Apple M1 Microarchitecture Research by Dougall Johnson

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

FRSQRTS (scalar, H)

Test 1: uops

Code:

  frsqrts h0, h0, h1
  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)030809191e3f4e51inst 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
10044037300000613407251000100010005319084018403740373258338951000100020004037403711100110001073216333473100040384038403840384038
10044037300000613407251000100010005319084018403740373258338951000100020004037403711100110000073316323473100040384038403840384038
10044037300000613407251000100010005319084018403740373258338951000100020004037403711100110000073316333473100040384038403840384038
10044037311000613407251000100010005319084018403740373258338951000100020004037403711100110001373316333473100040384038403840384038
10044037300000613407251000100010005319084018403740373258338951000100020004037407311100110002073316333473100040384038403840384038
10044037300000613407251000100010005319084018403740373258338951000100020004037403711100110000073316333473100040384038403840384038
10044037300000613407251000100010005319084018403740373258338951000100020004037403711100110000373316333473100040384038403840384038
10044037300000613407251000100010005319084018403740373258338951000100020004037403711100110000073316333473100040384038403840384038
10044037300000613407251000100010005319084018403740373258338951000100020004037403711100110002073316333473100040384038403840384038
10044037300000613407251000100010005319084018403740373258338951000100020004037403711100110007073316333473100040384038403840384038

Test 2: Latency 1->2

Code:

  frsqrts h0, h0, h1
  movi v0.16b, 1
  movi v1.16b, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 4.0037

retire (01)cycle (02)0308181e3f4e51inst 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)9fa0a6a8acbranch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204400372990006139407251010010010000100100005005706908140018400374003738108338745101002001000020020000400374003711102011009910010010000100000007101161139479100001004003840038400384003840038
10204400373000006139407251010010010000100100005005706908040018400374003738108338745101002001000020020000400374003711102011009910010010000100000007101161139479100001004003840038400384003840038
102044003729900996139407251010010010006100100005005706908040018400374003738108338745101002001000020020000400374003711102011009910010010000100000007101161139479100001004003840038400384003840038
10204400373000006139407251010010010000100100005005706908040018400374003738108338745101002001000020020000400374003711102011009910010010000100000007101161139479100001004003840038400384003840038
10204400373001006139407251010010010000100100005005706908040018400374003738108338745101002001000020020000400374003711102011009910010010000100000007101161139479100001004003840038400384003840038
102044003730000072639407251010010010000100100005005706908040018400374003738108338745101002001000020020000400374003711102011009910010010000100000007100161139479100001004003840038400384003840038
10204400372990006139407251010010010000100100005005706908040018400374003738108338745101002001000020020000400374003711102011009910010010000100000007101161139479100001004003840038400384003840038
10204400373000006139407251010010010000100100005005706908140018400374003738108338745101002001000020020000400374003711102011009910010010000100000007101161139479100001004003840038400384003840038
10204400373000006139407251010010010000100100005005706908040018400374003738108338745101002001000020020000400374003711102011009910010010000100000007101161139479100001004003840038400384003840038
10204400372990006139407251010010010000100100005005706908140018400374003738108338745101002001000020020000400374003711102011009910010010000100000007101161139479100001004003840038400384008040038

1000 unrolls and 10 iterations

Result (median cycles for code): 4.0037

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)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)9fa0a1a8acc2branch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002440037300000000061394072510010101000010100005057069080400180400374003738130338767100102010000202000040037400371110021109101010000100000006402162239473010000104003840038400384003840038
1002440037300000000082394072510010101000010100005057069081400180400374003738130338767100102010000202000040037400371110021109101010000100000006402162239473010000104003840038400384003840038
1002440037300000000066394072510010101000010100005057069081400180400374003738130338767100102010000202000040037400371110021109101010000100000006402162239473010000104003840038400384003840038
1002440037300000000061394072510010101000010100005057069080400180400374003738130338767100102010000202000040037400371110021109101010000100000006402162239473010000104003840038400384003840038
1002440037300000000061394072510010101000010100005057069080400180400374003738130338767100102010000202000040037400371110021109101010000100000006402162239473010000104003840038400384003840038
1002440037299000000061394072510010101000010100005057069081400180400374003738130338767100102010000202000040037400371110021109101010000100000006402162239473010000104003840038400384003840038
1002440037300000000061394072510010101000010100005057069081400180400374003738130338767100102010000202000040037400371110021109101010000100000006402162239473010000104003840038400384003840038
1002440037300000000082394072510010101000010100005057069080400180400374003738130338767100102010000202000040037400372110021109101010000100000006402162239473010000104003840038400384003840038
1002440037300000000061394072510010101000010100005057069080400180400374003738130338767100102010000202000040037400371110021109101010000100000006402162239512410000104041940465404624046240452
100244041630311188118870472753933519910063111005413113327157194721402980404504036538165413894210455221145220209684046040513101100211091010100001022029703208286812339817310000104046540466405014022640466

Test 3: Latency 1->3

Code:

  frsqrts h0, h1, h0
  movi v0.16b, 1
  movi v1.16b, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 4.0037

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)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)9fa0a6a8acc2c5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020440037300061394072510100100100001001000050057069084001840037400373811563874110100200100082002001640037400371110201100991001001000010000000111717016003949025100001004003840038400384003840038
102044003730006139407251010010010000100100005005706908400184003740037381157387401010020010008200200164003740037111020110099100100100001000000011171701600394890100001004003840038400384003840038
102044003729906639407251010010010000100100005225706908400184003740037381083387451010020010000200200004003740037111020110099100100100001000010000071011611394790100001004003840038400384003840038
102044003730006139407251010010010000100100005005706908400184003740037381083387451010020010000200200004003740037111020110099100100100001000003000071011611394790100001004003840038400384003840038
102044003730006139407251010010010000100100005005706908400184003740037381083387451010020010000200200004003740037111020110099100100100001000000000071011611394790100001004003840038400384003840038
1020440037300072639407251010010010000100100005005706908400184003740037381083387451010020010000200200004003740037111020110099100100100001000000000071011611394790100001004003840038400384003840038
102044003729906139407251010010010000100100005005706908400184003740037381083387451025020010000200200004003740037111020110099100100100001000000000071011611394790100001004003840038400384003840038
102044003730006139407251010010010000100100005005706908400184003740037381083387451010020010000200200004003740037111020110099100100100001000000000071011611394790100001004003840038400384003840038
102044003730006139407251010010010000100100005005706908400184003740037381083387451010020010000200200004003740037111020110099100100100001000000000071011611394790100001004003840038400384003840038
102044003730006139407251010010010000100100005005706908400184003740037381083387451010020010000200200004003740037111020110099100100100001000000000071011611394790100001004003840038400384003840038

1000 unrolls and 10 iterations

Result (median cycles for code): 4.0037

retire (01)cycle (02)030718191e1f3f4e51inst 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)9fa0a8accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002440037300000006139407251001010100001010000505706908140018400374003738130338767100102010000202000040037400371110021109101010000100006403163339473010000104003840038400384003840038
1002440037300000006139407251001010100001010000505706908140018400374003738130338767100102010000202000040037400371110021109101010000100006403163339473010000104003840038400384003840038
1002440037300000006139407251001010100001010000505706908140018400374003738130338767100102010000202000040037400371110021109101010000100006403163339473010000104003840038400384003840038
1002440037300000006139407251001010100001010000505706908040018400374003738130338767100102010000202000040037400371110021109101010000100006403163339473010000104003840038400384003840038
1002440037300000006139407251001010100001010000505706908040018400374003738130338767100102010000202000040037400371110021109101010000100006403163339473010000104003840038400384003840038
1002440037300000006139407251001010100001010000505706908140018400374003738130338767100102010000202000040037400371110021109101010000100006403163339473010000104003840038400384003840038
1002440037300000006139407251001010100001010000505706908040018400374003738130338767100102010000202000040037400371110021109101010000100006403163339473010000104003840038400384003840038
1002440037300000006139407251001010100001010000505706908140018400374003738130338767100102010000202000040037400371110021109101010000100006403163339473010000104003840038400384003840038
10024400372990000012439407251001010100001010000505706908140018400374003738130338767100102010000202000040037400371110021109101010000100006403163339473010000104003840038400384003840038
1002440037300000906139407251001010100001010000505706908140018400374003738130338767100102010000202000040037400371110021109101010000100006403163339473010000104003840038400384003840038

Test 4: throughput

Count: 8

Code:

  frsqrts h0, h8, h9
  frsqrts h1, h8, h9
  frsqrts h2, h8, h9
  frsqrts h3, h8, h9
  frsqrts h4, h8, h9
  frsqrts h5, h8, h9
  frsqrts h6, h8, h9
  frsqrts h7, h8, h9
  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)9fa0a1a6a7a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8020420059150000000042258010010080000100800005006400000200212004020040997339998801002008000020016000020040200401180201100991001008000010000000000511021611200370800001002004120041200412004120041
8020420040150000000042258010010080000100800005006400000200212004020040997339998801002008000020016000020040200401180201100991001008000010000000000511011611200370800001002004120041200412004120041
8020420040150000000042258010010080000100800005006400000200212004020040997339998801002008000020016000020040200401180201100991001008000010000000090511011611200370800001002004120041200412004120041
8020420040150000000042258010010080000100800005006400001200212004020040997339998801002008000020016000020040200401180201100991001008000010000000000511011611200370800001002004120041200412004120041
8020420040150000000042258010010080000100800005006400001200212004020040997339998801002008000020016000020040200401180201100991001008000010000000000511011611200370800001002004120041200412004120041
8020420040150000000042258010010080000100800005006400001200212004020040997339998801002008000020016000020040200401180201100991001008000010000003000511011611200370800001002004120041200412004120041
8020420040150000000042258010010080000100800005006400000200212004020040997339998801002008000020016000020040200401180201100991001008000010000000000511011611200370800001002004120041200412004120041
80204200401500000000422580100100800001008000050064000002002120040200409973399988010020080000200160000200402004011802011009910010080000100000000990511021611200370800001002004120041200412004120041
8020420040150000000042258010010080000100800005006400000200212004020040997339998801002008000020016000020040200401180201100991001008000010000000000511011611200370800001002004120041200412004120041
8020420040150000000042258010010080000100800005006400000200212004020040997339998801002008000020016000020040200401180201100991001008000010000000000511011611200370800001002004120041200412004120041

1000 unrolls and 10 iterations

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

retire (01)cycle (02)03091e3f51inst 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)9fa6a8a9accfd0d5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420049150008325800101080000108000050640000120021200402004099963100208001020800002016000020040200401180021109101080000100000502001161120037080000102004120041200412004120041
8002420040150004125800101080000108000050640000120021200402004099963100208001020800002016000020040200401180021109101080000100000502001161120037080000102004120041200412004120041
800242004015000354125800101080000108000050640000120021200402004099963100208001020800002016000020040200401180021109101080000100000502001161120037080000102004120041200412004120041
80024200401500010425800101080000108000050640000120021200402004099963100208001020800002016000020040200401180021109101080000100101440502001161120037080000102004120041200412004120041
8002420040150008525800101080000108000050640000120021200402004099963100208001020800002016000020040200401180021109101080000100330147502001161120037080000102004120041200412004120041
8002420040150004125800101080000108000050640000120021200402004099963100208001020800002016000020040200401180021109101080000100000502001161120037080000102004120041200412004120041
8002420040150004125800101080000108000050640000120021200402004099963100208001020800002016000020040200401180021109101080000100000502001161120037080000102004120041200412004120041
8002420040150008525800101080000108000050640000120021200402004099963100208001020800002016000020040200401180021109101080000100000502001161120037080000102004120041200412004120041
8002420040150004125800101080000128000050640000120021200402004099963100208001020800002016000020040200401180021109101080000100000502001161120037080000102004120041200412004120041
80024200401500010625800101080000108000050640000120021200402004099963100208001020800002016000020040200401180021109101080000100000502001161120037080000102004120041200412004120041