Apple M1 Microarchitecture Research by Dougall Johnson

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

FRECPE (vector, 8H)

Test 1: uops

Code:

  frecpe v0.8h, v0.8h
  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
100430372306119002510001000100010445103018303730372738328951000100010003037303711100110000073116112919100030383038303830383038
100430372306119002510001000100010445113018303730372738328951000100010003037303711100110001073116112919100030383038303830383038
100430372306119002510001000100010445113018303730372738328951000100010003037303711100110000973116112919100030383038303830383038
10043037230611900251000100010001044511301830373037273832895100010001000303730371110011000116573116112919100030383038303830383038
100430372206119002510001000100010445113018303730372738328951000100010003037303711100110000673116112919100030383038303830383038
100430372206619002510001000100010445113018303730372738328951000100010003037303711100110004073116112919100030383038303830383038
100430372206119002510001000100010445113018303730372738328951000100010003037303711100110000073116112919100030383038303830383038
100430372306119002510001000100010445113018303730372738328951000100010003037303711100110000073116112919100030383038303830383038
1004303722061190025100010001000104451030183037303727383289510001000100030373037111001100011073116112919100030383038303830383038
100430372208219002510001000100010445113018303730372738328951000100010003037303711100110000073116112919100030383038303830383038

Test 2: Latency 1->2

Code:

  frecpe v0.8h, v0.8h
  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)9fa8a9acc2cdcfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102043003722500000611990025101001001000010010000500106745113001830037300372858832874510100200100002001000030037300371110201100991001001000010000000734116112991912100001003013230086300383003830038
10204300372250003061199002510100100100001001000050010674511300183003730037285883287451010020010000200100003003730037111020110099100100100001000000071011611299190100001003003830038300383003830038
102043003722400000103199002510100100100001001000050010674511300183003730037285883287451010020010000200100003003730037111020110099100100100001000000071011611299190100001003003830038300383003830038
10204300372250000061199002510100100100001001000050010674510300183003730037285883287451010020010000200100003003730037111020110099100100100001000000071021611299190100001003003830038300383003830038
102043003722509581370429451977213410200124100721271037864410687391302343045930319287122829032105072061050321210336304653031810110201100991001001000010010320829289233024326100001003035930463304623036930462
1020430321228196798792103199002510100100100001001000050010674511300183003730037285883287451010020010000200100003003730037111020110099100100100001000000071011611299190100001003003830038300383003830038
102043003722500000264199002510100100100001001000050010674511300183003730037285883287451010020010000200100003003730037111020110099100100100001000000071011611299190100001003003830038300383003830038
102043003722500000491199002510100100100001001000050010674511300183003730037285883287681010020010000200100003003730037111020110099100100100001000000071011611299190100001003003830038300383003830038
1020430037225000510103199002510100100100001001000050010674511300183003730037285883287451010020010000200100003003730037111020110099100100100001000000071011611299190100001003003830038300383003830038
102043003722500000193199002510100100100001001000050010674511300183003730037285883287451010020010000200100003003730037111020110099100100100001000000071011611299190100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)030f1e3f4e51inst 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)9fa0a1a6a8acc2c5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037225008219900251001010100001010000501067451130018300373003728610328767100102010000201000030037300371110021109101010000100001300006402162229919010000103003830038300383003830038
10024300372240016619900251001010100001010000501067451130018300373003728610328767100102010000201000030037300371110021109101010000100000000006402162229919010000103003830038300383003830038
10024300372250014519900251001010100001010000501067451030018300373003728610328767100102010000201000030037300371110021109101010000100000000006402162229919010000103003830038300383003830038
10024300372250012419900251001010100001010000501067451030018300373003728610328767100102010000201000030037300371110021109101010000100000000006402162229919010000103003830038300383003830038
10024300372240012419900251001010100001010000501067451030018300373003728610328767100102010000201000030037300371110021109101010000100000000006402162229919010000103003830038300383003830038
10024300372250047819900251001010100001010000501067451030018300373003728610328767100102010000201000030037300371110021109101010000100000000006402162229919010000103003830038300383003830038
10024300372240012419900251001010100001010000501067451030018300373003728610328767100102010000201000030037300371110021109101010000100000000006402162229919010000103003830038300383003830038
10024300372250016619900251001010100001010000501067451130018300373003728610328767100102010000201000030037300371110021109101010000100000000006402162229919010000103003830038300383003830038
10024300372240014519900251001010100001010000501067451030018300373003728610328767100102010000201000030037300371110021109101010000100000000006402162229919010000103003830038300383003830038
10024300372240156119900251001010100001010000501067451030018300373003728610328767100102010000201000030037300371110021109101010000100000000006402162229919010000103003830038300383003830038

Test 3: throughput

Count: 8

Code:

  frecpe v0.8h, v8.8h
  frecpe v1.8h, v8.8h
  frecpe v2.8h, v8.8h
  frecpe v3.8h, v8.8h
  frecpe v4.8h, v8.8h
  frecpe v5.8h, v8.8h
  frecpe v6.8h, v8.8h
  frecpe v7.8h, v8.8h
  movi v8.16b, 9

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 1.0005

retire (01)cycle (02)03041e3f51inst 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)9fa0a6a8acc5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8020480040600103925801001008000010080000500640000180020800398003969971669993801002008000820080008800398003911802011009910010080000100000311151170160080036800001008004080040800408004080040
8020480039599003925801001008000010080000500640000180020800398003969965969981801022008001420080014800398003911802011009910010080000100000022251271231180035800001008004080040800408004080040
8020480039599006926801001008000010080002500640012180020800398003969965969980801022008001420080014800398003911802011009910010080000100000022251281231180035800001008004080040800408004080040
8020480039600006926801001018000010080002500640012180069800398003969965969980801022008001420080014800398003911802011009910010080000100000022251281231180035800001008004080040800408004080040
8020480088599006925801001008000010080002500640012180020800398003969965969981801022008001420080014800398003911802011009910010080000100000022251271231180035800001008004080040800408004080040
80204800395990025725801001008000010080002500640012180020800398003969965969980801022008001420080014800398003911802011009910010080000100000022251281231180035800001008004080040800408004080040
80204800395990069258010010080000100800025006400121800208003980039699651069980801022008001420080014800398003911802011009910010080000100000022251271231180035800001008004080040800408004080040
80204800396000069258010010080000100800025006400121800208003980039699651069981801022008001420080014800398003911802011009910010080000100000022251281231180035800001008004080040800408004080040
80204800396000069258010010080000100800025006400121800208003980039699651069981801022008001420080014800398003911802011009910010080000100000022251281231180035800001008004080040800408004080040
80204800396000069258010010080000100800025006400121800208003980039699651069981801022008001420080014800398003911802011009910010080000100000022251281231180035800001008004080040800408004080040

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 1.0005

retire (01)cycle (02)030b18191e3f51inst 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)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)9fa8acc2branch mispredict (cb)cdcfd0d5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002480040599000050258001010800001080000506400000800208003980039699860370019800102080000208000080039800391180021109101080000100000050200316448003480000108004080040800408004080040
8002480039600000050258001010800001080000506400001800208003980039699860370019800102080000208000080039800391180021109101080000100000050200516438003480000108004080040800408004080040
8002480039600000050258001010800001080000506402030800208003980039699860370019800102080000208000080039800391180021109101080000100000050200216448003480000108004080040800408004080040
8002480039599000050258001010800001080000506400000800208003980039699860370019800102080000208000080039800391180021109101080000100000050200416448003480000108004080040800408004080040
8002480039599000050258001010800001080000506400000800208003980039699860370019800102080000208000080039800391180021109101080000100000050200416248003480000108004080040800408004080040
80024800395990006650258001010800001080000506400000800208003980039699860370019800102080000208000080039800881180021109101080000100000050200316538003480000108004080040800408004080040
8002480039599000050258001010800001080000506400001800208003980039699860370019800102080000208000080039800391180021109101080000100000050200316448003480000108004080040800408004080040
8002480039600000050258001010800241080000506400000800208003980039699860370019800102080000208000080039800391180021109101080000100000050200316348003480000108004080040800408004080040
80024800395990001250258001010800001080000506400001800208003980039699860370019800102080000208000080039800391180021109101080000100300050200416338003480000108004080040800408004080040
8002480039599000350258001010800001080000506400000800208003980039699860370019800102080000208000080039800401180021109101080000100000050200416448003480000108004080040800408004080040