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, 4S)

Test 1: uops

Code:

  saddl2 v0.4s, v0.8h, v1.8h
  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
100420371506116872510001000100026468002018203720371572318951000100020002037203711100110000073216221787100020382038203820382038
100420371506116872510001000100026468002018203720371572318951000100020002037203711100110001073216221787100020382038203820382038
100420371508216872510001000100026468002018203720371572318951000100020002037203711100110000073216221787100020382038203820382038
100420371506116872510001000100026468002018203720371572318951000100020002037203711100110000073216221787100020382038203820382038
100420371506116872510001000100026468002018203720371572318951000100020002037203711100110000073216221787100020382038203820382038
1004203715126116872510001000100026468002018203720371572318951000100020002037203711100110000673216221787100020382038203820382038
100420371506116872510001000100026468002018203720371572318951000100020002037203711100110000073216221787100020382038203820382038
100420371506116872510001000100026468002018203720371572318951000100020002037203711100110000073216221787100020382038203820382038
100420371506116872510001000100026468002018203720371572318951000100020002037203711100110000073216221787100020382038203820382038
100420371506116872510001000100026468002018203720371572318951000100020002037203711100110000073216221787100020382038203820382038

Test 2: Latency 1->2

Code:

  saddl2 v0.4s, v0.8h, v1.8h
  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)030b1e3f4e51inst 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)9fa1acc5cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102042003715013557196872510100105100001051015250028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001002216007101161119825100001002003820038200382003820038
10204200371500061196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000007101161119791100001002003820038200382003820038
102042003715001261196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000007101161119791100001002003820038200382003820038
10204200371500061196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000007101161119791100001002003820038200382003820038
102042003715002161196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000007101161119791100001002003820038200382003820038
102042003715001561196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000007101161119791100001002003820038200382003820038
10204200371500061196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000007101161119791100001002003820038200382003820038
102042003715000441196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000007101241119791100001002003820038200382003820038
10204200371500061196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000007101161119791100001002003820038200382003820038
10204200371500061196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000007101161119791100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)03191e3f4e51inst 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)9fa8cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002420037150006119687251001010100001010000502847680200182003720037184443187671001020100002020000200372003711100211091010100001006403163319785010000102003820038200382003820038
10024200371500017019687251001010100001010000502847680200182003720037184443187671001020100002020000200372003711100211091010100001006403163319785010000102003820038200382003820038
1002420037150006119687251001010100001010000502847680200182003720037184443187671001020100002020000200372003711100211091010100001006403163319785010000102003820038200382003820038
10024200371500014719687251001010100001010000502847680200182003720037184443187671001020100002020000200372003711100211091010100001006403163319785110000102003820038200382003820038
1002420037150006119687431001011100001010000502847680200182003720037184443187671001020100002020000200372003711100211091010100001006403163319785010000102003820038200382003820038
1002420037150006119687251001010100001010000502847680200182003720037184443187671001020100002020000200372003711100211091010100001006403163319785010000102003820038200382003820038
10024200371500012619687251001010100001010000502847680200182003720037184443187671001020100002020000200372003711100211091010100001006403163319785010000102003820038200382003820038
100242003715000122319687251001010100001010000502847680200182003720037184443187671001020100002020000200372003711100211091010100001006403163319785010000102003820038200382003820038
1002420037150006119687251001010100001010000502847680200182003720037184443187671001020100002020000200372003711100211091010100001006403243319785010000102003820038200382003820038
1002420037150006119687251001010100001010000602847680200182003720037184443187671001020100002020000200372003711100211091010100001006403163319785010000102003820038200382003820038

Test 3: Latency 1->3

Code:

  saddl2 v0.4s, v1.8h, v0.8h
  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)033f4e51inst 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)9faccdcfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102042003715061196872510100100100001001000050028476800200182003720037184223187651010020010000200200002003720037111020110099100100100001000071021622197910100001002003820038200382003820038
1020420037149214196872510100100100001001000050028476800200182003720037184223187451010020010000200200002003720037111020110099100100100001000071021622197910100001002003820038200382003820038
102042003715061196872510100100100001001000050028476800200182003720037184223187451010020010000200200002003720037111020110099100100100001000071021622197910100001002003820038200382003820038
1020420037150170196872510100100100001001000050028476800200182003720037184223187451010020010000200200002003720037111020110099100100100001000071021622197910100001002003820038200382003820038
10204200371501261968725101001001000010010000500284768002001820037200371842231874510100200100002002000020037200371110201100991001001000010000712216221979112100001002003820038200382003820038
102042003715061196872510100100100001001000050028476800200182003720037184223187451010020010000200200002003720037111020110099100100100001000071021622197910100001002003820038200382003820038
102042003715084196872510100100100001001000050028476800200182003720037184223187451010020010000200200002003720037111020110099100100100001000071021622197910100001002003820038200382003820038
102042003715061196872510100100100001001000050028476800200182003720037184223187451010020010000200200002003720037111020110099100100100001000071021622197910100001002003820038200382003820038
102042003715061196872510100100100001001000050028476800200182003720037184223187451010020010000200200002003720037111020110099100100100001000071021622197910100001002003820038200382003820038
102042003715084196872510100100100001001000050028476800200182003720037184223187451010020010000200200002003720037111020110099100100100001000071021622198980100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)0318191e3f4e51inst 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)9fa8acc2cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002420037150000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000640716541978510000102003820038200382003820038
1002420037150000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000640516541978510000102003820038200382003820038
1002420037150000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010004640516451982310000102003820038200382003820038
1002420037150000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000640516551978510000102003820038200382003820038
1002420037150000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000640516541978510000102003820038200382003820038
1002420037150000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000640450451978510000102003820038200382003820038
10024200371500006211967625100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000640516551978510000102003820038200382003820038
1002420084150110611968725100101010000101000050284896312001820037200371844431876710010201000020200002003720037111002110910101000010000640416541978510000102008520038200382003820038
100242003715000156611968725100101010012101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000640516551978510000102003820038200382003820038
1002420037150000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000640516551978510000102003820038200382003820038

Test 4: throughput

Count: 8

Code:

  saddl2 v0.4s, v8.8h, v9.8h
  saddl2 v1.4s, v8.8h, v9.8h
  saddl2 v2.4s, v8.8h, v9.8h
  saddl2 v3.4s, v8.8h, v9.8h
  saddl2 v4.4s, v8.8h, v9.8h
  saddl2 v5.4s, v8.8h, v9.8h
  saddl2 v6.4s, v8.8h, v9.8h
  saddl2 v7.4s, v8.8h, v9.8h
  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)030818191e1f3f51inst 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
8020420060150000004025801001008000010080000500640000020019200382003899733999680100200800002001600002003820038118020110099100100800001000000000511021611200350800001002003920039200392003920039
802042003815000000386625801001008000010080000500640000020019200382003899733999680100200800002001600002003820038118020110099100100800001000000000511011612200350800001002003920039200392003920039
8020420038150000004025801251258000012580000500640000020019200382003899733999580100200800002001600002003820038118020110099100100800001000000030511011611200350800001002003920039200392003920039
8020420038150000004025801001008000010080000500640000020019200382003899733999680100200800002001600002003820038118020110099100100800001000000000511011611200350800001002003920039200392003920039
8020420038150000004025801001008000010080000500640000020019200382003899733999580125200800002001600002003820038118020110099100100800001000000000511011621200350800001002003920039200392003920039
8020420038151000003925801001008000010080000500640000020019200382003899733999580100200800002001600002003820038118020110099100100800001000000000511011611200350800001002003920039200392003920039
80204200381500000022925801001008000010080000500640000020019200382003899733999680100200800002001600002003820038118020110099100100800001000000000511011611200350800001002003920039200392003920039
8020420038150000004025801001008000010080000590640000120019200382003899733999580100200800002001600002008520038118020110099100100800001000000000511011711200350800001002003920039200392003920039
8020420038150000004025801001008000010080000626640000020019200382003899733999580100200800002001600002003820038118020110099100100800001000000000511011611200350800001002003920039200392003920039
8020420038149000004025801001008000010080000500640000020019200382003899733999680100200800002001600002003820038118020110099100100800001000000100511011611200350800001002003920039200392003920039

1000 unrolls and 10 iterations

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

retire (01)cycle (02)03080b18191e1f3f51inst 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)5f60696d6edispatch 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
80024200391500000003925800101080000108000050640000002001920038200389996310018800102080000201600002003820038118002110910108000010000000050202216171720035080000102003920039200392003920039
80024200381500000003925800101080000108000050640000002001920038200389996310018800102080000201600002003820038118002110910108000010000000050201516181420035080000102003920039200392003920039
80024200381500000003925800101080000108000050640000002001920038200389996310018800102080000201600002003820038118002110910108000010000000050201516171520035080000102003920039200392003920039
80024200381500000003948800101080000108000050640000002001920038200389996310018800102080000201600002003820038118002110910108000010000000050201816181820035080000102003920039200392003920039
80024200381500000003925800101080000108000050640000002001920038200389996310018800102080000201600002003820038118002110910108000010000000050201816151820035080000102003920039200392003920039
80024200381500000003925800101080000108000050640000002001920038200389996310018800102080000201600002003820038118002110910108000010000000050201716171720035080000102003920039200392003920039
80024200381500000003925800101080000108000050640000002001920038200389996310018800102080000201600002003820038118002110910108000010000000050201816171720035080000102003920039200392003920039
80024200381500000003925800101080000108000050640000002001920038200389996310018800102080000201600002003820038118002110910108000010000000050201916181520035080000102003920039200392003920039
80024200381500000003925800101080000108000050640000002001920038200389996310018800102080000201600002003820038118002110910108000010000000050201816151820035080000102003920039200392003920039
80024200381500000003925800101080000108000050640000002001920038200389996310018800102080000201600002003820038118002110910108000010000000050201816151820035080000102003920039200392003920039