Apple M1 Microarchitecture Research by Dougall Johnson

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

SHRN2 (8H)

Test 1: uops

Code:

  shrn2 v0.16b, v1.8h, #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)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)a8a9accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
1004303723061254825100010001000398313030183037303724153289510001000200030373037111001100000073116112630100030383038303830383038
1004303723061254825100010001000398313030183037303724153289510001000200030373037111001100000073116112630100030383038303830383038
1004303722061254825100010001000398313130183037303724153289510001000200030373037111001100000073116112630100030383038303830383038
1004303722061254825100010001000398313030183037303724153289510001000200030373037111001100000073116112630100030383038303830383038
1004303723061254825100010001000398313030183037303724153289510001000200030373037111001100000073116112630100030383038303830383038
1004303722061254825100010001000398313030183037303724153289510001000200030373037111001100000073116112630100030383038303830383038
1004303723061254825100010001000398313030183037303724153289510001000200030373037111001100000073116112630100030383038303830383038
1004303723061254825100010001000398313030183037303724153289510001000200030373037111001100000073116112630100030383038303830383038
1004303722061254825100010001000398313030183037303724153289510001000200030373037111001100000073116112630100030383038303830383038
1004303722061254825100010001000398313030183037303724153289510001000200030373037111001100000073116112630100030383038303830383038

Test 2: Latency 1->1

Code:

  shrn2 v0.16b, v1.8h, #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)030b1e1f3f4e51inst 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)9fa0a7a8acc2cdcfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020430037225000762954825101001001000010010000500427731313001830037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731313001830037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
10204300372250001052954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731313001830037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731303001830037300372826532874510100200100002002099030086300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037224000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037225000612954825101001001000010010000500427731303001830084300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
1020430037224000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03070b18191e1f3f4e51inst 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)9fa6a8acc5cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037225000000612954825100101010000121000050427731303001830037300372828703287671001220100002020000300373003711100211091010100001000006422162229630010000103003830038300383003830038
10024300372250000005362954825100101010000101000055427731313001830037300372828703287671001020106522020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
1002430037233000000612954825100171010000101000050427731303001830037300372828703287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
1002430037225000000612954825100101010000101000050427731313001830037300372828703287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
1002430037225000000612954825100101010000101000050427731313001830037300372828703287671001220100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
1002430037225000000612954825100101010000101000050427731303001830037300372828703287671001220100002020000300373003711100211091010100001000006402162229630210000103003830038300383003830038
1002430037224100000612954825100101010000101000050427731303001830037300372828703287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
1002430037225000000612954825100101010000101000050427731303001830037300372828703287671001020100002020000300373003711100211091010100001000006403162229630010000103003830038300383003830038
1002430037225000000612954825100101010000101000050427731303001830037300372828703287671001020101792020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
10024300372250000004412954825100101010000101000050427731303001830037300372828703287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038

Test 3: Latency 1->2

Code:

  shrn2 v0.16b, v0.8h, #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)031e3f4e51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa0a1a8a9acc5branch mispredict (cb)cdcfd6e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372250612954725101001001000010010000500427716030018300373003728271628740101002001000820020016300373003711102011009910010010000100000001117171629646100001003003830038300383003830038
10204300372250612954725101001001000010010000500427716030018300373003728271628741101002001000820020016300373003711102021009910010010000100000001117171629646100001003003830038300383003830038
10204300372250612954725101001001000010010000500427716030018300373003728271728740101002001000820020016300843003711102011009910010010000100020001117171629646100001003003830038300383003830038
10204300372250612954725101001001000010010000500427716030018300373003728271728740101002001000820020016300373003711102011009910010010000100000001117171629645100001003003830038300383003830038
10204300372250612954725101001001000010010000500427716030018300373003728271728740101002001000820020016300373003711102011009910010010000100000001117181629645100001003003830038300383003830038
10204300372250612954725101001001000010010000500427716030018300373003728271728740101002001000820020016300373003711102011009910010010000100004031117171629645100001003003830038300383003830038
10204300372250612954725101001001000010010000500427716030018300373003728271628740101002001000820020016300373003711102011009910010010000100000001117171629646100001003003830038300383003830038
10204300372250612954725101001001000010010000500427716030018300373003728271728741101002001000820020016300373003711102011009910010010000100000001117171629646100001003003830038300383003830038
10204300372250612954725101001001000010010000500427716030018300373003728271728741101002001000820020016300373003711102011009910010010000100000001117181629645100001003003830038300383003830038
10204300372250612954725101001001000010010000500427716030018300373003728271628741101002001000820020016300373003711102011009910010010000100000001117171629646100001003003830038300383003830038

1000 unrolls and 10 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)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)9fa6a8a9cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100243003722500612954725100101010000101000050427716003001830037300372828632876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038
100243003722400612954725100101010000101000050427716003001830037300372828632876710010201000020200003003730037111002110910101000010200640316332962910000103003830038300383003830038
100243003722500612954725100101010000101000050427716013001830037300372828632876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038
100243003722500612954725100101010000101000050427716003001830037300372828632876710010241000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038
100243003722500612954725100101010000101000050427716003001830037300372828632876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038
100243003722500612954725100101010000101000050427716003001830037300372828632876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038
100243003722500612954725100101010000101000050427716003001830037300372828632876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038
100243003722400822954725100101010000101000050427716003001830037300372828632876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038
1002430037225015612954725100101010000101000050427716003001830037300372829132876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038
100243003722500612954725100101010000101000050427716003001830037300372828632876710010201000020200003003730037111002110910101000010000640316332962910000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  movi v0.16b, 0
  shrn2 v0.16b, v8.8h, #3
  movi v1.16b, 0
  shrn2 v1.16b, v8.8h, #3
  movi v2.16b, 0
  shrn2 v2.16b, v8.8h, #3
  movi v3.16b, 0
  shrn2 v3.16b, v8.8h, #3
  movi v4.16b, 0
  shrn2 v4.16b, v8.8h, #3
  movi v5.16b, 0
  shrn2 v5.16b, v8.8h, #3
  movi v6.16b, 0
  shrn2 v6.16b, v8.8h, #3
  movi v7.16b, 0
  shrn2 v7.16b, v8.8h, #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)030b191e1f3f51inst 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)9fa0a7a8a9acc2c5branch mispredict (cb)cdcfd5d6inst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
160204200881500048029258011610080016100800285006401960200452006520065612801282008002820016005620065200651116020110099100100160000100000000111101190160200621600001002006620066200662006620066
16020420065151000029258011610080016100800285006401961200452006520065612801282008002820016005620065200651116020110099100100160000100000000111101190160200621600001002006620066200662006620066
160204200651500021029258011610080016100800285006401961200452006520065612801282008002820016005620065200651116020110099100100160000100000000111101190160200621600001002006620066200662006620066
160204200651500036029258011610080016100800285006401960200452006520065612801282008002820016005620065200651116020110099100100160000100000000111101190161200621600001002006620066200662006620066
1602042006515100330292580116100800161008002850064019602004520065200651479801282008002820016005620065200651116020110099100100160000100000000111101190160200621600001002006620066200662006620066
16020420065151001259829908022410080124100801325006401961200452014720147612801282008002820016005620065200651116020110099100100160000100000000111101190160200621600001002006620066200662006620066
160204200651500015029321802231008001610080028505641044020045201472014761280128200800282001600562006520147111602011009910010016000010040025300111101190160200621600001002045820164200662011520199
16020420065150116029258011610080016100802395006410520200452006520065612802332008024020016005620065200651116020110099100100160000100400000111101430160200621600001002006620149200662006620066
16020420065151020050487802211008011210080133500641036120118201542015513128012820280136202160056200652006511160201100991001001600001000010301111011911823200621600001002006620066200662006620066
16020420065150006071258011610080016100800285006401961200452006520065612801282008002820016005620065200651116020110099100100160000100000000111101190160200621600001002006620066200662006620066

1000 unrolls and 10 iterations

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

retire (01)cycle (02)031e3a3f51inst 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)5f60696b6d6escheduler 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)9fa8cfd0d2icache miss (d3)d5d6d9dadbddinst fetch restart (de)e0ea? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
16002420076150604252580010108000010800005064000011200420200462004632280010208000020160000200482004811160021109101016000010010048311252011118212004315160000102004720049201532004920049
1600242004815060452580010108000010800005064000011200380200462004632280010208000020160000200462005011160021109101016000010110043611182041118182004515160000102004920049200472004720053
16002420046151210452580010108000010800005064000011200470200462005032280010208000020160000200462004611160021109101016000010010045311232221123182004331160000102004720047200472004720047
1600242005015000452580010108000010800005064000011200470200462004632280010208000020160000200462004611160021109101016000010010042321192021220162004515160000102004920047200472004920049
16002420046150480732580010108000010800005064000011200380200462004632280010208000020160000200482004811160021109101016000010010042311202611120192004515160000102004920051200472004720047
160024200501503420452580010108000010800005064000011200390200462004832280010208000020160000200502004611160021109101016000010010040311192011118182004516160000102004720047200472004720138
1600242004815000452580010108000010800005064000011200390200462004832280010208000020160000200502004811160021109101016000010010045311232222119222004330160000102005320047200512004720047
1600242004815000452580010108000010800005064000011200370200462004632280010208000020160000200462004611160021109101016000010010040311172421121172004316160000102004720049200532004720047
16002420048150120452580010108000010800005064000011200370200482004832280010208000020160000200482004811160021109101016000010010043311182021121222004316160000102004720049200492004920047
1600242004615000452580010108000010800005064000001200460200462004632280010208000020160000200462004811160021109101016000010010043611192241120232004315160000102004720047200492004920049

Test 5: throughput

Count: 16

Code:

  shrn2 v0.16b, v16.8h, #3
  shrn2 v1.16b, v16.8h, #3
  shrn2 v2.16b, v16.8h, #3
  shrn2 v3.16b, v16.8h, #3
  shrn2 v4.16b, v16.8h, #3
  shrn2 v5.16b, v16.8h, #3
  shrn2 v6.16b, v16.8h, #3
  shrn2 v7.16b, v16.8h, #3
  shrn2 v8.16b, v16.8h, #3
  shrn2 v9.16b, v16.8h, #3
  shrn2 v10.16b, v16.8h, #3
  shrn2 v11.16b, v16.8h, #3
  shrn2 v12.16b, v16.8h, #3
  shrn2 v13.16b, v16.8h, #3
  shrn2 v14.16b, v16.8h, #3
  shrn2 v15.16b, v16.8h, #3
  movi v16.16b, 17

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.2503

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)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)9fa8acc5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
16020440058300030251601081001600081001600205001280132040020400394003919977061999016012020016003220032006440039400391116020110099100100160000100001111011801600400361600001004008940040400404004040040
16020440039299030251601081001600081001600205001280132040061400394003919977061999016012020016003220032006440039400391116020110099100100160000100001111011801600400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132140020400394003919977061999016012020016003220032006440039400391116020110099100100160000100001111011801600400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132140020400394003919977061999016012020016003220032006440039400391116020110099100100160000100001111011801600400361600001004004040040400404004040040
16020440039299030251601081001600081001600205001280132140020400394003919977061999016012020016003220032006440039400391116020110099100100160000100001111011801600400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132040020400394003919977061999016012020016003220032006440039400391116020110099100100160000100001111011801600400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132140020400394003919977061999016012020016003220032006440039400391116020110099100100160000100001111011801600400361600001004004040040400404004040040
16020440039299030251601081001600081001600205001280132140020400394003919977061999016012020016003220032006440039400391116020110099100100160000100001111011801600400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132140020400394003919977061999016012020016003220032006440039400391116020110099100100160000100001111011801600400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132040020400394003919977061999016012020016003220032006440039400391116020110099100100160000100001111011801600400361600001004004040040400404004040040

1000 unrolls and 10 iterations

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

retire (01)cycle (02)030b18191e1f3a3f51inst 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)5f6061696b6d6edispatch 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)91inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa1a6a8acc2cfd0d2icache miss (d3)d5d6d9dadbddinst fetch restart (de)e0eaeb? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
160024400512990000004625160010101600001016000050128000011540020040039400391999632001916001020160000203200004003940039111600211090101016000010000001002211216162127640036155160000104004040040400404004040040
16002440039300000000462516001010160000101600005012800001154002004003940039199963200191600102016000020320000400394003911160021109010101600001000000100228317162117740036155160000104004040040400404004040040
16002440039300000000462516001010160000101600005012800001154002004003940039199963200191600102016000020320000400394003911160021109010101600001000000100228227162124640036155160000104004040040400404004040040
16002440039300000000462516001010160000101600005012800000154002004003940039199968200191600102016000020320210400394013011160021109010101600001000000100228214162114440036156160000104010440040400404004040040
16002440039300100000942516001010160000101600005012800000154002004003940039199963200481600102016000020320636400914003911160021109010101600001000000100228217162114440036155160000104004040040400404004040040
160024400392990000005212516001010160000101600005012800001154002004003940039199963200191600102016000020320000400394003911160021109010101600001000000100228214162114340036155160000104004040040400404004040040
16002440039300000000462516001010160000101600005012800001154002004003940039199963200191600102016000020320000400394003911160021109010101600001000000100228214362116640036155160000104010240040400404004040040
160024400393000111013329681462516001010160000101600005012800001154031504003940039199963200191600102016000020320000400394003911160021109010101600001002030100228213162116440036155160000104004040040400404004040040
16002440039299000000462516001010160000101600005012800001154002004003940039199963200191600102016000020320000400394003911160021109010101600001000100100228213162114440036155160000104004040040400404025040040
160024400393000001200462516001010160000101600005012800001154002004003940039199963200191600102016000020320000400394003911160021109010101600001000000100228214162114340036305160000104004040040400404004040040