Apple M1 Microarchitecture Research by Dougall Johnson

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

UQSUB (vector, 2D)

Test 1: uops

Code:

  uqsub v0.2d, 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)03071e3f4e51inst 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
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110000373216222630100030383038303830383038
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110000073216222630100030383038303830383038
1004303723006125482510001000100039831303018303730372415328951000100020003037303711100110000073216222630100030383038303830383038
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110000073216222630100030383038303830383038
10043037221910525482510001000100039831313018303730372415328951000100020003037303711100110000073216222630100030383038303830383038
1004303722006125482510001000100039831313018303730372415328951000100020003037303711100110000073216222630100030383038303830383038
1004303722006125482510001000100039831313018303730372415328951000100020003037303711100110000073216222630100030383038303830383038
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110000073216222630100030383038303830383038
10043037230025125482510001000100039831313018303730372415328951000100020003037303711100110000073216222630100030863038303830383038
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110000073216222630100030383038303830383038

Test 2: Latency 1->2

Code:

  uqsub v0.2d, 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)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)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)9fa0a1a6a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372250000000319295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000000000710116112963425100001003003830038300383003830038
1020430037225000000083729548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000000071011611296340100001003003830038300383003830038
1020430037225000000045129548251010010010000100100005004277313130018300373003728265328745101002001000020020000300373003711102011009910010010000100000000071011611296340100001003003830038300383003830038
1020430037225000000019129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000000071011611296340100001003003830038300383003830038
10204300372250000000105229548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000000071011611296340100001003003830038300383003830038
10204300372250000000455295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000000000710116112963424100001003003830038300383003830038
10204300372250000000150629548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000000071011611296340100001003003830038300383003830038
1020430037225010000019129548251010010010000100100005004277313030018300373008328265328745101002001000020020000300373003711102011009910010010000100000000071011611296340100001003003830038300383003830038
1020430037225000000021029548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000000071011611296340100001003003830038300383003830038
1020430037224000000012429530251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000000071011611296340100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)031e1f3f4e51inst 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)9fa8a9accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037225120149729548251001010100001010000504277313130018300373003728287328767100102010000202000030037300371110021109101010000101036402162229630010000103003830038300383003830038
10024300372250067329548251001010100001010000504277313130018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313130018300373003728287328767100102010000202000030037300371110021109101010000100036402162229630010000103003830038300383003830038
1002430037225006129548251001010100001010000504277313130018302283003728287328767100102010000202000030037300371110021109101010000101206402162229630010000103003830038300383003830038
10024300372250088129548251001010100001010000504277313130018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038
1002430037225006129548251001010100081010000504277313130018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250084929548251001010100001010000504277313130018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038
1002430037224006129548251001010100001010000504277313130018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250098229548251001010100001010000504277313130018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250090629548251001010100001010000504277313130018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038

Test 3: Latency 1->3

Code:

  uqsub v0.2d, v1.2d, v0.2d
  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)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)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)9fa0a1a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372250612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
10204300372249612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
1020430037224507612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
10204300372250612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
1020430037224390612954825101001001000010010000500427731313001830037300372826532874510100200100002002000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
10204300372250612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
1020430037225531612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
1020430037225528612954825101001001000010010000500427731313001830037300372826532874510100200100002002000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
10204300372250612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
102043003722536612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038

1000 unrolls and 10 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)9fa8accfd5d6daddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037233612954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010006402160222963010000103003830038300383003830038
1002430037225612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010006402160222963010000103003830038300383003830038
1002430037224612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010006402160222963010000103003830038300383003830038
10024300372252082954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010106402160222963010000103003830038300383003830038
1002430037225612954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010006402160222963010000103003830038300383003830038
1002430037224612954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010006402160222963010000103003830038300383003830038
10024300372251452954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010006402160222963010000103003830038300383003830038
10024300372253312954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010006402160222963010000103003830038300383003830038
10024300372241282954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010006403160222963010000103003830038300383003830038
10024300372251682954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010006402160222963010000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  uqsub v0.2d, v8.2d, v9.2d
  uqsub v1.2d, v8.2d, v9.2d
  uqsub v2.2d, v8.2d, v9.2d
  uqsub v3.2d, v8.2d, v9.2d
  uqsub v4.2d, v8.2d, v9.2d
  uqsub v5.2d, v8.2d, v9.2d
  uqsub v6.2d, v8.2d, v9.2d
  uqsub v7.2d, v8.2d, v9.2d
  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)03080b18191e3f51inst 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)9fa8a9acc2cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
802042005815000005441258010010080000100800005006400002002020039200399973399978010020080000200160000200392003911802011009910010080000100000051102161120036800001002004020040200402004020040
80204200391500000041258010010080000100800005006400002002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040
80204200391500000041258010010080000100800005006400002002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040
80204200391500000341258010010080000100800005006400002002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040
802042003915000002762258010010080000100800005006400002002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040
802042003915000001841258010010080000100800005006400002002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040
802042003915000001841258010010080000100800005006400002002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040
80204200391500000641258010010080000100800005006400002002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002009120040200402004020040
80204200391500000041258010010080000100800005006400002002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040
80204200391500000041258010010080000100800005006400002002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040

1000 unrolls and 10 iterations

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

retire (01)cycle (02)030408090b1e1f3a3f51inst 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)9fa8accfitlb miss (d4)d5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
800242004815000004200402580010108000010800005064000012002020039200399996310019800102080000201600002003920039118002110910108000010005021191617142003680000102004020040200402004020040
800242003915010100008925800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100050211181610172003680000102004020040200402004020040
800242003915010100008925800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100050211161617172003680000102004020040200402004020040
800242003915010100008925800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100050211151617172003680000102004020040200402004020040
800242003915010110008925800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100050211171617172003680000102004020040200402004020040
80024200391501010135008925800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100050211101617142003680000102004020040200402004020040
8002420039150101012008925800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100050211171613172003680000102004020040200402004020040
800242003915010100008925800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100050211171614172003680000102004020040200402004020040
800242003915010100008925800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100050211171617172003680000102004020040200402004020040
800242003915010100008925800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100050211171614172003680000102004020040200402004020040