Apple M1 Microarchitecture Research by Dougall Johnson

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

FRINTX (vector, 4H)

Test 1: uops

Code:

  frintx 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)0308090b181e1f3a3f4e51inst 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)a1a8a9acc2cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
100430372200000008225472510001000100039816013018303730372414328951000100010003037303711100110000000073116112629100030383038303830383038
100430372200000006125472510001000100039816013018303730372414328951000100010003037303711100110000000073116112629100030383038303830383038
100430372200000006125472510001000100039816013018303730372414328951000100010003037303711100110000000073116112629100030383038303830383038
1004303722000000010325472510001000100039816013018303730372414328951000100010003037303711100110000000073116112629100030383038303830383038
100430372200000006125472510001000100039816013018303730372414328951000100010003037303711100110000102638094124112645100030383085308630383085
100430842301011320017725384910001008115039816013054303730372418828951000116211633084303721100110000002618273116112675100030853038308530743085
1004308523111114188110325384210081000100039951213018303730852415328951000116211623084308521100110004042815073116112629100030383038303830383038
100430372300000008425472510001000100039816013018303730372414328951000100010003037303711100110000000073116112629100030383038303830383038
1004303723000000012625472510001000100039816013018303730372414328951000100010003037303711100110000000073116112629100030383038303830383038
10043037220000120025125472510001000100039816013018303730372414328951000100010003037303711100110000000073116112629100030383038303830383038

Test 2: Latency 1->2

Code:

  frintx 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)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)9fa1a6a9acc2c5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020430037224025506129547251010010010000100100005004277160030018300373003728271628740101002001000820010008300373003711102011009910010010000100000001117180160029646100001003003830038300383003830038
10204300372240006129547251010010010000100100005004277160130018300373003728271628745101002001000020010000300373003711102011009910010010000100000000007101161129633100001003003830038300383003830038
10204300372250006129547251010010010000100100005004277160130018300373003728264328745101002001000020010000300373003711102011009910010010000100000000007101161129633100001003003830038300383003830038
102043003722501506129547251010010010000100100005004277160130018300373003728264328745101002001000020010000300373003711102011009910010010000100000020007101161129633100001003003830038300383003830038
10204300372250006129547251010010010000100100005004277160030018300373003728264328745101002001000020010000300373003711102011009910010010000100000000007101161129695100001003003830038300383003830038
10204300372250006129547251010010010000100100005004277160030018300373003728264328745101002001000020010000300373003711102011009910010010000100000000007101161129633100001003003830038300383003830038
10204300372250006129547251010010010000100100005004277160030018300373003728264328745101002001000020010000300373003711102011009910010010000100000000007101161129633100001003003830038300383003830038
10204300372250006129547251010010010000100100005004277160030018300373003728264328745101002001000020010000300373003711102011009910010010000100000000007101161129633100001003003830038300383003830038
10204300372250006129547251010010010000100100005004277160030018300373003728264328745101002001000020010000300373003711102011009910010010000100001000007101161129633100001003003830038300383003830038
102043003722503072629547251010010010000100100005004277160030018300373003728264328745101002001000020010000300373003711102011009910010010000100000000007101161129633100001003003830038300383003830038

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)9fa0a1a7a8a9acc2c5cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100243003722500045001032954725100101010000101000050427716013001830037300372828632876710010201000020100003003730037111002110910101000010000100006402162229629010000103003830038300383003830038
100243003722400012001032954725100101010000101000050427716013001830037300372828632876710010201000020100003003730037111002110910101000010000103006402162229629010000103003830038300383003830038
1002430037225000600612954725100101010000101000050427716003001830037300372828632876710010201000020100003003730037111002110910101000010000100006402162229629010000103003830038300383003830038
100243003722500036001032954725100101010000101000050427716013001830037300372828632876710010201000020100003003730037111002110910101000010000103006402162229629010000103003830038300383003830038
1002430037225000000612954725100101010000101000055427716013001830037300372828632876710010201000020100003003730037111002110910101000010020100006402162229629110000103003830038300383008530084
100243003722400048001032953845100191010000101015050427986413005430037300372828632878510010201000020100003003730037111002110910101000010000103006402162229629010000103003830038300383003830038
1002430037225000300612954725100101010000101000050427716013001830037300372828632876710010201000020100003003730037111002110910101000010000103006402162229629010000103003830038300383003830038
100243003722500036008182954725100101010000101000050427716013001830083300372828632876710010201000020100003003730037411002110910101000010000120006402162229629010000103003830038300383003830038
1002430037225000510038942952015410037181004013104506042812161300183003730037282863287671001020100002010000300373003721100211091010100001024110193082068241012229629010000103003830038300383003830038
100243003722501711708811335294821171002810100001010000504277160130018300373003728286328786112172010000241032430037300371110021109101010000102411225201207482162229629010000103018030038300853017930038

Test 3: throughput

Count: 8

Code:

  frintx v0.4h, v8.4h
  frintx v1.4h, v8.4h
  frintx v2.4h, v8.4h
  frintx v3.4h, v8.4h
  frintx v4.4h, v8.4h
  frintx v5.4h, v8.4h
  frintx v6.4h, v8.4h
  frintx 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)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)cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
802042003915090302580108100800081008002050064013212002020039200399977699908012020080032200800322003920039118020110099100100800001000011151184163420036800001002004020040200402004020040
802042003915004102580108100800081008002050064013212002020039200399977699908012020080032200800322003920039118020110099100100800001000011151184163320036800001002004020040200402004020040
80204200391500302580108100800081008002050064013212002020039200399977699908012020080032200800322003920039118020110099100100800001000011151183163320086800001002004020040200402004020040
80204200391506302580108100800081008002050064013212002020039200399977699908012020080032200800322003920039118020110099100100800001000011151184164320036800001002004020040200402004020040
80204200391500302580108100800081008002050064013212002020039200399977699908012020080032200800322003920039118020110099100100800001000011151183164420036800001002004020040200402004020040
80204200391500302580108100800081008002050064013212002020039200399977699908012020080032200800322003920039118020110099100100800001000011151184164320036800001002004020040200402004020040
802042003915030302580108100800081008002050064013212002020039200399977699908012020080032200800322003920039118020110099100100800001000011151183163420036800001002004020040200402004020040
8020420039150453302580108100800081008002050064013212002020039200399977699908012020080032200800322003920039118020110099100100800001000011151183164320036800001002004020040200402004020040
80204200391500302580108100800081008002050064013212002020039200399977699908012020080032200800322003920039118020110099100100800001000011151183163320036800001002004020040200402004020040
80204200391500302580108100800081008002050064013212002020039200399977699908012020080032200800322003920039118020110099100100800001000011151184163420036800001002004020040200402004020040

1000 unrolls and 10 iterations

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

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)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)9fa1a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
80024200391500120822580010108000010800005064000012002020039200399996310019800102080000208000020039200391180021109101080000100135020116112003680000102004020040200402004020040
800242003915003604025800101080000108000050640000020020200392003999963100198001020800002080000200392003911800211091010800001000215020116112003680000102004020040200402004020040
80024200391500270822580010108000010800005064000002002020039200399996310019800102080000208000020039200391180021109101080000100105020116112003680000102004020040200402004020040
80024200391490300402580010108000010800005064000002002020039200399996310019800102080000208000020039200391180021109101080000100135020116112003680000102004020040200402004020040
8002420039150000822580010108000010800005064000012002020039200399996310019800102080000208000020039200391180021109101080000100135020116112003680000102004020040200402004020040
800242003915001860822580010108000010800005064000012002020039200399996310019800102080000208000020039200391180021109101080000100135020116112003680000102004020040200402004020040
80024200391500300402580010108000010800005064000012002020039200399996310019800102080000208000020039200391180021109101080000100135020116112003680000102004020040200402004020040
80024200391500390822580010108000010800005064000002002020039200399996310019800102080000208000020039200391180021109101080000100035020116112003680000102004020040200402004020040
800242003915003090822580010108000010800005064000012002020039200399996310019800102080000208000020039200391180021109101080000100005020116112003680000102004020040200402004020040
8002420039150090402580010108000010800005064000012002020039200399996310019800102080000208000020039200391180021109101080000100135020116112003680000102004020040200402004020040