Apple M1 Microarchitecture Research by Dougall Johnson

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

SQSHRUN2 (2D)

Test 1: uops

Code:

  sqshrun2 v0.4s, v1.2d, #3
  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)031e1f3f4e51inst 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
1004303722008225482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
10043037233015625482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
10043037232106125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723906125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
10043037232106125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
10043037233306125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038

Test 2: Latency 1->1

Code:

  sqshrun2 v0.4s, v1.2d, #3
  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)030818191e3f4e51inst 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)9fa8accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020430037225001111612954825101001001000010010000500427731313001830037300372826513287451010020010000200200003003730037111020110099100100100001000071021622296340100001003003830038300383003830038
1020430037224000061295392510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000071021622296340100001003003830038300383003830038
1020430037224000661295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000071021622296340100001003003830038300383003830038
1020430037225000361295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000071021622296340100001003003830038300383003830038
1020430037225000061295482510100100100001001000050042773131300183003730037282803287451010020010000200200003003730037111020110099100100100001000071031622296340100001003003830038301833003830038
10204300372250004861295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000071021622296340100001003003830038300383003830038
1020430037225000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000071021622296340100001003003830038300383003830038
1020430037225000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000071021622296340100001003003830038300383003830038
1020430037225000061295482510100100100001261000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000071021622296340100001003003830038300383003830038
1020430037225000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000071021622296340100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03081e1f3f4e51inst 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? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024300372250300612954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038
1002430037225000612954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038
10024300372250180612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038
100243003722500053629548251001010100001010000504277313130018300373003728287202876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038
10024300372250270612954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103008630038300383003830038
1002430037224000612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038
1002430037225030612954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038
1002430037225000612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038
1002430037225090612954825100101010000101000050427731313001830037300372829032876710010201000020200003003730037111002110910101000010103640216222963010000103003830038300383003830038
1002430037225055509352953925100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010000640216222963010000103003830038300383003830038

Test 3: Latency 1->2

Code:

  sqshrun2 v0.4s, v0.2d, #3
  movi v0.16b, 1

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)030b1e3f4e51inst 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)9fa8acc5branch mispredict (cb)cdcfd5d6e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102043003722500612954725101001001000010010000500427716003001803003730037282717287411010020010008200200163003730037111020110099100100100001000011171701629646100001003003830038300383003830038
1020430037225006129547251010010010000100100005004277160130018030037300372827110287541010020010008200200163003730037111020110099100100100001000011171801629646100001003003830038300383003830038
1020430037225007262954725101001001000010010000500427716013001803003730037282716287401010020010008200200163003730037111020110099100100100001000011171811629646100001003003830038300383003830038
102043003722500612954725101001001000010010000500427716013001803003730037282716287411010020010008200200163003730037111020110099100100100001000011171701629645100001003003830038300383003830038
1020430037225005362954725101001001000010010000500427716013001803003730037282717287401010020010008200200163003730037111020110099100100100001000011171701629646100001003003830038300383003830038
1020430037225001562952925101001001000010010000500427716013001803003730037282717287401010020010008200200163003730037111020110099100100100001000011171801629645100001003003830038300383003830038
102043003722410612954725101001001000010010000500427716013001803003730037282926287401010020010008200200163003730037111020110099100100100001000011171801629645100001003003830038300383003830038
102043003722400612954725101001001000010010000500427716003001803003730037282717287411010020010008200200163003730037111020110099100100100001000011171801629646100001003003830038300383003830038
102043003722500612954725101001001000010010000500427716003001803003730037282716287401010020010008200200163003730037111020110099100100100001000011171701629645100001003003830038300383003830038
102043003722500612954725101001001000010010000500427716013001803003730037282717287411010020010008200200163003730037111020110099100100100001000011171701629646100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03080a18191e3f4e51inst 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)9fa8acc2cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037225003036612954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000640716332962910000103003830038300383003830038
100243003722500000612954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038
1002430037225000001032954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000640316332970110000103003830038300383003830038
1002430037225000024612954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038
1002430037225000012612954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010302640316332962910000103003830038300383003830038
100243008422500000612954725100101010008101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010030640316332962910000103003830038300383003830038
1002430037225000012612954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038
100243003722500000612954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038
100243003722400000612954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038
100243003722511000612954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  movi v0.16b, 0
  sqshrun2 v0.4s, v8.2d, #3
  movi v1.16b, 0
  sqshrun2 v1.4s, v8.2d, #3
  movi v2.16b, 0
  sqshrun2 v2.4s, v8.2d, #3
  movi v3.16b, 0
  sqshrun2 v3.4s, v8.2d, #3
  movi v4.16b, 0
  sqshrun2 v4.4s, v8.2d, #3
  movi v5.16b, 0
  sqshrun2 v5.4s, v8.2d, #3
  movi v6.16b, 0
  sqshrun2 v6.4s, v8.2d, #3
  movi v7.16b, 0
  sqshrun2 v7.4s, v8.2d, #3
  movi v8.16b, 9

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.2508

retire (01)cycle (02)033f51inst 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)9fa8acc5branch mispredict (cb)cdcfd6e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1602042009015129258011610080016100800285006401960200452006520065612801282008002820016005620065200651116020110099100100160000100001111011916200621600001002006620066200662006620066
1602042006515029258011610080016100800285006401960200452006520065612801282008002820016005620065200651116020110099100100160000100001111011916200621600001002006620066200662006620066
1602042006515029258011610080016100800285006401960200452006520065612801282008002820016005620065200651116020110099100100160000100001111011916200621600001002006620066200662014920066
1602042006515029258011610080016100800285006401960200452006520065612801282008002820016005620065200651116020110099100100160000100001111011916200621600001002006620066200662006620066
1602042006515029258011610080016100800285006401960200452006520065612801282008002820016005620065200651116020110099100100160000100001111011916200621600001002006620066200662006620066
1602042006515029258011610080016100800285006401960200452006520065612801282008002820016005620065200651116020110099100100160000100001111011916200621600001002006620066200662006620066
1602042006515029258011610080016100800285006401960200452006520065612801282008002820016005620065200651116020110099100100160000100001111011916200621600001002006620066200662006620066
1602042006515029258011610080016100800285006401960200452006520065612801282008002820016005620065200651116020110099100100160000100001111011916200621600001002006620066201472006620066
1602042006515029258011610080016100800285006401960200452006520065612801282008002820016005620065200651116020110099100100160000100001111011916200621600001002006620066200662006620066
1602042006515029258011610080016100800285006401960200452006520065612801282008002820016005620065200651116020110099100100160000100001111011916200621600001002006620066200662006620066

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2529

retire (01)cycle (02)03081e3f51inst 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)5f6061696d6escheduler 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)9fa1a8acc2cfd0d2icache miss (d3)d5d6d9dadbddinst fetch restart (de)e0ea? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
160024204091500018712580010108000010800005064000011520027200462004632280010208000020160000200462004611160021109101016000010000010031821102121112122004315160000102004720047200472004720047
16002420240150004525800101080000108000050640000115200272004620046322800102080000201600002004620046111600211091010160000100000100308217202111172004315160000102004720047200472004720047
16002420277150030512580010108000010800005064000011520027200502004632280010208000020160000200462005011160021109101016000010001170100368218202118132004315160000102004720047200472004720047
16002420241150005125800101080000108000050640000115200312005020050322800102080000201600002004620050111600211091010160000100000100341132112442211142004730160000102005120051200472005120051
16002420217150002372580010108000010800005064000001520027200462004632280010208000020160000200462004611160021109101016000010000010040831142041213132004330160000102005120047200512005120047
160024202501500071025800101080000108000050640000115200272004620046322800102080000201600002004620050111600211091010160000100060010035821132421113132004315160000102004720047200472005120047
1600242023615000452580010108000010800005064000011520027200462004632280010208000020160000200462005011160021109101016000010000010031112212204111172004330160000102004720047200472004720047
1600242023615003452580010108000010800005064000011520027200462004632280010208000020160000200502004611160021109101016000010000010038113211202229132004715160000102004720051200512005120051
1600242021515000452580010108000010800005064000011520031200502004632280010208000020160000200502005011160021109101016000010000010034113211244211382004730160000102005120051200512005120051
160024202201500051258001010800001080000506400000152003120050200503228001020800002016000020050200501116002110910101600001000156010036112213204228142004715160000102004720047200472004720047

Test 5: throughput

Count: 16

Code:

  sqshrun2 v0.4s, v16.2d, #3
  sqshrun2 v1.4s, v16.2d, #3
  sqshrun2 v2.4s, v16.2d, #3
  sqshrun2 v3.4s, v16.2d, #3
  sqshrun2 v4.4s, v16.2d, #3
  sqshrun2 v5.4s, v16.2d, #3
  sqshrun2 v6.4s, v16.2d, #3
  sqshrun2 v7.4s, v16.2d, #3
  sqshrun2 v8.4s, v16.2d, #3
  sqshrun2 v9.4s, v16.2d, #3
  sqshrun2 v10.4s, v16.2d, #3
  sqshrun2 v11.4s, v16.2d, #3
  sqshrun2 v12.4s, v16.2d, #3
  sqshrun2 v13.4s, v16.2d, #3
  sqshrun2 v14.4s, v16.2d, #3
  sqshrun2 v15.4s, v16.2d, #3
  movi v16.16b, 17

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

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

retire (01)cycle (02)03181e1f3f51inst 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)9fa0a1a6a8acc2c5branch mispredict (cb)cdcfd5d6e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1602044006030000030251601081001600081001600205001280132400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000000011110118516400361600001004004040040400404004040040
1602054003930006030251601081001600081001600205001280132400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000000011110118016400361600001004004040040400404004040040
1602044003930000030251601081001600081001600205001280132400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000000011110118016400361600001004004040040400404004040040
1602044003929900030251601081001600081001600205001280132400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000000011110118016400361600001004004040040400404004040040
1602044003930000030251601081001600081001600205001280132400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000000011110118016400361600001004004040040400404004040040
16020440039300000324251601081001600081001600205001280132400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000000011110118016400361600001004004040040400404004040040
16020440039300000220251601081001600081001600205001280132400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000016011110118016400361600001004004040040400404004040040
160204400393000007225160108100160008100160020500128013240020401004003919977619990160120200160032200320064400394010011160201100991001001600001000000473011110118016400361600001004004040040400404004040040
1602044003929900030251601081001600081001600205001280132400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000000011110118016400361600001004004040040400404004040040
1602044008929900030251601081001600081001600205001280132400204003940039199776199901601202001600322003200644003940039111602011009910010016000010000000011110118016400361600001004004040040400404004040198

1000 unrolls and 10 iterations

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

retire (01)cycle (02)03071e3f51inst 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)9fa1a8acc5cfd0d2icache miss (d3)d5d6d9dadbddinst fetch restart (de)e0eaebec? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
16002440039300004625160010101600001016000050128000011400204003940039199963200191600102016000020320000400394003911160021109101016000010000010022312231621119184003615100160000104004040040400404004040040
16002440039299005225160010101600001016000050128000011400204003940039199963200191600102016000020320000400394003911160021109101016000010000010024312201641119194003615100160000104004040040400404004040040
1600244003930000462516001010160000101600005012800001140020400394003919996320019160010201600002032000040039400391116002110910101600001000301002362115162211918400363050160000104004040040400404004040040
16002440039299004625160010101600001016000050128000001400204003940039199963200451600102016000020320000400394003911160021109101016000010000010022321201621118184003630100160000104004040040402444004040040
16002440039300004625160010101600001016000050128000011400204003940039199963200191600102016000020320000400394003911160021109101016000010000010022611201622218184003630100160000104004040040400404004040040
160024400393000071125160010101600001016000050128000011400204003940039199963200461600102016000020320000400394003911160021109101016000010000010024312151621118174003615100160000104004040040400404004040040
1600244003930000462516001010160000101600005012800000140020400394003919996320019160010201600002032000040039400391116002110910101600001000301002231217162111717400361550160000104004040040400404004040040
16002440039300005225160010101600001016000050128075611400204003940039199963200191600102016000020320000400394003911160021109101016000010003010024311181621120154003615100160000104004040040400404004040040
1600244003930000462516001010160000101600005012800001140020400394003919996320019160010201600002032000040039400391116002210910101600001000001002261120162211914400361550160000104004040040402504004040040
1600244003930000522516001010160000101600005012800000140020400394003919996320019160010201600002032000040039400391116002110910101600001000001002261220162121821400363050160000104004040040400404004040040