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, 4S)

Test 1: uops

Code:

  sqrdmulh v0.4s, v0.4s, 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)03191e1f3f4e51inst 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)a6accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
100430372301506125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
10043037230006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
10043037230006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372301506125482510001000100039831303018303730372415328951000100020003037303711100110000073116112627100030383038303830383038
10043037230006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
10043037230006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
10043037230006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372304506125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
10043037230006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372300010325482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038

Test 2: Latency 1->2

Code:

  sqrdmulh v0.4s, v0.4s, 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)0308091e3f4e51inst 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)9fa6a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102043003722500061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
102043003722400061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
102043003722500061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
102043003722500061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
102043003722501061295482510100100100001001000050042773131300183003730037282659287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
1020430037225000251295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
102043003722500061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
102043003722500061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
102043003722400061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
102043003722501061295482510100100100001001000050042773131300183008530037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03080b18191e1f3f4e51inst 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)9fa0a1a6a7a8acc2cdcfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100243003722500000061295482510010101000010100005042773133001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038
100243003722500000061295482510010101000010100005042773133001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038
100243003722500000061295482510010101000010100005042773133001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038
100243003722500000061295482510010101000010100005042773133001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038
100243003722500000061295482510010101000010100005042773133001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038
10024300372250000003853295482510010101000010100005042773133001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038
1002530037225000018061295482510010101000010100005042777993001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038
100243003722500000061295482510010101000010100005042773133001830037300372828732876710010201000020200003003730037111002110910101000010000003006402162229630010000103003830038300383003830038
100243003722500000061295482510010101000010100005042773133001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038
1002430037225000018061295482510010101000010100005042773133001830037300372828732876710010201000020200003003730037111002110910101000010000000006402162229630010000103003830038300383003830038

Test 3: Latency 1->3

Code:

  sqrdmulh v0.4s, v1.4s, 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)03080b191e1f3f4e51inst 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)9fa0a8acc5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372250000061295482510100100100001001000050042773133001830037300372827262874010100200100082002001630037300371110201100991001001000010000311171731600296460100001003003830038300383003830038
10204300372250000061295392510100100100001001000050042773133001830037300372827262874510100200100002002000030037300371110201100991001001000010000300071011611296340100001003003830038300383003830038
10204300372250000061295482510100100100001001000050042773133001830037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037225000006129548251010010010000100100005004277313300183003730037282653287451010020010000200200003003730037111020110099100100100001000017400071011611296340100001003003830038300383003830038
10204300372250000061295482510100100100001001000050042773133001830037300372826532874510100200100002002000030037300371110201100991001001000010000600071011611296340100001003003830038300383003830038
102043003722400090612954825101001001000010010000500427731330018300373003728265328745101002001000020020000300373003711102011009910010010000100001200071011611296340100001003003830038300383003830038
1020430037226000006129548251010010410000100100005004277313300183003730037282653287451010020010000200200003003730037111020110099100100100001000012300071011611296340100001003003830038300383003830038
10204300372250000061295482510100100100001001000050042773133001830037300372826532874510100200100002002000030037300371110201100991001001000010000600071011611296340100001003003830038300383003830038
10204300372250000061295482510100100100001001000050042773133001830037300372826532874510100200100002002000030037300371110201100991001001000010000300071011611296340100001003003830038300383003830038
10204300372250000061295482510100100100001001000050042773133001830037300372826532874510100200100002002000030037300371110201100991001001000010000300071011611296340100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)030708090a18191e1f3a3f4e51inst 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)9fa0a8acbranch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024300372240000000006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000306402162229630010000103003830038300383003830038
100243003722500000000061295482510010101000010100005042773133001830037300372828732876710010201000020200003003730037111002110910101000010008406402162229630010000103003830038300383003830038
1002430037225000000000612954825100101010000101000050427731330018300373003728287328767100102010000202000030037300371110021109101010000100018006402162229630010000103003830038300383003830038
10024300372250000000006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
10024300372250000000006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000306402162229630010000103003830038300383003830038
10024300372240000000006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000606402162229630010000103003830038300383003830038
10024300372250000000006129548251001012100001010000504277313300183003730037282873287671001220100002020000300373003711100211091010100001000306402163229630010000103003830038300383003830038
10024300372250000000006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000906401162229630010000103003830038300383003830038
10024300372250000000006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000306402162229630010000103003830038300383003830038
10024300372240000000006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000306402162229630010000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  sqrdmulh v0.4s, v8.4s, v9.s[1]
  sqrdmulh v1.4s, v8.4s, v9.s[1]
  sqrdmulh v2.4s, v8.4s, v9.s[1]
  sqrdmulh v3.4s, v8.4s, v9.s[1]
  sqrdmulh v4.4s, v8.4s, v9.s[1]
  sqrdmulh v5.4s, v8.4s, v9.s[1]
  sqrdmulh v6.4s, v8.4s, v9.s[1]
  sqrdmulh v7.4s, v8.4s, 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)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)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)9fa0a1a6a7a8acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
802042005815000000004125801001008000010080000500640000120020200392003999733999780100200800002001600002003920039118020110099100100800001000000000511021611200360800001002004020040200402004020040
802042003915000000002312580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100000022570511011611200360800001002004020040200402004020040
8020420039150000000041258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000000150511011611200360800001002004020040200402004020040
80204200391501000000412580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100000022660511011611200360800001002004020040200402004020040
802042003915000000004125801001008000010080000500640000020020200392003999733999780100200800002001600002003920039118020110099100100800001000000000511011611200360800001002004020040200402004020040
802042003915000000004125801001008000010080000500640000020020200392003999733999780100200800002001600002003920039118020110099100100800001000000000511011611200360800001002004020040200402004020040
802042003915000000004125801001008000010080000500640000020020200392003999733999780100200800002001600002003920039118020110099100100800001000000000511011611200360800001002004020040200402004020040
802042003914900000004125801001008000010080000500640000020020200392003999733999780100200800002001600002003920039118020110099100100800001000000000511011611200360800001002004020040200402004020040
802042003915000000004125801001008000010080000500640000020020200392003999733999780100200800002001600002003920039118020110099100100800001000000000511011611200360800001002004020040200402004020040
802042003915000000004125801001008000010080000500640000020020200392003999733999780100200800002001600002003920039118020110099100100800001000000030511011611200360800001002004020040200402004020040

1000 unrolls and 10 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)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)9fa8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
80024200481500004025800101080000108000050640000020020020039200399996310019800102080000201600002003920039118002110910108000010005020416232003680000102004020040200402004020040
80024200391500004025800101080000108000050640000020020020039200399996310019800102080000201600002003920039118002110910108000010005020416442003680000102004020040200402004020040
8002420039150001240258001010800001080000506400000200200200902012910005310019800102080000201600002003920039118002110910108000010005054416432003680000102004020040200402004020040
80024200391500004025800101080000108000050640000020020020039200399996310019800102080000201600002003920039118002110910108000010005020516552003680000102004020040200402004020040
80024200391500004025800101080000108000050640000020020020039200399996310019800102080000201600002003920039118002110910108000010005020316452003680000102004020040200402004020040
80024200391500004025800101080000108000050640000020024020039200399996310019800102080000201600002003920039118002110910108000010005020316442003680000102004020040200402004020040
80024200391500004025800101080000108000050640000020020020039200399996310019800102080000201600002003920039118002110910108000010035020416332003680000102004020040200402004020040
80024200391500004025800101080000108000050640000020020020039200399996310019800102080000201600002003920039118002110910108000010005020316352003680000102004020040200402004020040
80024200391500004025800101080000108000050640000020020020039200399996310019800102080000201600002003920039118002110910108000010005020416342003680000102004020040200402004020040
80024200391500004025800101080000108000050640000020020020039200399996310019800102080000201600002003920039118002110910108000010005020416442003680000102004020040200402004020040