Apple M1 Microarchitecture Research by Dougall Johnson

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

SQRDMLAH (vector, 2S)

Test 1: uops

Code:

  sqrdmlah v0.2s, v1.2s, v2.2s
  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)03080b18191e1f3a3f4e51inst 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)a0a1a6a7a8a9aaacc2cfd5d6ddinst fetch restart (de)e0ea? fp/simd (ee)f5f6f7f8fd
10043037230000000612548251000100010003983130301830373037241532895100010003000303730371110011000000000000731161126300100030383038303830383038
100430372200000002972548251000100010003983130301830373037241532895100010003000303730371110011000000000000731161126300100030383038303830383038
10043037230000000612548251000100010003983130301830373037241532895100010003000303730371110011000000000000731161126300100030383038303830383038
10043037220000000822548251000100010003983130301830373037241532895100010003000303730371110011000000000000731161126300100030383038303830383038
10043037230000600612548251000100010003983130301830373037241532895100010003000303730371110011000000000000731161126300100030383038303830383038
10043037220000000612548251000100010003983130301830373037241532895100010003000303730371110011000000000000731161126300100030383038303830383038
10043037220000000612548251000100010003983130301830373037241532895100010003000303730371110011000000000000731161126300100030383038303830383038
10043037230000000612548251000100010003983130301830373037241532895100010003000303730371110011000000000000731161126300100030383038303830383038
10043037230000000612548251000100010003983130301830373037241532895100010003000303730371110011000000000000731161126300100030383038303830383038
10043037230000000612548251000100010003983130301830373037241532895100010003000303730371110011000000000000731161126300100030383038303830383038

Test 2: Latency 1->1

Code:

  sqrdmlah v0.2s, v1.2s, v2.2s
  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)0318191e1f3a3f4e51inst 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)9fa1a6a8a9accfd0d5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102043003722400000962954825101001001000010010000500427731313001830037300372826532874510100200100002003000030037300371110201100991001001000010000000710021622296340100001003003830038300383003830038
102043003722500000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000000712021622296340100001003003830038300383003830038
10204300372250000020762954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000000710121633296700100001003003830038300383003830038
1020430037225006900612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000000710121622296340100001003003830038300383003830038
102043003722500000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000000712121622296340100001003003830038300383003830038
102043003722500900612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000000710121622296340100001003003830038300383003830038
102043003722500000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000000710121622296340100001003003830038300383003830038
102043003722500000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000000710121623296340100001003003830038300383003830038
102043003722500000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000000710121622296340100001003003830038300383003830038
1020430037225000006129548251010010010000100101495004277313030018300373003728265628745101002001000020030000300373003711102011009910010010000100001200710121632296340100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03080b1e3a3f4e51inst 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)9fa0a8acc5cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100243003722500015362954825100101010000101000050427731303001830037300372828703287671001020100002030000300373003711100211091010100001000000642816682963010000103003830038300383003830038
100243003722500026129548251001010100001010000504277313030018300373003728287032876710010201000020300003003730037111002110910101000010000006441016882963010000103003830038300383003830038
1002430037225000161295482510010101000010100005042773130300183003730037282870328767100102010000203000030037300371110021109101010000100000064210161092963010000103003830038300383003830038
100243003722500013462954825100101010000101000050427731303001830037300372828703287671001020100002030000300373003711100211091010100001000000642816862963010000103003830038300383003830038
10024300372250001612954825100101010000101000050427731303001830037300372828703287671001020100002030000300373003711100211091010100001001900642916862963010000103003830038300383003830038
1002430122224001215712954825100101010000101000050427731313001830037300372828703287671001020100002030000300373003711100211091010100001000000642616872963010000103003830038300383003830038
100243003722500017262954825100101010000101000050427731303001830037300372828703287671001020100002030000300373003711100211091010100001000000642816782963010000103003830038300383003830038
10024300372250001822954825100101010000101000050427731303001830037300372828703287671001020100002030000300373003711100211091010100001000000642816772963010000103003830038300383003830038
10024300372250001612954825100101010000101000050427731303001830037300372828703287671001020100002030000300373003711100211091010100001000003642816862963010000103003830038300383003830038
10024300372250001822954825100101010000101000050427731303001830037300372828703287671001020100002030000300373003711100211091010100001000000642816992963010000103003830038300383003830038

Test 3: Latency 1->2

Code:

  sqrdmlah v0.2s, v0.2s, v1.2s
  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)0318191e3f4e51inst 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)9fa0a8acbranch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020430037225000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000007102161129634100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
10204300372250072612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731313001830037300372826532874510100200100002003000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731313001830037300372826532874510100200100002003000030037300371110201100991001001000010000007101161129670100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731313001830037300372826532874510100200100002003000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
1020430037224000612954825101001001000010010000500427731303001830037300372826532879510100200100002003000030037300371110201100991001001000010000007101161129634100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000017101161129634100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731313001830037300372826532874510100200100002003000030037300375110201100991001001000010000007101161129634100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03070b191e3f4e51inst 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)9fa0a1a6a8a9accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037225000357346295482510010101000010100005042773131300180300373003728287328767100102010000203000030037300371110021109101010000100000006403164329630010000103003830038300383003830038
1002430037225000333726295482510010101000010100005042773131300180300373003728287328767100102010000203000030037300371110021109101010000100000006404164329630010000103003830038300383003830038
1002430037225010318251295482510010101000010100005042773131300180300373003728287328767100102010000203000030037300371110021109101010000100000006402163329630010000103003830038300383003830038
100243003722500035761295482510010101000010100005042773131300180300373003728287328767100102010000203000030037300371110021109101010000100000006404164329630010000103003830038300383003830038
100243003722500035161295482510010101000010100005042773131300180300373003728287328767100102010000203000030037300371110021109101010000100000006404163429630010000103003830038300383003830038
10024300372250000254295482510010101000010100005042773131300180300373003728287328767100102010000203000030037300371110021109101010000100000006404164429630010000103003830038300383003830038
100243003722500028561295482510010101000010100005042773131300180300373003728287328767100102010000203000030037300371110021109101010000100000006404163429630010000103003830038300383003830038
10024300372250006061295482510010101000010100005042773131300180300373003728287328767100102010000203000030037300371110021109101010000100000036404164329630010000103003830038300383003830038
1002430037224000061295482510010101000010100005042773131300180300373003728287328767106092010000203000030037300371110021109101010000100000006404164429630010000103003830038300383003830038
10024300372250000726295482510010101000010100005042773131300180300373003728287328767100102010000203000030037300373110021109101010000100000006404163429630010000103003830038300383003830038

Test 4: Latency 1->3

Code:

  sqrdmlah v0.2s, v1.2s, v0.2s
  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)0318191e1f3f4e51inst 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)9fa0a6a7a8acc5cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204301802254315061295482510100100100001001000050042773131300183003730037282653287451010020010000200300003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
1020430037225000061295482510100100100001001000050042773130300183003730037282653287451010020010000200300003003730037111020110099100100100001000000554807101161129634100001003003830038300383003830038
1020430037225000061295482510100100100001001000050042773130300183003730037282653287451010020010000200300003003730037111020110099100100100001000000007101161129702100001003003830038300383003830038
1020430037225000061295482510100100100001001000050042773130300183003730037282653287451010020010000200300003003730037111020110099100100100001000000907101161129634100001003003830038300383003830038
102043003722500006129548251010010010000100100005004277313030018300373003728265328745101002001000020030000300373003711102011009910010010000100000011407101161129634100001003003830038300383003830038
10204300372240027061295482510100100100001001000050042773131300183003730037282653287451010020010000200300003003730037111020110099100100100001000000907101161129634100001003003830038300383003830038
1020430037225000061295482510100100100001001000050042773130300183003730037282653287451010020010000200300003003730037111020110099100100100001000000907101161129634100001003003830038300383003830038
10204300372250000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000019907101161129634100001003003830038300383003830038
10204300372250000612954825101001001000010010000500427731303001830037300372826532874510100200100002003000030037300371110201100991001001000010000002407101161129634100001003003830038300383003830038
1020430037225000061295482510100100100001001000050042773130300183003730037282653287451010020010000200300003003730037111020210099100100100001000000307101161129634100001003003830038300383003830038

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)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? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100243003722500034729548251001010100001010000504277313030018300373003728287328767100122010000203000030037300371110021109101010000100006842162229630210000103003830038300383003830038
100243003722500069829548251001010100001010000504277313030018300373003728287328767100122010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250006129548251001212100001210000604277313030018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250006129548251001010100001010000504277313030018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250096129548251001010100001010000504277313030018300373003728287328767100102010000203000030037300371110021109101010000100006404164429632010000103003830038300383003830038
100243003722510312429530251001010100001010000504277313030054300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250006129548251001010100001010000504277313030018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250006129548251001010100001010000504277313030018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
100243003722500025129548251001212100001010000504277313030018300373003728287328767100102010000203000030037300371110021109101010000100006402162229630010000103003830038300383003830038
100243003722500072629548251001010100001010000504277313030018300373003728287328767100102010000203000030037301311110021109101010000100006402162229630010000103003830038300383003830038

Test 5: throughput

Count: 8

Code:

  movi v0.16b, 0
  sqrdmlah v0.2s, v8.2s, v9.2s
  movi v1.16b, 0
  sqrdmlah v1.2s, v8.2s, v9.2s
  movi v2.16b, 0
  sqrdmlah v2.2s, v8.2s, v9.2s
  movi v3.16b, 0
  sqrdmlah v3.2s, v8.2s, v9.2s
  movi v4.16b, 0
  sqrdmlah v4.2s, v8.2s, v9.2s
  movi v5.16b, 0
  sqrdmlah v5.2s, v8.2s, v9.2s
  movi v6.16b, 0
  sqrdmlah v6.2s, v8.2s, v9.2s
  movi v7.16b, 0
  sqrdmlah v7.2s, v8.2s, v9.2s
  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)03070a1e3f51inst 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)60696b6d6escheduler 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)9fa8acc5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1602042008915111029258011610080016100800285006401961200450200652006561280128200800282002400842006520065111602011009910010016000010041181111011921611200621600001002006620066200662006620066
16020420065151110392580100100800001008000050064000002004502006420064322801002008000020024000020064200641116020110099100100160000100000001011221622200611600001002006520065200652006520065
1602042006415000039258010010080000100800005006400000200450200642006432280100200800002002400002006420064111602011009910010016000010001410001011221622200611600001002006520065200652006520065
16020420064150000392580100100800001008000050064249612004502006420064322801002008000020024000020064200641116020110099100100160000100000001011221622200611600001002006520065200652006520065
16020420064150000392580100100800001008000050064000002004502006420064322801002008000020024000020064200641116020110099100100160000100000001011221622200611600001002006520065200652006520065
16020420064150000392580100100800001008000050064000002004502006420064322801002008000020024000020064200641116020110099100100160000100000001011221622200611600001002006520065200652006520065
16020420064150000392580100100800001008000050064000002004502006420064322801002008000020024000020064200641116020110099100100160000100000001014421622200611600001002006520065200652006520065
160204200641510003925801001008000010080000500640000020045020064200643228010020080000200240000200642006411160201100991001001600001002160001011221622200611600001002006520065200652006520065
1602042006415000039258010010080000100800005006400000200450200642006432280100200800002002400002006420064111602011009910010016000010001440001011221622200611600001002006520065200652006520065
16020420064151000392580100100800001008000050064000002004502006420064322801002008000020024000020064200641116020110099100100160000100000001011221622200611600001002006520065200652006520065

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)9fa0a8acbranch mispredict (cb)cfd0d2icache miss (d3)d5d6d9dadbddinst fetch restart (de)e0? int output thing (e9)eaec? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
16002420080150045258001212800001280000626400000120027200462004632280012208000020240000200462004611160021109101016000010001801003231115202111082004321543160000102004720047200512004720047
16002420046150045258001212800001280000626400000120027200462004632280012208000020240000200462004611160021109101016000010003010030311720211610200432150160000102004720047200472004720047
1600242004615004525800121280000128000062640000112002720046200503228001220800002024000020050200461116002110910101600001000001003031182021174200432150160000102004720047200472005120150
1600242005015005125800121280000128000062640000112002720046200463228001220800002024000020046200461116002110910101600001000001003061172021147200472150160000102004720047200472005120047
1600242004615004525800121280000128000062640000112002720050200463228001220800002024000020046200461116002110910101600001000601003431142021147200432150160000102004720047200512004720047
1600242004615004525800121280000128000062640000112003120046200503228001220800002024000020046200461116002110910101600001000301002931172021174200432150160000102004720047200472004720047
16002420046150052025800121280000128000062640000112003120046200463228001220800002024000020046200461116002110910101600001000001003061142021147200472150160000102004720047200472005120051
16002420046150124525800121280000128000062642512012002720046200463228001220800002024000020046200461116002110910101600001000001002931172021147200432150160000102004720047200512004720047
16002420046150045258001212800001280000626400001120031200462004632280012208000020240000200462004611160021109101016000010001801003031152022177200432150160000102004720051200472005120047
160024200461500452580012128000012800006264000001200272004620046322800122080000202400002005020046111600211091010160000100012010030312420211410200432150160000102005120047200472004720051

Test 6: throughput

Count: 16

Code:

  sqrdmlah v0.2s, v16.2s, v17.2s
  sqrdmlah v1.2s, v16.2s, v17.2s
  sqrdmlah v2.2s, v16.2s, v17.2s
  sqrdmlah v3.2s, v16.2s, v17.2s
  sqrdmlah v4.2s, v16.2s, v17.2s
  sqrdmlah v5.2s, v16.2s, v17.2s
  sqrdmlah v6.2s, v16.2s, v17.2s
  sqrdmlah v7.2s, v16.2s, v17.2s
  sqrdmlah v8.2s, v16.2s, v17.2s
  sqrdmlah v9.2s, v16.2s, v17.2s
  sqrdmlah v10.2s, v16.2s, v17.2s
  sqrdmlah v11.2s, v16.2s, v17.2s
  sqrdmlah v12.2s, v16.2s, v17.2s
  sqrdmlah v13.2s, v16.2s, v17.2s
  sqrdmlah v14.2s, v16.2s, v17.2s
  sqrdmlah v15.2s, v16.2s, v17.2s
  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)03080b18191e1f373a3f4e51inst 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)9fa0a1a6a8a9acc2c5cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1602044007229900003601708502516063410016009810016000050024388650400200400394003919973319997160100200160000200480000400394004911160201100991001001600001000001000010110116114003601600001004005240040400404004040050
160204400393000000600010602516011710016001710016000050012800000400200400394004819973319997160100200160000200480000400394003911160201100991001001600001000001030010110116114004501600001004004040040400404004040040
160204400393000000150008502516010010016000010016000050012800000400200400394003919973319997160100200160000200480000400394003911160201100991001001600001000000030010110116114004501600001004004040040400404004040040
1602044004830000001501708502516010010016001710016000050023989990400330400394004819973319997160100200160000200480000400394004011160201100991001001600001000001030010110116114003601600001004004040040400404004040040
1602044003929900001500011602516011710016000010016000050012800000400200400394003919973320030160100200160000200480000400394004011160201100991001001600001000001030010110116114003601600001004004140040402474022840049
16020440039299000021000411425016146712216146812716152464623687280406180401654009820000319997160100200160000200480000400394003911160201100991001001600001000201030010110116114003601600001004004040040400404004040040
16020440039299000000180135502516010010016000010016000050012800001400200400394004819973319997160100200160000200480000400484003911160201100991001001600001000003002010110116114003701600001004004040040400404004040040
16020440039300003090371667582516010012216000010016000050012800001400210400394003919973320041160100200160374200481257400394024311160201100991001001600001000001015480010110216114004661600001004004040287400404004040040
160204400393000101150008502516010012416000010016000050012800001400200400484003919973319997160100200160335200481368400394016311160201100991001001600001002000054580010110116114003701600001004004040040400404004040049
160204400393000000000013202516010010016000010016000050023989991400200400394004819973320006160100200160000200480000400394003911160201100991001001600001000000000010110116114003601600001004004040041400524004040040

1000 unrolls and 10 iterations

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

retire (01)cycle (02)0307080a0b18191e1f373a3f4e51inst 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)5f6061696d6edispatch 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)9fa0a1a6a8a9acc2branch mispredict (cb)cfd0d2icache miss (d3)itlb miss (d4)d5d6d9dadbddinst fetch restart (de)e0? int output thing (e9)eaeb? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
160024400403000000009000460251602041016010010160000503575105015400214003940039199963200191600102016000020480000400394004911160021109101016000010000303001002411320101642211144003604012160000104004640040400404004040040
1600244003930000000000007302516002710160000101600005012800001154002040039400391999632002816001020160000204800004003940039111600211091010160000100000030010022841011162111110400360206160000104005740049400404004040040
1600244003930000000090007702516001010160018101600005012800001154002940039400391999632002816018220160106204800004003940039111600211091010160000100001000010022841012162111310400360206160000104024940049400404004040040
1600244003930000000000005202516001010160017101600006212800001154002040039400391999632001916001020160000204800004003940039111600211091010160000100001030010022114101116211912400360207160000104004340040400404004040050
1600244003930000000000004602516002810160168101600005012800001154002040039400391999632001916001020160000204800004003940103111600211091010160000100020030010022841010162111110400360207160000104004940049400404004040040
160024400393000001000018014302516002810160000101600005024827201154002140039400391999632001916001020160000204800004003940039111600211091010160000100020030010022841011162111112400360206160000104005240166402094027540040
160024400483000001101200031102516001010160000101600005012800001154002040039400391999682002816001020160000204800004003940049211600211091010160000100001430010022841012162111110400360206160000104013340049400494004940049
16002440039300001000120008802516002810160000101601065023989991154002040039400391999632001916001020160000204803214004840040111600211091010160000100001000010022841013162111110400360206160000104005840040400504004040040
16002440039300000000000051602516001011160000131600005013203251154002940101400392000532001916001020160000204800004003940039211600211091010160000100000200010042841012162111313400450206160000104007040040400404004040040
160024400393000100003030004642516001010160000101600005012800001154010840209400392001432010516018720160000204800004018340101111600211091010160000100002030010022841010162111110400360206160000104005240040400404004040092