830038300383003830069
10204300372250000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000000071011611296700100001003003830038300383003830038
10204300372250000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000000071011611296340100001003003830038300383003830038
102043003722500000000612948516110196140100481401104370642854551302703037130320282923328864105612201115122421990303723008481102011009910010010000100000010221484867383352988632100001003037330359303693037130363
10204303722280117710566160455829485162102031391005614111043734428681213030630369300852830237288671118422210990224200003003730037111020110099100100100001000000000279915623298507100001003037830323303223037430180
1020430276227111678070061295482510100100100321201000054442773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000000071011611296340100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)030b1e1f3f4e51inst 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)9fa8accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037225000612954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010006403162229630010000103003830038300383003830038
1002430037225000612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010006402162229630010000103003830038300383003830038
1002430037225000612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010006402162229630010000103003830038300383003830038
1002430037225060612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010006402162229630010000103003830038300383003830038
1002430037225000612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010006402162229630010000103003830038300383003830038
1002430037225000612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010006402162229630010000103003830038300383003830038
1002430037224000612954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010006402162229630010000103003830038300383003830038
100243003722500061295482510010101000010100005042773131300183003730037282873287671001020100002020000300373003711100211091010100001002016402162229630010000103003830038300383003830038
10024300372250006129548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100186402162229630010000103003830038300383003830038
1002430037225000612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010006402162229630010000103003830038300383003830038

Test 3: Latency 1->3

Code:

  sqrdmulh s0, s1, v0.s[1]
  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)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)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)9fa8accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020430085225001001060295484510100100100001181000052242773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000071014811296340100001003003830038300383003830038
10204300372250000061295482510100100100001001000055542773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000371011611296340100001003003830038300383003830038
102043003722500012061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001001840171011611296340100001003003830038300383022830038
102043003722400012061295482510100100100001001000050042773130300183003730037282653287451010020010000200209983003730037111020110099100100100001001371011611296340100001003003830038300383003830038
102043003722500012061295092510100100100001001000050042773130300183003730037282653287451010020010000200200003003730228111020110099100100100001000071011611296660100001003018330038300383003830087
10204300372251109352822954825101001001000010010000500427731303001830037300372826518287451010020010000200200003003730037111020110099100100100001000071011611296340100001003003830038300383003830038
10204300372250001217682295482510100100100001041000050042773130300183003730037282653287451010020010000200209903003730037111020110099100100100001001071011611296340100001003003830038300383003830038
1020430037225000002736295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000671011611296340100001003003830038300383003830038
102043003722500012061295482510100100100001001044750042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001001071011611296340100001003003830038300383003830134
10204300372250004201032954825101001001000010010000500427731303001830169300372826532874510100200100002002000030037300371110201100991001001000010002771011613296340100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03191e3f4e51inst 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)9fa8acbranch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037224006129548251001010100001010000504277313130018030037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313130018030037300372828732876710010201000020200003003730037111002110910101000010100640216222963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313130018030037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313130018030037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103008430038300383003830038
1002430037225006129548251001010100001010000504277313130018030037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313130018030037300372828732876710010201000020200003003730037111002110910101000010430640216222963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313130018030037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313130018030037300372828732876710010201000020200003003730037111002110910101000010100640216222963010000103003830038300383003830038
1002430037224006129548251001010100001010000504277313130018030037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313130018030037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  sqrdmulh s0, s8, v9.s[1]
  sqrdmulh s1, s8, v9.s[1]
  sqrdmulh s2, s8, v9.s[1]
  sqrdmulh s3, s8, v9.s[1]
  sqrdmulh s4, s8, v9.s[1]
  sqrdmulh s5, s8, v9.s[1]
  sqrdmulh s6, s8, v9.s[1]
  sqrdmulh s7, s8, v9.s[1]
  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)030708090a0b18191e1f3a3f51inst 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)9fa8acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8020420058150100100000248258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000051141016410200360800001002004020040200402004020040
80204200391501001000002482580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100000511491649200360800001002004020040200402004020040
80204200391501001000002482580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100000511491699200360800001002004020040200402004020040
8020420039150100100000248258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000051141116910200360800001002004020040200402004020040
80204200391501001000002712580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100000518971677200360800001002004020040200402004020040
802042003915010010000024825801001008000010080000500640000020020200392003999733999780100200800002001600002003920039118020110099100100800001000005114616910200360800001002004020040200402004020040
8020420039150100100000248258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000051149161210200360800001002004020040200402004020040
80204200391501001000002482580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100000511491677200360800001002004020040200402004020040
80204200391501001000002482580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100000511491697200360800001002004020040200402004020040
802042003915010010000024825801001008000010080000500640000020020200392003999733999780100200800002001600002003920039118020110099100100800001000005114101699200360800001002004020040200402004020040

1000 unrolls and 10 iterations

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

retire (01)cycle (02)030918191e1f3a3f51inst 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)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)9fa1a6a8a9accfd0d5d6d9daddinst fetch restart (de)e0? int output thing (e9)ea? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420048150000000402580010108000011800005064000000200202003920039999631001980010208000020160000200392003911800211091010800001000000502004160067200360080000102004020040200402004020040
8002420039150000000402580010108000010800005064000000200202003920039999631001980010208000020160000200392003911800211091010800001000000502007160044200360080000102004020040200402004020040
8002420039150000000402580010108000010800005064000000200202003920039999631001980010208000020160000200392003911800211091010800001000000502006160043200360080000102004020040200402004020040
80024200391500000002514880109128009711801066164077600200622009120093100046100478001020800002016000020039200391180021109101080000104002475503704360034200870080000102004020040200402004020040
80024200391500001200822580010108000010800005064000000200202003920039999631001980010208000020160000200392011211800211091010800001000000502003160044200360080000102004020040200402004020040
8002420039150000000402580010108000010800005064000000200202003920039999631001980010208000020160000200392003911800211091010800001000000502004160044200360080000102004020040200402004020040
8002420039150000000402580010108000010800005064000000200202003920039999631001980010208000020160000200392003911800211091010800001000000502003160066200360080000102004020040200402004020040
80024200391500000008952580010108000010800005064000000200202003920039999631001980010208000020160000200392003911800211091010800001000000502003160066200360080000102004020040200402004020040
8002420039150000000402580010108000010800005064000000200202003920039999631001980010208000020160000200392003911800211091010800001000000502004160076200360080000102004020040200402004020040
8002420039150000000402580010108000010800005064000000200202003920039999631001980010208000020160000200392003911800211091010800001000000502003160044200360080000102004020040200402004020040