Apple M1 Microarchitecture Research by Dougall Johnson

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

BCAX (vector, 16B)

Test 1: uops

Code:

  bcax v0.16b, v0.16b, v1.16b, v2.16b
  movi v0.16b, 1
  movi v1.16b, 2
  movi v2.16b, 3

(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)03093f4e51inst 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)a0a8acc2cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
10042037150611687251000100010002646800201820372037157231895100010003000203720371110011000000073216111787100020382038203820382038
10042037150611687251000100010002646801201820372037157231895100010003000203720371110011000000073116111787100020382038203820382038
10042037150611687251000100010002646801201820372037157231895100010003000203720371110011000000073116111787100020382038203820382038
10042037150611687251000100010002646801201820372037157231895100010003000203720371110011000000073116111787100020382038203820382038
10042037160611687251000100010002646801201820372037157231895100010003000203720371110011000000073116111787100020382038203820382038
10042037160611687251000100010002646800201820372037157231895100010003000203720371110011000000073116111787100020382038203820382038
10042037150611687251000100010002646801201820372037157231895100010003000203720371110011000003073116111787100020382038203820382038
10042037160611687251000100010002646801201820372037157231895100010003000203720371110011000000073116111787100020382038203820382038
10042037160611687251000100010002646801201820372037157231895100010003000203720371110011000000073116111787100020382038203820382038
10042037160611687251000100010002646801201820372037157231895100010003000203720371110011000000073116111787100020382038203820382038

Test 2: Latency 1->2

Code:

  bcax v0.16b, v0.16b, v1.16b, v2.16b
  movi v0.16b, 1
  movi v1.16b, 2
  movi v2.16b, 3

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 2.0037

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)9fa1a6a8accfd0d5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020420037150000010319687251010010010000100100005002847680120018200372003718422318745101002001000020030000200372003711102011009910010010000100000071002162219791100001002003820038200382003820038
1020420037150000045519687251010010010000100100005002847680120018200372003718422318745101002001000020030000200372003711102011009910010010000100000071002162219791100001002003820038200382003820038
102042003715000006119687251010010010000100100005002847680120018200372003718422318745101002001000020030000200372003711102011009910010010000100000071012162219791100001002003820038200382003820038
102042003715000006119687251010010010000100100005002847680020018200372003718422318745101002001000020030000200372003711102011009910010010000100201073412162219791100001002003820038200382003820038
102042003715000016119687251010010010000100100005002847680120018200372003718422318745101002001000020030000200372003711102011009910010010000100000071012162219791100001002003820038200382003820038
102042003715000008419687251010010010000100100005002847680020018200372003718422318745101002001000020030000200372003711102011009910010010000100000071012162219791100001002003820038200382003820038
1020420037150000061196872510100100100001001000050028476800200182003720037184223187451010020010000200300002003720037111020110099100100100001000038071012162219791100001002003820038200382003820038
10204200371500001124519687251010010010000100100005002847680020018200372003718422318745101002001000020030000200372003711102011009910010010000100000071012162219791100001002003820038200382003820038
1020420037150000019119687251010010010000100100005002847680020018200372003718422318745101002001000020030000200372003711102011009910010010000100000071012162219791100001002003820038200382003820038
102042003715000183073219687251010010010000100100005002847680020018200372003718422318745101002001000020030000200372003711102011009910010010000100000071012162219791100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)03080b18191e1f3a3f4e51inst 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)9fa0a1a6a7a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100242003715000009001031968725100101010000101000050284768002001820037200371844431876710010201000020300002003720037111002110910101000010000010006403162219785010000102003820038200382003820038
10024200371500000000611968725100101010000101000050284768002001820037200371844431876710010201000020300002003720037111002110910101000010000000006402162219785010000102003820038200382003820038
10024200371500000000611968725100101010000101000050284768002001820037200371844431876710010201000020300002003720037111002110910101000010000010006402163219785010000102003820038200382003820038
10024200371500000000611968725100101010000101000050284768002001820037200371844431876710010201000020305042003720037111002110910101000010030000006402162219785010000102003820038200382003820038
10024200841500100010804461968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010000010026402162219785010000102003820038200382003820038
100242003715000000002511968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010000000006402162219785010000102003820038200382003820038
10024200371500000000611968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010000000006402162319785010000102003820038200382003820038
10024200371500000000611968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010000000006402162119785010000102003820038200382003820038
1002420037150000012001871968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010000010006402162319785010000102003820038200862003820038
1002420037150001114400611968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010000000006402162219785010000102003820038200382003820038

Test 3: Latency 1->3

Code:

  bcax v0.16b, v1.16b, v0.16b, v2.16b
  movi v0.16b, 1
  movi v1.16b, 2
  movi v2.16b, 3

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)03191e3a3f4e51inst 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)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)9fa1a8accfd0d5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020420037150000841968725101001001000010010000500284768020018200372003718422031874510100200100002003000020037200371110201100991001001000010000071002162219791100001002003820038200382003820038
10204200371500004851968725101001001000010010000500284768020018200372003718422031874510100200100002003000020037200371110201100991001001000010001371002162219791100001002003820038200382003820038
10204200371500004661968725101001001000010010000500284768020018200372003718422031874510100200100002003000020037200371110201100991001001000010000071012162219791100001002003820038200382003820038
10204200371500001491968725101001001000010010000500284768020018200372003718422031874510100200100002003000020037200371110201100991001001000010000071012162219791100001002003820038200382003820038
10204200371500001491968725101001001000010010000500284768020018200372003718422031874510100200100002003000020037200371110201100991001001000010000071012162219791100001002003820038200382003820038
10204200371500001261968725101001001000010010000500284768020018200372003718422031874510100200100002003000020037200371110201100991001001000010000071012162219791100001002003820038200382003820038
10204200371500004801968725101001001000010010000500284768020018200372003718422031874510100200100002003000020037200371110201100991001001000010000071012162219791100001002003820038200382003820038
102042003715000013011968725101001001000010010000500284768020018200372003718422031874510100200100002003000020037200371110201100991001001000010000071012163219791100001002003820038200382003820038
1020420037150000611968725101001001000010010000500284768020018200372003718422031874510100200100002003000020037200371110201100991001001000010000071012162219791100001002003820038200382003820038
1020420037150000611968725101001001000010010000500284768020018200372003718422031874510100200100002003000020037200371110201100991001001000010000671012163219791100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)030718191e3f4e51inst 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
10024200371501006611968725100101010012101000050284768002001820037200371844431876710010201000020300002003720037111002110910101000010000640316221978510000102003820038200382003820038
100242003715000001661968725100101010000101000050284768002001820037200371844431876710010201000020300002003720037111002110910101000010000640216121978510000102003820038200382003820038
10024200371500000611968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010000662216221978510000102003820038200382008420038
10024200371500000611968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010000640216221978510000102003820038200382003820038
10024200371500000611968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010050640216221978510000102003820038200382003820038
10024200371500000611968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010000640216321978510000102003820038200382003820038
100242003715000001491968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010000640216221978510000102003820038200382003820038
10024200371500000611968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010010640216221978510000102003820038200382003820038
100242003715000003451968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010000640216221978510000102003820038200382003820038
100242003715000003691968725100101010000101000050284768012001820037200371844431876710010201000020300002003720037111002110910101000010000640216221978510000102003820038200382003820038

Test 4: Latency 1->4

Code:

  bcax v0.16b, v1.16b, v2.16b, v0.16b
  movi v0.16b, 1
  movi v1.16b, 2
  movi v2.16b, 3

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)030818191e3a3f4e51inst 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)9fa8a9acc2cfd0d5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204200371500000014719687251010010010000100100005002847680020018200372003718422318745101002001000020030000200372003711102011009910010010000100000071002163219791100001002003820038200382003820038
1020420037150000006119687251010010010000100100005002847680020018200372003718422318745101002001000020030000200372003711102011009910010010000100000071002162219791100001002003820038200382003820038
1020420037150000006119687251010010010000100100005002847680020018200372003718422318745101002001000020030000200372003711102011009910010010000100000071002163219791100001002003820038200382003820038
10204200371500000032819687251010010010000100100005002847680020018200372003718422318745101002001000020030000200372003711102011009910010010000100000071002162219791100001002003820038200382003820038
1020420037150000006119687251010010010000100100005002847680120018200372003718422318745101002001000020030000200372003711102011009910010010000100000071003162219791100001002003820038200382003820038
10204200371500000014919687251010010010000100100005002847680120018200372003718422318745101002001000020030000200372003711102011009910010010000100000071013162219791100001002003820038200382003820038
10204200371500000063119687251010010010000100100005002847680120018200372003718422318745101002001000020030000200372003711102011009910010010000100400071012162219791100001002003820038200382003820038
10204200371500000016619687251010010010000100100005002847680020018200372003718422318745101002001000020030000200372003711102011009910010010000100106071012162219791100001002003820038200382003820038
10204200371500000010319687251010010010000100100005002847680120090200372003718426318745102562001000020030498200372003711102011009910010010000100000071012162219791100001002003820038200382003820038
10204200371500000015119687251010010010000102100005002847680120018200852003718422318745102562001000020030000200372003711102011009910010010000100000071012162219791100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)03081e3f4e51inst 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)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)9fa0a9acbranch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024200371500061196872510010101000010100005028476802001820037200371844401918767100102010000203048920086200371110021109101010000100000661216221978510000102003820038200382003820038
1002420037150006119687251001010100001010000502847680200182003720037184440318767100102010000203000020037200371110021109101010000100000640216221978510000102003820038200382003820038
1002420037150006119687251001010100001010000502847680200182003720037184440318767100102010000203000020037200371110021109101010000100000640216221978510000102003820038200382003820038
1002420037150006119687251001010100001010000502847680200182003720037184440318767100102010000203000020037200371110021109101010000100000640216221978510000102003820038200382003820038
1002420037150006119687251001010100001010000502847680200182003720037184447318767100102010000203000020037200371110021109101010000100000640216221978510000102003820038200382003820038
1002420037150006119687251001010100001010000502847680200182003720037184440318767100102010000203000020037200371110021109101010000100000640216221978510000102003820038200382003820038
1002420037150006119687251001010100001010000502847680200182003720037184440318767100102010000203000020037200371110021109101010000100000640216221978510000102003820038200382003820038
1002420037150006119687251001010100001010000502847680200182003720037184440318767100102010000203000020037200371110021109101010000100000640216221978510000102003820038200382003820038
10024200371500248419667251001010100001010000502847680200182003720037184440318767100102010000203000020037200371110021109101010000100000640216321978510000102003820038200382003820038
10024200371490186119687251001010100001010000502847680200182003720037184440318767100102010000203000020037200371110021109101010000100000640216221978510000102003820038200382003820038

Test 5: throughput

Count: 8

Code:

  bcax v0.16b, v8.16b, v9.16b, v10.16b
  bcax v1.16b, v8.16b, v9.16b, v10.16b
  bcax v2.16b, v8.16b, v9.16b, v10.16b
  bcax v3.16b, v8.16b, v9.16b, v10.16b
  bcax v4.16b, v8.16b, v9.16b, v10.16b
  bcax v5.16b, v8.16b, v9.16b, v10.16b
  bcax v6.16b, v8.16b, v9.16b, v10.16b
  bcax v7.16b, v8.16b, v9.16b, v10.16b
  movi v8.16b, 9
  movi v9.16b, 10
  movi v10.16b, 11

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

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

retire (01)cycle (02)031e1f3a3f51inst 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)9fa8accdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
80204200601501350040258010010080000100800005006400001200192003820038997339996801002008000020024000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
8020420038150120040258010010080000100800005006400001200192003820038997339996801002008000020024088220038200381180201100991001008000010000051101161120035800001002003920039200392003920039
802042003815000040258010010080000100800005006400001200192003820038997339996801002008000020024000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
802042003815000040258010010080000100800005006400001200192003820038997339996801002008000020024000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
802042003815000040258010010080000100800005006400001200192003820038997339996801002008000020024000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
802042003815000040258010010080000100800005006400001200192003820038997339996801002008000020024000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
8020420038150000617258010010080000100800005006400001200192003820038998239996801002008000020024000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
802042003815000040258010010080000100800005006400001200192003820038997339996801002008000020024000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
802042003815000040258010010080000100800005006400001200192003820038997339996801002008000020024000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
8020420038150210040258010010080000100801125006400001200192003820038997339996801002008000020024000020088200381180201100991001008000010000051101161120035800001002003920039200392003920039

1000 unrolls and 10 iterations

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

retire (01)cycle (02)030b1e1f3f51inst 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)9fa8acbranch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420039150000392580010108000010800005064000020019200382003899963100188001020800002024000020038200381180021109101080000100005020151616142003580000102003920039200392003920039
80024200381500150392580010108000010800005064000020019200382003899963100188001020800002024000020038200381180021109101080000100005020111610152003580000102003920039200392003920039
800242003815003902162580010108000010800005064000020019200382003899963100188001020800002024000020038200381180021109101080000100005020151614162003580000102003920039200392003920039
8002420038150000392580010108000010800005064000020019200382003899963100188001020800002024000020038200381180021109101080000100005020161615112003580000102003920039200392003920039
8002420038150000392580010108000010800005064000020019200382003899963100188001020800002024000020038200381180021109101080000100005020161611162003580000102003920039200392003920039
800242003815000039258001010800001080000506400002001920038200389996310018800102080000202400002003820038118002110910108000010000502015161392003580000102003920039200392003920039
8002420038150000392580010108000010800005064000020019200382008899963100188001020800002024000020038200381180021109101080000100005020151615142003580000102003920039200392003920039
80024200381500270392580010108000010800005064000020019200382003899963100188001020800002024000020038200381180021109101080000100005020161615112003580000102003920039200392003920039
80024200381500420392580010108000010800005064000020019200382003899963100188001020800002024000020038200381180021109101080000100005020151612162003580000102003920039200392003920039
8002420038150000392580010108000010800005064000020019200382003899963100188001020800002024000020038200381180021109101080000100005020151613122003580000102003920039200392003920039