Apple M1 Microarchitecture Research by Dougall Johnson

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

FRINTP (vector, 4H)

Test 1: uops

Code:

  frintp v0.4h, v0.4h
  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)030a1e1f3f4e51inst 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)9da6a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
1004303724000612547251000100010003981600301830373037241432895100010001000303730371110011000001373216222629100030383038303830383038
1004303723000612547251000100010003981600301830373037241432895100010001000303730371110011000000073216222629100030383038303830383038
10043037240006125472510001000100039816013018303730372414328951000100010003037303711100110000015073216222629100030383038303830383038
10043037240088912547251000100010003981601301830373037241432895100010001000303730371110011000000373216222629100030383038303830383038
1004303723000612547251000100010003981600301830373037241432895100010001000303730371110011000000073216222629100030383038303830383038
1004303724000612547251000100010003981601301830373037241432895100010001000303730371110011000000073216222629100030383038303830383038
1004303723000822547251000100010003981600301830373037241432895100010001000303730371110011000000073216222629100030383038303830383038
1004303723000612547251000100010003981601301830373037241432895100010001000303730371110011000000073216222629100030383038303830383038
1004303723000612547251000100010003981600301830373037241432895100010001000303730371110011000000073216222629100030383038303830383038
1004303723000612547251000100010003981600301830373037241432895100010001000303730371110011000000073216222629100030383038303830383038

Test 2: Latency 1->2

Code:

  frintp v0.4h, v0.4h
  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)03181e3f4e51inst 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)9fa8acc5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102043003723310726295472510100100100001001000050042771600300183003730084282716287411010020010008200100083003730037111020110099100100100001000011171701600296450100001003003830038300383003830038
1020430037233001240295472510100100100001001000050042771600300183003730037282716287411010020010008200100083003730037111020110099100100100001000011171801600296460100001003003830038300383003830038
1020430037233001204295472510100100100001001000050042771601300183003730037282643287451010020010000200100003003730037111020110099100100100001000000071011611296330100001003003830038300383003830038
1020430037233001158295472510100100100001001000050042771601300183003730037282643287451010020010000200100003003730037111020110099100100100001000000071011611296330100001003003830038300383003830038
1020430037232001089295472510100100100001001000050042771601300183003730037282643287451010020010000200100003003730037111020110099100100100001001300071011611296330100001003003830038300383003830038
102043003723300566295474410127100100001151000050042771601300183003730037282643287451010020010000200100003003730037111020110099100100100001000000071011611296330100001003003830038300383003830038
102043003723300922295472510100100100001001000050042771601300183003730037282643287451010020010000200100003003730037111020110099100100100001000000071011611296330100001003003830038300383003830038
1020430037232001065295472510100100100001001000050042771601300183003730037282643287451010020010000200100003003730037111020110099100100100001000000071011611296330100001003003830038300383003830038
1020430037233001240295472510100100100001001000050042771601300183003730037282643287451010020010000200100003003730037111020110099100100100001000000071011611296330100001003003830038300383003830038
1020430037232001037295472510100100100001001000050042771601300183003730037282643287451010020010000200100003003730037111020110099100100100001000000071011611296330100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)0318191e3a3f4e51inst 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)9fa6a8acc2branch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037225000061295472510010101000010100005042771601300183003730037282863287671001020100002010000300373003711100211091010100001000000640216222962910000103003830038300723003830038
1002430037225000061295472510010101000010100005042771601300183003730037282863287671001020100002010000300373003711100211091010100001009278100640216222962910000103003830038300383003830038
100243003722500017402295472510010101000010100005042771600300183003730037282863287671001020100002010000300373003711100211091010100001000000640216222962910000103003830038300383003830038
1002430037225000061295472510010101000010100005042771600300183003730037282863287671001020100002010000300373003711100211091010100001000000640216222962910000103003830038302263003830038
1002430037225000061295472510010101000010100005042771600300183003730037282863287671001020100002010000300373003711100211091010100001000000640216222962910000103003830038300383003830038
1002430037225000061295472510010101000010100005042771600300183003730037282863287671001020100002010000300373003711100211091010100001000000640216222962910000103003830038300383003830038
1002430037225000061295472510010101000010100005042771600300183003730037282863287671001020100002010000300373003711100211091010100001000000640216222962910000103003830038300383003830038
1002430037225000061295472510010101000010100005042771600300183003730037282863287671001020100002010000300373003711100211091010100001003000640216222962910000103003830038300383003830038
100243003722400450061295472510010101000010100005042771600300183003730037282863287671001020100002010000300373003711100211091010100001000000640216222969410000103003830038301333003830038
1002430037225000061295472510010101000010101505042771600300183003730037282863287671001020100002010000300373003711100211091010100001000000640216222962910000103003830038300383003830038

Test 3: throughput

Count: 8

Code:

  frintp v0.4h, v8.4h
  frintp v1.4h, v8.4h
  frintp v2.4h, v8.4h
  frintp v3.4h, v8.4h
  frintp v4.4h, v8.4h
  frintp v5.4h, v8.4h
  frintp v6.4h, v8.4h
  frintp v7.4h, v8.4h
  movi v8.16b, 9

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.2505

retire (01)cycle (02)030809181e1f3f51inst 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)9fa0a7a8a9acc5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
80204200391550000030258010810080008100800205006401322002020039200399977699908012020080032200800322003920039118020110099100100800001000000011151182160020036800001002004020040200402004020040
80204200391550000030258010810080008100800205006401322002020039200399977699908012020080032200800322003920039118020110099100100800001000000011151180160020080800001002004020040200402004020040
80204200391560000030258010810080008100800205006401322002020039200399977699908012020080032200800322003920039118020110099100100800001000000011151180160020036800001002004020040200402004020040
80204200391550000030258010810080008100800205006401322002020039200399977699908012020080032200800322003920039118020110099100100800001000000011151180160020036800001002004020040200402004020040
802042003915500000312258010810080008100800205006401322002020039200399977699908012020080032200800322003920039118020110099100100800001000000311151180160020036800001002004020040200402004020040
80204200391550000030258010810080008100800205006401322002020039200399977699908012020080032200800322003920039118020110099100100800001000000011151180160020036800001002004020040200402004020040
80204200391550000030258010810080008100800205006401322002020039200399977699908012020080032200800322003920039118020110099100100800001000000011151180160020036800001002004020040200402004020040
8020420039155000013230258010810080008100800205006401322002020039200399977699908012020080032200800322003920039118020110099100100800001000010011151180160020036800001002004020040200402004020040
80204200391550000072258010810080008100800205006401322002020039200399977699908012020080032200800322003920039118020110099100100800001000000011151180160020036800001002004020040200402004020040
80204200391550000030258010810080008100800205006401322002020039200399977699908012020080032200800322003920039118020110099100100800001000000011151180160020036800001002004020040200402004020040

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2505

retire (01)cycle (02)0304081e1f3f51inst 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)9fa0a1a8accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420039155101201101258001010800001080000506400000200202003920039999631001980010208000020800002003920039118002110910108000010000050202716142520036080000102004020040200402004020040
80024200391560000351258001010800001080000506400001200202003920039999631001980010208000020800002008920039118002110910108000010220050202616182320036080000102004020040200402004020040
800242003915500901290258010810800001080107506400001200202003920039999631001980010208000020800002003920039118002110910108000010000350202416262520036080000102004020040200402004020040
80024200391550000359258001010800001080000506400000200202003920039999631001980010208000020800002003920039118002110910108000010000050202316192620036080000102004020040200402004020040
80024200391550000407258001010800001080000506400000200202003920039999631001980010208000020800002003920039118002110910108000010000050202716252620036080000102004020040200402004020040
80024200391550000251258001010800001080000506400000200202003920039999631001980010208000020800002003920039118002110910108000010000050202516252620036080000102004020040200402004020040
80024200391550000191258001010800001080000506400000200202003920039999631001980010208000020800002003920039118002110910108000010001050202616132820036080000102004020040200402004020040
80024200391550000982258001010800001080000506400001200202003920039999631001980010208000020800002003920039118002110910108000010000050202516262420036080000102004020040200402004020040
800242003915500001682580010108000010800005064000002002020039200399996121001980010208000020800002003920092118002110910108000010000050202516222320036080000102004020040200402004020040
80024200391550012082258001010800001080000506400000200202003920039999631001980010208000020800002003920039118002110910108000010000050202616252720036080000102004020040200402004020040