Apple M1 Microarchitecture Research by Dougall Johnson

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

SABD (vector, 4H)

Test 1: uops

Code:

  sabd v0.4h, v0.4h, v1.4h
  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)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)a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
1004303723012825482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372306125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372206125482510001000100039831303018303730372415328951000100020003037303711100110000673116112630100030383038303830383038
1004303723010325482510001000100039831313018303730372415328951000100020003037303711100110000373116112630100030383038303830383038
100430372306125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372206125482510001000100039831313018303730372415328951000100020003037303711100110001073116112630100030383038303830383038
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372336125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372206125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372306125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038

Test 2: Latency 1->2

Code:

  sabd v0.4h, v0.4h, v1.4h
  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)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)9fa1a8acc5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102043003722400612954825101001001000010010000500427731303001803003730037282727287411010020010008200200163003730037111020110099100100100001000701117180160029646100001003003830038300383003830038
102043003722500612954825101001001000010010000500427731303001803003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722500612954825101001001000010010000500427731313001803003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722400612954825101001001000010010000500427731313001803003730037282653287451010020010000200200003003730037111020110099100100100001000000007401161129634100001003003830038300383003830038
102043003722500612954825101001001000010010000500427731303001803003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722500612954825101001001000010010000500427731313001803003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722500612954825101001001000010010000500427731303001803003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722500612954825101001001000010010000500427731303001803003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300853003830038
102043003722500612954825101001001000010010000500427731303001803003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038
102043003722500612954825101001001000010010000500427731303001803003730037282653287451010020010000200200003003730037111020110099100100100001000000007101161129634100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)0307080b18191e1f3f4e51inst 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)9fa0a1a6a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024300372250000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
10024300372250000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
10024300372250000000612954825100101010000101000050427731303001803003730037282873287671001020100002020000300373003711100211091010100001000000006402168329904210000103036630367303693036030369
10024303582270106780461645262948514110044121002415108945642868121302700303193036828313292880410962221114020219523027430318811002110910101000010222121954827693724329846310000103036830370304183037030367
10024303662276117739661650282948516010079141005612108947742868120302700303683036628291392889811055201115024222983013230367811002110910101000010022002023006832943429884210000103037230368304213037030320
10024303682270010000612954825100101010000101000050427731303001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
10024300372250000000612954825100101010000101000050427731303001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
10024300372250000000612954825100101010000101000050427731303001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
10024300372240000000612954825100101010000101000050427731303001803003730037282873287671001020100002020000300373003711100211091010100001000000306402162229630010000103003830038300383003830038
10024300372250000000822954825100101010000101000050427731303001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038

Test 3: Latency 1->3

Code:

  sabd v0.4h, v1.4h, 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)03080b18191e1f3f4e51inst 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
1020430037225002000287029548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000009071011611296340100001003003830038300383003830038
1020430037225000000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000041090071011611296340100001003003830038300383003830038
1020430037224000000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000000174071011611296340100001003003830038300383003830038
102043003722500000061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000000099071011611296340100001003003830038300383003830038
102043003722500000061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000000090071011611296340100001003003830038300383003830038
102043003722500000061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000001000710116112963420100001003003830038300383003830038
10204300372250000006129548251010010010000100100005004277313030018300373013128265328745101002001000020020000300373003711102011009910010010000100000000071011611296340100001003003830038300383003830038
102043003722500000040002948516210191145100161451104373942868120302703032230372282863628800110322261114922022316301333035981102011009910010010000100220236224080848389212988630100001003032430410303613037130134
10204303752281077936176523829473182102051541006414511192744428816903030630417304192827637288901149322811327226229483032430419101102011009910010010000100002022211049082106122995836100001003046930465305103018130467
1020430359228019811887926129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100000000071011611296340100001003003830038300383003830038

1000 unrolls and 10 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)696b6d6edispatch 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? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037225061295482510010101000010100005042773133001803003730037282873287671001020100002020000300373003711100211091010100001003640316222963010000103003830038300383003830038
1002430037225061295482510010101000010100005042773133001803003730037282873287671001020100002020000300373003711100211091010100001009640216222963010000103003830038300383003830038
100243003722506129548251001010100001010000504277313300180300373003728287328767100102010000202000030037300371110021109101010000104815640216222963010000103003830038300383003830038
100243003722406129548251001010100001010000504277313300180300373003728287328767100102010000202000030037300371110021109101010000100177640216222963010000103003830038300383003830038
100243003722506129548251001010100001010000504277313300180300373003728287328767100102010000202000030037300371110021109101010000100114640216222963010000103003830038300383003830038
100243003722506129548251001010100001010000504277313300180300373003728287328767100102010000202000030037300371110021109101010000100129640216222963010000103003830038300383003830038
10024300372250612954825100101010000101000050427731330018030037300372828732876710010201000020200003003730037111002110910101000010012640216222963010000103003830038300383003830038
10024300372250612954825100101010000101000050427731330018030037300372828732876710010201000020200003003730037111002110910101000010069640216222963010000103003830038300383003830038
100243003722506129548251001010100001010000504277313300180300373003728287328767100102010000202000030037300371110021109101010000100177640216222963010000103003830038300383003830038
1002430037225061295482510010101000010100005042773133001803003730037282873287671001020100002020000300373003711100211091010100001006640216222963010000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  sabd v0.4h, v8.4h, v9.4h
  sabd v1.4h, v8.4h, v9.4h
  sabd v2.4h, v8.4h, v9.4h
  sabd v3.4h, v8.4h, v9.4h
  sabd v4.4h, v8.4h, v9.4h
  sabd v5.4h, v8.4h, v9.4h
  sabd v6.4h, v8.4h, v9.4h
  sabd v7.4h, v8.4h, v9.4h
  movi v8.16b, 9
  movi v9.16b, 10

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

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

retire (01)cycle (02)03070a0e1e3a3f51inst 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)9fa8acc5cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8020420058150112024825801001008000010080000500640000020020200392003999733999780100200800002001600002003920039118020110099100100800001001605114101691020036800001002004020040200402004020040
80204200391501100248258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010013051149169920036800001002004020040200402004020040
8020420039150110024825801001008000010080307500640000020020200392003999733999780100200800002001600002003920039118020110099100100800001000126051141116111120036800001002004020040200402004020040
802042003915011002482580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100060051149169920036800001002004020040200402004020040
80204200391501100248258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010010051147167920036800001002004020040200402004020040
80204200391501100248258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000051141216111120036800001002004020040200402004020040
80204200391501100248258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000051149169920036800001002004020040200402004020040
802042003915011002482580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100013505114101691020036800001002004020040200402004020040
8020420039150110302482580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100230511411169920036800001002004020040200402004020040
80204200391501100248258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000051149169920036800001002004020040200402004020040

1000 unrolls and 10 iterations

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

retire (01)cycle (02)0318191e1f3f51inst 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)9fa6a8acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)eb? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
800242004015000004025800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100000502071666200360080000102004020040200402004020040
800242003915000004025800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100000502051656200360080000102004020040200402004020040
800242003915000004025800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100000502081675200360080000102004020040200402004020040
800242003915000006325800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100000502051656200360080000102004020040200402004020040
800242003915000004025800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100000503861676200360580000102004020040200402004020040
8002420039150000040258001010800001080000506400000200202003920039999631001980010208000020160000200392003911800211091010800001000005020616552003601180000102004020040200402004020040
80024200391500000402580010108000010800005064000002002020091200399996111004780010208031520160608200962003911800211091010800001000005020616552008601180000102004020040200402004020040
8002420039150000040258001010800001080000506400000200202003920039999631001980010208000020160000200392003911800211091010800001000690502091786200360680000102004020040200402004020040
8002420039150000040258001010800001080000506400001200202003920039999631001980010208000020160000200392003911800211091010800001000660502081675200360680000102004020040200402004020040
800242003915000004025800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100000502061657200360680000102004020040200402004020040