Apple M1 Microarchitecture Research by Dougall Johnson

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

UQSHRN (4S)

Test 1: uops

Code:

  uqshrn v0.4h, v0.4s, #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)0307181e1f3f4e51inst 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)a6a8a9accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
10043037220000612547251000100010003981600301830373037241432895100010001000303730371110011000000073216222629100030383038303830383038
10043037230000612547251000100010003981601301830373037241432895100010001000303730371110011000010073216222629100030383038303830383038
10043037230000612547251000100010003981600301830373037241432895100010001000303730371110011000020073216222629100030383038303830383038
10043037230000612547251000100010003981600301830373037241432895100010001000303730371110011000000073216222629100030383038303830383038
100430372300003462547251000100010003981600301830373037241432895100010001000303730371110011000000073216222629100030383038303830383038
1004303723000061254725100010001000398160130183037303724143289510001000100030373037111001100001105773216222629100030383038303830383038
10043037230000612547251000100010003981601301830373037241432895100010001000303730371110011000000073216222629100030383038303830383038
10043037230000612547251000100010003981600301830373037241432895100010001000303730371110011000000073216222629100030383038303830383038
10043037220000612547251000100010003981601301830373037241432895100010001000303730371110011000000073216222629100030383038303830383038
10043037230000612547251000100010003981601301830373037241432895100010001000303730371110011000010073216222629100030383038303830383038

Test 2: Latency 1->2

Code:

  uqshrn v0.4h, v0.4s, #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)03091e3f4e5051inst 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)9fa7a8a9acc2branch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372250061295473002125101001001000010010000500427716003001830037300372826432874510100200100002001000030037300371110201100991001001000010000000071001611296330100001003003830038300383003830038
1020430037225006129547025101641001000010010000500427716003001830037300372826432874510100200100002001000030037300371110201100991001001000010000000071011611296330100001003003830038300383003830038
1020430037225006129547025101001001000010010000500427716003001830037300372826432874510100200100002001000030037300371110201100991001001000010000000071011611296330100001003003830038300383003830038
10204300372240061295470251010010010000100100005004277160030018300373003728264328745101002001000020010000300373003711102011009910010010000100027030071011611296330100001003003830038300813003830038
1020430037225006129547025101001001000010010000500427716003001830037300372826432874510100200100002001000030037300371110201100991001001000010024028502073511611296330100001003003830038300863013230038
10204300372251061295470251010010010000100100005004277160030018300373003728264328745101002001000020010000300373003711102011009910010010000100033030071011611296330100001003003830038300383003830038
10204300372250061295470251010010010000100106005004277160030018300373003728264328745101002001000020010000300373003711102011009910010010000100034000071011611296330100001003003830038300383003830038
10204300372250938429538025101291001000010010000500427716003001830037300372826432874510100200100002001000030037300371110201100991001001000010000030071011611296330100001003003830038300383003830038
1020430037225006129547098101001001000011110000500427716003001830037300372826432874510100200100002001000030037300371110201100991001001000010000000071011611296330100001003003830038300383003830038
1020430037225006129547025101001001000010010000500427716003001830037300372826432875010257200100002001000030037300371110201100991001001000010000000071011611296330100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)030b18191e1f3a3f4e51inst 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)9fa0a1a6a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024300372250000006129547251001010100001010150504277160030018300373003728286328767100102010000201000030037300371110021109101010000100003501206402162229629010000103003830038300383003830038
1002430037225000000612954725100101010000101000050427716003001830037300372828632876710010201000020100003003730037111002110910101000010000620006402162229629010000103003830038300383003830038
1002430037225000000612954725100101010000101000050427716013001830037300372828632876710010201000020100003003730037111002110910101000010000290906402162229629010000103003830038300383003830038
100243003722500000061295472510010101000010100005042771600300183003730037282863287671001020100002010000300373003711100211091010100001000010306402162229629010000103003830038300383003830038
100243003722500000061295472510010101000010100005042771600300183003730037282863287671001020100002010000300373003711100211091010100001000010306402162229629010000103003830038300383003830038
100243003722500000061295472510010101000012100005042771600300183003730037282863287671001020100002010000300373003711100211091010100001000000606402162229629010000103003830038300383003830038
10024300372250000007262954725100101010000101000050427716003001830037300372828632876710010201000020109843032130323711002110910101000010222141687907723652529966110000103032030323304053037130369
100243041622706692461614106294841531006818100561611198714282568030270303703036628309312891511065201114124109843031930417811002110910101000010002041938106402162229629010000103003830038300383003830038
10024300372250879246161459929493801007213100561111200724286624130270304043036828314332889911062261114222111553036930369811002110910101000010403021953147874733329944310000103037030416303663036730370
1002430371228087106526406492294572461005816100801811500554290680130378305093055528324502887211512201198428116363051130558911002110910101000010200023042206402162229629010000103003830038300383003830038

Test 3: throughput

Count: 8

Code:

  uqshrn v0.4h, v8.4s, #3
  uqshrn v1.4h, v8.4s, #3
  uqshrn v2.4h, v8.4s, #3
  uqshrn v3.4h, v8.4s, #3
  uqshrn v4.4h, v8.4s, #3
  uqshrn v5.4h, v8.4s, #3
  uqshrn v6.4h, v8.4s, #3
  uqshrn v7.4h, v8.4s, #3
  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)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)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? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8020420058150030258010810080008100800205006401321200202003920039997769990801202008003220080032200392003911802011009910010080000100001115118116200360800001002004020040200402004020040
8020420039150030258010810080008100800205006401320200202003920039997769990801202008003220080032200392003911802011009910010080000100001115118016200360800001002004020040200402004020040
80204200391500302580108100800081008002050064013212002020039200399977610024801202008003220080032200392003911802011009910010080000100001115118016200360800001002004020040200402004020040
8020420039150030258010810080008100800205006401321200202003920039997769990801202008003220080032200392003911802011009910010080000100361115118016200360800001002004020040200402004020040
8020420039150030258010810080008100800205006401321200202003920039997769990801202008003220080032200392003911802011009910010080000100001115118016200360800001002004020040200402004020040
8020420039150030258010810080008100800205006401320200202003920039997769990801202008003220080032200392003911802011009910010080000100001115118016200360800001002004020040200402004020040
8020420039150030258010810080008100800205006401320200202003920039997769990801202008003220080032200392003911802011009910010080000100001115118016200360800001002004020040200402004020040
80204200391500695258010810080008100800205006401320200202003920039997769990801202008003220080032200392003911802011009910010080000100101115118016200360800001002004020040200402004020040
80204200391500505258010810080008100800205006401321200202003920039997769990801202008003220080032200392003911802011009910010080000100101115118016200360800001002004020040200402004020040
8020420039150030258010810080008100800205006401321200202003920039997769990801202008003220080032200392003911802011009910010080000100001115118016200360800001002004020040200402004020040

1000 unrolls and 10 iterations

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

retire (01)cycle (02)03041e3f51inst 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)9fa0a8acc5cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420051150104025800101080000108000050640000020020200392003999963100198001020800002080000200392003911800211091010800001001005020111612162003680000102004020040200402004020040
8002420039150004025800101080000108000050640000020020200392003999963100198001020801402080000200392003911800211091010800001000005020171616132003680000102004020040200402004020040
8002420039150004025801071080000108000050640000020020200392003999963100198001020800002080000200902003911800211091010800001005005020141617132003680000102004020040200402004020040
800242003915000402580010108000010800005064000002002020039200399996310019800102080000208000020039200392180021109101080000100012015020161616132003680000102004020040200402004020040
8002420039150004025800101080000108000050640000020020200392003999963100198001020800002080000200392003911800211091010800001000005020121616122003680000102004020040200402004020040
80024200391500013525800101080000108000050640000020020200392003999963100198001020800002080000200392003911800211091010800001003005020141611122003680000102004020040200402004020040
8002420039150004025800101080000108000050640000020020200392003999963100198001020800002080000200392003911800211091010800001000005020161613132003680000102004020040200402004020040
80024200391500040258001010800001080000506400001200202003920039999631001980010208000020800002003920039118002110910108000010031605020111614142003680000102004020040200402004020040
8002420039150004025800101080000108000050640000020020200392003999963100198001020800002080000200392003911800211091010800001000005020161611162003680000102004020040200402004020040
8002420039150004025800101080000108000050640000120020200392003999963100198001020800002080000200392003911800211091010800001003005020161614142003680000102004020040200402004020040