Apple M1 Microarchitecture Research by Dougall Johnson

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

UMLAL2 (vector, 4S)

Test 1: uops

Code:

  umlal2 v0.4s, v1.8h, v2.8h
  movi v0.16b, 1
  movi v1.16b, 2
  movi v2.16b, 3

(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)0308090b18191e1f3f4e51inst 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)a0a1a6a8a9acc2cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
10043037280000000612548251000100010003983131301830373037241532895100010003000303730371110011000000000073216112630100030383038303830383038
100430378700001003192548251000100010003983130301830373037241532895100010003000303730371110011000002020073116112630100030383038303830383084
10043084280010000612548251000100010003983130301830373037241532895100010003000303730371110011000000000073116112630100030383038303830383038
10043037280000000612548251000100010003983130301830373037241532895100010003000303730371110011000000000073116112630100030383038303830383038
10043037280000000612548251000100010003983130301830373037241532895100010003000303730371110011000000000073116112630100030383038303830383038
10043037280000000612548251000100010003983130301830373037241532895100010003000303730371110011000000000073116112630100030383038303830383038
10043037260000000612548251000100010003983130301830373037241532895100010003000303730371110011000000000073116112630100030383038303830383038
10043037260000000612548251000100010003983131301830373037241532895100010003000303730371110011000000000073116112630100030383038303830383038
100430372600000006125482510001000100039831303018303730372415328951000100030003037303711100110000002200073116112630100030383038303830383038
10043037260000000612548251000100010003983131301830373037241532895100010003000303730371110011000000000073116112630100030383038303830383038

Test 2: Latency 1->1

Code:

  umlal2 v0.4s, v1.8h, v2.8h
  movi v0.16b, 1
  movi v1.16b, 2
  movi v2.16b, 3

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03081e3a3f4e51inst 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)9fa8a9accfd0d5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020430037225000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000071013163229634100001003003830038300383003830038
1020430037225100612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010010071012162229634100001003003830038300383003830038
1020430037224000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000071013162229634100001003003830038300383003830038
10204300372250011032954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000071012162229634100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731303001830037300372826532874510252200100002003050130037300371110201100991001001000010000071012162229634100001003003830038300383003830038
1020430037225000612954845101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000071212162229634100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731313001830037300372826532874510100200100002003000030037300371110201100991001001000010000071012162229634100001003003830038300383003830038
1020530037225000612954825101001001000010010000500427731313001830037300372826532874510100200100002003000030037300371110201100991001001000010000071214162229634100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731313001830037300372826532874510100200100002003000030037300371110201100991001001000010000071012162229634100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731313001830037300372826532874510100200100002003000030037300371110201100991001001000010000071012162229634100001003003830038300383003830038

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)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)9fa0a1a6a7a8a9acc2branch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100243003722400000082295482510010101000010100005042773131300183003730037282873287671001020100002030000300373003711100211091010100001000000024006402162229630010000103003830038300383003830038
10024300372250000008429548251001010100001010000504277313030018300373003728287328767100102010000203000030037300371110021109101010000100000000006402162229630110000103003830038300383003830038
10024300372250000006129548251001010100001010000504277313030018300373003728287328767100102010000203000030037300371110021109101010000100000000006402162229630010000103003830038300383003830038
10024300372250000008429548251001010100001010000504277313030018300373003728287328767100102010000203000030037300371110021109101010000100000000006402162229630010000103003830038300383003830083
10024300372250000006129548251001010100001010000504277313030018300373003728287328767100102010000203000030037300371110021109101010000100000000006402162229630010000103003830038300383003830038
100243003722500000018929548251001010100001010000504277313030018300373003728287328767100102010000203000030037300371110021109101010000100000000006402162229630010000103003830038300383003830038
10024300372250000008429548251001010100001010000504277313130018300373003728287328767100102010000203000030037300371110021109101010000100000000006402162229630110000103003830038300383003830038
100243003722500000016829548251001010100001010000504277313030018300373003728287328767100102010000203000030037300371110021109101010000100000000006402162229630010000103003830038300383003830038
10024300372250000008429548251001010100001010000504277313030018300373003728287328767100102010000203000030037300371110021109101010000100000000006402162229630010000103003830038300383003830038
100243003722500000057429548251001010100001010000504277313030018300373003728299328767100102010000203000030037300371110021109101010000100000000006402162329630010000103003830038300383003830038

Test 3: Latency 1->2

Code:

  umlal2 v0.4s, v0.8h, v1.8h
  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)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)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)9fa1a6a7a8acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020430037225000000612954825101001001000010010000500427731313001830037300372826532874510100200100002003000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037225010000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037225000000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
10204300372250000007262954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037225000000612954825101001001000010010000500427731313001830037300372826532874510100200106542003000030037300741110201100991001001000010000003071011611296340100001003003830038300383003830038
10204300372250000870612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037224000000612954825101001221000012010149500427731303001830037300372826532874510425200100002003000030037300371110201100991001001000010000100071011611296340100001003003830038300383003830038
1020430086225000100612954869101001001000010010000605427867003001830037300372826532874510100200100002003000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037225010000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037225000000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300851110201100991001001000010000000071011611296340100001003003830086300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)0307081e3f4e51inst 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)9fa0a1a6a8a9acc2cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037225019612954825100101010000101000050427867030018300373003728287328767103112010000203000030037300371110021109101010000100000000640316332963010000103003830038300383003830038
10024300372250001032954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100000000640316332963010000103003830038300383003830038
10024300372250012612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100000000640316332963010000103003830038300383003830038
10024300372250012612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100000002640316332963010000103003830038300383003830038
1002430037225000612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100000000640316332963010000103003830038300383003830038
1002430037225000612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100000000640316332963010000103003830038300383003830038
1002430037225000612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100000000640316332963010000103003830038300383003830038
1002430037225000612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100000000640316332963010000103003830038300383003830038
1002430037225000612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100000000640316332963010000103003830038300383003830038
1002430037225000612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100000000640316332963010000103003830038300383003830038

Test 4: Latency 1->3

Code:

  umlal2 v0.4s, v1.8h, 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)03080b181e3f4e51inst 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)9fa6a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102043003722500006129548251010010010000100100005004277313300183003730037282653287451010020010000200300003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
1020430037224000276629548251010010010000100100005004277313300183003730037282653288171010020010000200300003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
1020430037225000216129548251010010010000100100005004277313300183003730037282653287451010020010000200300003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
102043003722500006129548251010010010000100100005004277313300183003730037282653287451010020010000200300003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
1020430037225000156129548251010010010000100100005004277313300183003730037282653287451010020010000200300003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
10204300372250003336129548251010010010000100100005004277313300183003730037282653287451010020010000200300003003730037111020110099100100100001000007101171129634100001003003830038300383003830038
1020430037225000156129548251010010010000100100005004277313300183003730037282653287801010020010000200300003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
1020430037225000246129548251010010010000100100005004277313300183003730037282653287451010020010000200300003003730037111020110099100100100001000107101161129634100001003003830038300383003830038
1020430037225002306129548641011211610008100100005004277313300183003730037282653287451010020010000200300003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
1020430132226000074929548251013210010000100100005004277313300183003730037282653287451010020010000200300003003730037111020110099100100100001000007101161129634100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03081e3f4e51inst 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)9fa0a8accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037225036612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100006405162229630010000103003830038300383003830038
100243003722500612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250271032954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
100243003722500612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
100243003722500612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
100243003722500612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
1002430037225004412954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
1002430037225027612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
1002430037225030612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250165612954825100101010000101000050427731330018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038

Test 5: throughput

Count: 8

Code:

  movi v0.16b, 0
  umlal2 v0.4s, v8.8h, v9.8h
  movi v1.16b, 0
  umlal2 v1.4s, v8.8h, v9.8h
  movi v2.16b, 0
  umlal2 v2.4s, v8.8h, v9.8h
  movi v3.16b, 0
  umlal2 v3.4s, v8.8h, v9.8h
  movi v4.16b, 0
  umlal2 v4.4s, v8.8h, v9.8h
  movi v5.16b, 0
  umlal2 v5.4s, v8.8h, v9.8h
  movi v6.16b, 0
  umlal2 v6.4s, v8.8h, v9.8h
  movi v7.16b, 0
  umlal2 v7.4s, v8.8h, v9.8h
  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): 0.2508

retire (01)cycle (02)031e3f51inst 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)60696d6escheduler 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)9fa8acc5cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1602042009115036392580100100800001008000050064000002004520064200643228010020080000200240000200642006411160201100991001001600001000001011121611200611600001002006520065200652006520065
160204200641500392580100100800001008000050064000002004520064200643228010020080000200240000200642006411160201100991001001600001000001011111611200611600001002006520065200652006520065
16020420064151348392580100100800001008000050064000002004520064200643228010020080000200240000200642006411160201100991001001600001000001011111611200611600001002006520065200652006520065
16020420064150152082580100100800001008000050064000012004520064200643228010020080000200240000200642006411160201100991001001600001000001011117211200611600001002006520065200652006520065
160204200641500392580100100800001008000050064000012004520064200643228010020080000200240000200642006411160201100991001001600001000001011111611200611600001002006520065200652006520065
160204200641510392580100100800001008000050064000002004520064200643228010020080000200240000200642006411160201100991001001600001000001011111611200611600001002006520065200652006520065
160204200641500672580100100800001008000050064000012004520064200643228010020080000200240000200642006411160201100991001001600001000001011111611200611600001002006520065200652006520065
160204200641506392580100100800001008000050064000002004520064200643228010020080000200240000200642006411160201100991001001600001000001011111611200611600001002006520065200652006520065
160204200641500392580100100800001008000050064000002004520064200643228010020080000200240000200642006411160201100991001001600001000001011111611200611600001002006520065200652006520065
160204200641500392580100100800001008000050064000002004520064200643228010020080000200240000200642006411160201100991001001600001000001011111611200611600001002006520065200652006520065

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2506

retire (01)cycle (02)031e3f51inst 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)5f60696d6escheduler 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)9fa7a8accfd0d2icache miss (d3)d5d6d9dadbddinst fetch restart (de)e0? int output thing (e9)eaec? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1600242006815021452980012128000012800006264000001200412006420051322800122080000202403122006020060111600211091010160000101001003261263422267200592402160000102006120061200612006120061
1600242006015133517880012128000012800006264000001200432008920062322800122080000202400002006020060111600211091010160000100001003162272542256200572402160000102006320063200522006320061
160024200601509512780012128000012800006264000010200412008620062322800122080000202400002006020062111600211091010160000100001003161283442255200572402160000102006120052200632006320052
160024200601509512980012128000012800006264000001200322006420060322800122080000202400002005120062111600211091010160000100001003162263642265200572401160000102006320061200612006120052
160024200601503512980012128000012800006264000001200412007320062322800122080000202400002006020060111600211091010160000100101003162273432256200592412160000102006120061200632006120061
1600242006015012932980012128000012800006264000001200412006420060322800122080000202400002006220062111600211091010160000100101003262263432277200572412160000102006120063200632006320061
1600242006215012512980012128000012800006264000001200412007320060322800122080000202400002006020060111600211091010160000100031003131162521166200482201160000102005220052200522005220052
1600242005115024452780012128000012800006264000011200322006420062322800122080000202400002006220060111600211091010160000100001003262253432267200592402160000102006120063200632006320150
1600242006015033512980012128000012800006264000001200412006420051322800122080000202400002005120051111600211091010160000100091003262273442267200592411160000102006120063200632006320063
1600242006015012512980012128000012800006264000001200432006420060322800122080000202400002006020060111600211091010160000100001003362263642278200572412160000102006120061200612006120063

Test 6: throughput

Count: 16

Code:

  umlal2 v0.4s, v16.8h, v17.8h
  umlal2 v1.4s, v16.8h, v17.8h
  umlal2 v2.4s, v16.8h, v17.8h
  umlal2 v3.4s, v16.8h, v17.8h
  umlal2 v4.4s, v16.8h, v17.8h
  umlal2 v5.4s, v16.8h, v17.8h
  umlal2 v6.4s, v16.8h, v17.8h
  umlal2 v7.4s, v16.8h, v17.8h
  umlal2 v8.4s, v16.8h, v17.8h
  umlal2 v9.4s, v16.8h, v17.8h
  umlal2 v10.4s, v16.8h, v17.8h
  umlal2 v11.4s, v16.8h, v17.8h
  umlal2 v12.4s, v16.8h, v17.8h
  umlal2 v13.4s, v16.8h, v17.8h
  umlal2 v14.4s, v16.8h, v17.8h
  umlal2 v15.4s, v16.8h, v17.8h
  movi v16.16b, 17
  movi v17.16b, 18

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.2502

retire (01)cycle (02)030b18191e1f373f51inst 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)5f60696b6d6edispatch 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)9fa0a1a8acc5branch mispredict (cb)cdcfd5d6dadbddinst fetch restart (de)e0? int output thing (e9)ea? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1602044005830000000173025160108100160008100160020500128013201400300403344003919977061999016012020016003220048009640039400401116020110099100100160000100000011110118316002240036001600001004004040040400414004040040
160204400493000000003925160108100160008100160020500128013201400200400394004019977061999116012020016003220048009640040400391116020110099100100160000100000011110118216002240036001600001004005040049400724004040040
160204400403000000005025160100100160000100160000500132000001400200400404003919973032000716010020016000020048000040049400481116020110099100100160000100000000010110316003340036001600001004004940040400494004940050
160204400393000000014125160100100160000100160000500128000001400290400394004919973032000616010020016000020048000040048400391116020110099100100160000100000000010110316002340045001600001004004040040400404004940049
1602044003930000000174125160100100160017100160000525128000001400210400404003919973032000716010020016000020048000040049400481116020110099100100160000100000000010110316003340037001600001004005040049400404005040040
160204400393000000005025160117100160017100160000500239902701400290400394004919973031999716010020016000020048000040040400391116020110099100100160000100000000010110216003340036001600001004004040143400494007240040
160204400393000000004125160100100160000100160000500239902701400200400394004819973031999716010020016000020048000040039400481116020110099100100160000100000000010110317003240036001600001004004040040400404004040040
1602044003930000000174125160117100160000100160000500128000001400290400494003919973031999716010020016000020048000040071400391116020110099100100160000100000000010110316003340036001600001004004040050400494004040040
1602044004830000000175125160117100160061100160000500239899901400200400394007119973032000716010020016000020048000040040400391116020110099100100160000100000000010110216003340037001600001004004040040400504004940050
1602044004830000000174125160100100160017100160000500128000001400200400484007119980031999716010020016000020048000040040400391116020110099100100160000100000000010110316003340095001600001004004940040400494004940049

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2502

retire (01)cycle (02)031e373f51inst 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)9fa8a9accfd0d2icache miss (d3)d5d6d9dadbddinst fetch restart (de)e0eaeb? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1600244005830069174625160010101600001016000050243886511400204004840039199963200281600102016000020480000400394003911160021109101016000010000100223117162115840036155160000104004040040400404004040040
16002440039300004625160010101600181016000050128000011400204004840048199963200191600102016000020480000400394003911160021109101016000010000100223115162116440036155160000104004040040400404005040050
160024400393000046251600101016000010160000501280000114003040039400391999632002816001020160000204800004003940039111600211091010160000100001002462231642244400463010160000104004040040400404004040040
160024400393002405225160010101600001016000050128000001400204003940048199963200191600102016000020480000400394003911160021109101016000010000100223114162114440037155160000104004040040400504004040040
1600244003930039046251600101016000010160000501280000114002040039400391999632002916001020160000204800004004940049111600211091010160000102001002462231642264400363010160000104004040040400404004040040
160025400393000052251600101016000010160000501280000014002040039400391999632001916001020160000204800004003940039111600211091010160000100001002462271642244400363010160000104005040050400404004040040
16002440039299005225160010101600001016000050128000001400204004840048199963200191600102016000020480000400394003911160021109101016000010000100223117162116440036155160000104004040040400404004040040
160024400393000046251600101016000010160000501280000114002040049400481999632001916001020160000204800004003940039111600211091010160000100001002462261641244400361510160000104004040040400404004040040
1600244003930001846251600101016000010160000501280000014002040039400391999632001916001020160000204800004003940049111600211091010160000100001002462241642144400363010160000104004040040400404004040040
16002440049300195052251600101016000010160000501280000114002040048400392004632001916001020160000204800004003940039111600211091010160000102031002462281641255400363010160000104004040040400404004040040