Apple M1 Microarchitecture Research by Dougall Johnson

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

SADDL2 (vector, 8H)

Test 1: uops

Code:

  saddl2 v0.8h, v0.16b, v1.16b
  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)696d6edispatch 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)a0a6a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
1004203715061168725100010001000264680201820372037157231895100010002000203720371110011000000073116111787100020382038203820382038
1004203715061168725100010001000264680201820372037157231895100010002000203720371110011000000073116111787100020382038203820382038
1004203715061168725100010001000264680201820372037157231895100010002000203720371110011000000073116111787100020382038203820382038
1004203715061168725100010001000264680201820372037157231895100010002000203720371110011000000073116111787100020382038203820382038
1004203715061168725100010001000264680201820372037157231895100010002000203720371110011000000073116111787100020382038203820382038
1004203715061168725100010001000264680201820372037157231895100010002000203720371110011000001373116111787100020382038203820382038
1004203715061168725100010001000264680201820372037157231895100010002000203720371110011000000073116111787100020382038203820382038
1004203715061168725100010001000264680201820372037157231895100010002000203720371110011000000073116111787100020382038203820382038
1004203715061168725100010001000264680201820372037157231895100010002000203720371110011000000073116111787100020382038203820382038
10042037150821687251000100010002646802018203720371572318951000100020002037203711100110000003673116111787100020382038203820382038

Test 2: Latency 1->2

Code:

  saddl2 v0.8h, v0.16b, v1.16b
  movi v0.16b, 1
  movi v1.16b, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 2.0037

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)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)9fa8c5cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102042008415106119687251010010010000100100005002847680020018200372003718422318745101002001000020020000200372003711102011009910010010000100007101161119791100001002003820038200382003820038
102042003715006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100007101161219791100001002003820038200382003820038
1020420037150017419687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100007101161119791100001002003820038200382003820038
102042003715006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100007101161119791100001002003820038200382003820038
1020420037150061196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001004707101161119791100001002003820038200382003820038
1020420037150015619687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100007101161119791100001002003820038200382003820038
1020420037150037519687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100007101161219791100001002003820038200382003820038
102042003714906119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100007101161119791100001002003820038200382003820038
102042003714906119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100007101161119791100001002003820038200382003820038
102042003715006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100007101161119791100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)0308090b18191e1f3a3f4e51inst 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
10024200371500000000061196872510010101000010100005028476800200182003720037184443187671001020100002020000200372003711100211091010100001000000006404164419785010000102003820038200382003820038
10024200371500000000061196872510010101000010100005028476800200182003720037184443187671001020100002020000200372003711100211091010100001000000006404164419785010000102003820038200382003820038
10024200371500000000061196872510010101000010100005028476800200182003720037184443187671001020100002020000200372003711100211091010100001000000006403163419949010000102003820038200382003820038
10024200371500000000061196872510010101000010100005028476800200182003720037184443187671001020100002020000200372003711100211091010100001000000006403162319785010000102003820038200382003820038
10024200371500000000061196872510010101000010100005028476800200182003720037184443187671001020100002020000200372003711100211091010100001000000006404162319785010000102003820038200382003820038
10024200371500000000061196872510010101004810100005028502460200182003720037184443187671001020100002020000200372003711100211091010100001000000006404164319785010000102003820038200382003820038
10024200371500000000082196872510010101000010100005028476800200182003720037184443187671001020100002020000200372003711100211091010100001044000006403163219785010000102003820038200382003820038
100242003715000000120061196872510010101000010100005028476800200182003720037184443187671001020100002020000200372003711100211091010100001000000006404163419785010000102003820038200382003820038
10024200371500000000061196872510010101000010100005028476800200182003720037184443187671001020100002020000200372003711100211091010100001000000006404162319785010000102003820038200382003820038
10024200371500000000061196872510010101000010101525028476801200182003720037184443187671001020100002020000200372003711100211091010100001000000306404164419785010000102003820038200382003820038

Test 3: Latency 1->3

Code:

  saddl2 v0.8h, v1.16b, v0.16b
  movi v0.16b, 1
  movi v1.16b, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)0308090b1e1f3f4e51inst 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)9fa0a1a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204200371500000082196872510100100100001001000050028476802001820037200371842231874510100200100002002000020037200371110201100991001001000010000007101161119791100001002003820038200382003820038
102042003715000000212196874310139107100121041015250028489632001820037200371842231874510100204101642042000020037200371110201100991001001000010000007321161119791100001002003820038200382003820038
10204200371500110082196872510100100100001001015256428476802001820037200371842231874510100200100002002000020037200371110201100991001001000010000007101161119791100001002003820038200382003820038
10204200371500009061196872510100100100001001000050028476802001820037200371842231874510100200100002002000020037200371110201100991001001000010000007101161119791100001002003820038200382003820038
10204200371500003061196872510100100100001001000050028476802001820037200371842271874510100200100002002000020037200371110201100991001001000010000007101161119791100001002003820038200382003820038
102042003715000000103196872510100100100001001000050028476802001820037200371842231874510100200100002002000020037200371110201100991001001000010000007101161119791100001002003820038200382003820038
10204200371500000061196872510100100100001001000050028476802001820037200371842231874510100200100002002000020037200371110201100991001001000010000007101161119791100001002003820038200382003820038
1020420037150000210156196872510100100100001001000050028476802001820037200371842231874510100200100002002000020037200371110201100991001001000010000007101161119791100001002003820038200382003820038
10204200371501000061196872510100100100001001000050028476802001820037200371842231874510100200100002002000020037200371110201100991001001000010000007101161119931100001002003820038200382003820038
10204200371500000061196872510100100100001001000050028476802001820037200371842231874510100200100002002000020037200371110201100991001001000010000007321161119827100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

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)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)9fa1a9accficache miss (d3)d5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100242003715006119687251001010100001010000502847680120018200372003718444318767100102010000202000020037200371110021109101010000100006400216221978510000102003820038200382003820038
1002420037150063119687251001010100001010150502847680120018200372003718444318767100102010000202000020037200371110021109101010000100006400216221978510000102003820038200382003820038
100242003715506119687251001010100001010000502847680120018200372003718444318767100102010000202000020132200841110021109101010000100006400216221978510000102003820038200382003820038
1002420037150029819687251001010100001010000502847680120018200372003718444318767100102010000202000020037200371110021109101010000100006400216221978510000102003820038200382003820038
100242003715006119687251001010100001010000502847680120018200372003718444318767100102010000202000020037200371110021109101010000100006400216221978510000102003820038200382003820038
100242003715006119687251001010100001010000502847680020018200372008418444318767100102010000202000020037200371110021109101010000100006400216221978510000102003820038200382003820038
1002420037150047519687251001010100001010000502847680020018200372003718444318767100102010000202000020037200371110021109101010000100006400216221978510000102003820038200382003820038
1002420037150048219687251001010100001010000502847680120018200372003718444318767100102010000202000020037200371110021109101010000100006400216221978510000102003820038200382003820038
100242003715006119687251001010100001010000502847680120018200372003718444318767100102010000202000020037200371110021109101010000100006400116221978510000102003820038200382003820038
100242003715006119687251001010100001010000502847680120018200372003718444318767100102010000202000020037200371110021109101010000100006400216221978510000102003820038200382003820038

Test 4: throughput

Count: 8

Code:

  saddl2 v0.8h, v8.16b, v9.16b
  saddl2 v1.8h, v8.16b, v9.16b
  saddl2 v2.8h, v8.16b, v9.16b
  saddl2 v3.8h, v8.16b, v9.16b
  saddl2 v4.8h, v8.16b, v9.16b
  saddl2 v5.8h, v8.16b, v9.16b
  saddl2 v6.8h, v8.16b, v9.16b
  saddl2 v7.8h, v8.16b, v9.16b
  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)033f51inst 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)9fa0acbranch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
80204200581504025801001008000010080000500640000200192003820038997339996801002008000020016000020038200381180201100991001008000010000051102161120035800001002003920039200392003920039
80204200381504025801001008000010080000500640000200192003820038997339996801002008000020016000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
80204200381504025801001008000010080000500640000200192003820038997339996801002008000020016000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
802042003815086125801001008000010080000500640000200192003820038997339996801002008000020016000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
80204200381504025801001008000010080000500640000200192003820038997339996801002008000020016000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
802042003815022925801001008000010080000500640000200192003820038997339996801002008000020016000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
80204200381504025801001008000010080000500640000200192003820038997339996801002008000020016000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
80204200381504025801001008000010080000500640000200192003820038997339996801002008000020016000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
80204200381504025801001008000010080000500640000200192003820038997339996801002008000020016000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039
802042003815041025801001008000010080000500640000200192003820038997339996801002008000020016000020038200381180201100991001008000010000051101161120035800001002003920039200392003920039

1000 unrolls and 10 iterations

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

retire (01)cycle (02)031e3a3f51inst 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)9faccficache miss (d3)d5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
800242004815000452580010108000010800005064000002001920038200389996310018800102080000201600002003820038118002110910108000010050207516552003580000102003920039200392003920039
8002420038150007492580010108000010800005064000002001920038200389996310018800102080000201600002003820038118002110910108000010050207716542003580000102003920039200392003920039
8002420038150002712580010108000010800005064000002001920038200389996310018800102080000201600002003820038118002110910108000010050206516552003580000102003920039200392003920039
800242003815000392580010108000010800005064000002001920038200389996310018800102080000201600002003820038118002110910108000010050206616652003580000102003920039200392003920039
800242003815000392580010108000010800005064000012001920038200389996310018800102080000201600002003820038118002110910108000010050206616552003580000102003920039200392003920039
800242003815000392580010108000010800005064000012001920038200389996310018800102080000201600002003820038118002110910108000010050205416452003580000102003920039200392003920039
800242003815000392580010108000010800005064000002001920038200389996310018800102080000201600002003820038118002110910108000010050207616562003580000102003920039200392003920039
800242003815000392580010108000010800005064000002001920038200389996310018800102080000201600002003820038118002110910108000010050206516552003580000102003920039200392003920039
800242003815000392580010108000010800005064000012001920038200389996310018800102080000201600002003820038118002110910108000010050207516662003580000102003920039200392003920039
800242003814900392580010108000010800005064000002001920038200389996310018800102080000201600002003820038118002110910108000010050207516452003580000102003920039200392003920039