Apple M1 Microarchitecture Research by Dougall Johnson

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

ZIP2 (vector, 8H)

Test 1: uops

Code:

  zip2 v0.8h, 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)0318191e1f3a3f4e51inst 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)a0a1a7a8a9acc2cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
1004203716000006116872510001000100026468002018203720371572318951000100020002037203711100110000001503073216221787100020382038203820382038
100420371500000611687251000100010002646800201820372037157231895100010002000203720371110011000000303073216221787100020382038203820382038
1004203716000001561687251000100010002646800201820372037157231895100010002000203720371110011000000100073216221787100020382038203820382038
1004203715000006116872510001000100026468002018203720371572318951000100020002037203711100110000002000073216221787100020382038203820382038
100420371600000611687251000100010002646800201820372037157231895100010002000203720371110011000000003073216221787100020382038203820382038
100420371600000611687251000100010002646800201820372037157231895100010002000203720371110011000000000073216221787100020382038203820382038
100420371600000611687251000100010002646800201820372037157231895100010002000203720371110011000000000073216221787100020382038203820382038
100420371600000611687251000100010002646800201820372037157231895100010002000203720371110011000000000073216221787100020382038203820382038
100420371500000611687251000100010002646800201820372037157231895100010002000203720371110011000000000073216221787100020382038203820382038
1004203715000880611687251000100010002646800201820372037157231895100010002000203720371110011000000000073216221787100020382038203820382038

Test 2: Latency 1->2

Code:

  zip2 v0.8h, 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)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)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
1020420037150006119687251010010010000100100005002847680020018200372003718422318745101002001000020020000200372003711102011009910010010000100500007102161119791100001002003820038202312003820038
1020420037150006119687251010010010000100100005002847680020018200372003718422318763101002001000020020000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
10204200371500015619687251010010010000100100005002847680020018200372003718422318745101002001017220020000200372003711102011009910010010000100100007101161119791100001002003820038200382003820038
1020420037150006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
1020420037150006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
1020420037150006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
1020420037150106119687251010010010000100100005002847680020018200372003718422318745101002001000020020000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
10204200371500024719687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100100007101161119791100001002003820038200382003820038
1020420037150006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
1020420037150006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)03080918191e1f3a3f4e51inst 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)9fa1a6a8a9accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024200371500000000611968725100101010000101000050284768012001820086200371844431876710010201000020200002003720037111002110910101000010000006405164319785010000102003820038200382003820038
10024200371500000000611968725100101010000101000050284768002001820037200371844431876710010201000020200002003720037111002110910101000010000006404164319785010000102003820038200382003820038
10024200371500000000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000006403163419785010000102003820038200382003820038
100242003715000000011031968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000006404164319785010000102003820038200382003820038
100242003715000000006119687251001010100001010000502847680020018200372003718444318767100102010000202000020037200371110021109101010000100000756404163419785010000102003820038200382003820038
10024200371500000000611968725100101010036141000050284768002001820037200371844431876710010201000020200002003720037111002110910101000010000006423163419785010000102003820038200382003820038
10024200371500000000611968725100101010000101000050284768002001820037200371844431876710010201000020200002003720037111002110910101000010000006404164319785010000102003820038200382003820038
10024200371500000000611968725100101010000101000050284768002001820037200371844431876710010201000020200002003720037111002110910101000010000006404164419785010000102003820038200382003820038
10024200371500000000611968725100101010000101000050284768002001820037200371844431876710010201000020200002003720037111002110910101000010000006404164419785010000102003820038200382003820038
100242003715000000001561968725100101010000101000050284768002001820037200371844431876710010201000020200002003720037111002110910101000010000036403164420037010000102003820038200382003820038

Test 3: Latency 1->3

Code:

  zip2 v0.8h, 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)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)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? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020420037155003201968725101001001000010010000500284768002001820037200371842231874510100200100002002000020037200371110201100991001001000010003071011611197910100001002003820038200382003820038
1020420037156001671968725101001001000010010000500284768002001820037200371842231874510100200100002002000020037200371110201100991001001000010000071011611197910100001002003820038200382003820038
1020420037156001241968725101001001000010010000500284768002001820037200371842231874510100200100002002000020037200371110201100991001001000010010071011611197910100001002003820038200382003820038
102042003715500611968725101001001000010010000500284768012001820037200371842231874510100200100002002000020037200371110201100991001001000010000071011611197910100001002003820038200382003820038
102042003715600841968725101001001000010010000500284768012001820037200371842231874510100200100002002000020037200371110201100991001001000010000071011611197910100001002003820038200382003820038
1020420037155004791968725101001001000010010000500284768012001820037200371842231874510100200100002002000020037200371110201100991001001000010000071011611197910100001002003820038200382003820038
102042003715510611968725101001001000010010000500284768012001820037200371842231874510100200100002002000020037200371110201100991001001000010000071011611197910100001002003820038200382003820038
10204200371550010019687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000710116111979126100001002003820038200382003820038
1020420037155001691968725101001001000010010000500284768012001820037200371842231874510100200100002002000020037200371110201100991001001000010000071011611197910100001002003820038200382003820038
1020420037156091561968725101001001000010010000500284768002001820037200371842231874510100200100002002000020037200371110201100991001001000010000071011611197910100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)03040818191e3f4e51inst 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)9fa1a6a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024200371551000076919687251001010100001010000502847680020018200372003718444318767100102010000202000020037200371110021109101010000100000006403162219785010000102003820038200382003820038
10024200371550000082019687251001010100001010000502847680020018200372003718444318767100102010000202000020037200371110021109101010000100000046402162219785010000102003820038200382003820038
10024200371550000132115919687251001010100001010000502847680020018200372003718444318767100102010000202000020037200371110021109101010000100000006402162219785010000102003820038200382003820038
1002420037155000006119687251001010100001010000502847680020018200372003718444318767100102010000202000020037200371110021109101010000100000006402162219785010000102003820038200382003820038
100242003715500000106819687251001010100001010000502847680020018200372003718444318767100102010000202000020037200371110021109101010000100000006402162219785010000102003820038200382003820038
1002420037155000006119687251001010100001010000502847680120018200372003718444318767100102010000202000020037200371110021109101010000100000006402162219800010000102003820038200382003820038
10024200371550000068119687251001010100001010000502847680020018200372003718444318767100102010000202000020037200371110021109101010000100000006402162219785010000102003820038200382003820038
10024200371550000095419687251001010100001010000502847680020018200372003718444318767100102010000202000020037200371110021109101010000100000006402162219785010000102003820038200382003820038
10024200371550000049519687251001010100001010000502847680020018200372003718444318767100102010000202000020037200371110021109101010000100000006402162219785010000102003820038200382003820038
10024200371560000014519687251001010100001010000502847680020018200372003718444318767100102010000202000020037200371110021109101010000100000006402162219785010000102003820038200382003820038

Test 4: throughput

Count: 8

Code:

  zip2 v0.8h, v8.8h, v9.8h
  zip2 v1.8h, v8.8h, v9.8h
  zip2 v2.8h, v8.8h, v9.8h
  zip2 v3.8h, v8.8h, v9.8h
  zip2 v4.8h, v8.8h, v9.8h
  zip2 v5.8h, v8.8h, v9.8h
  zip2 v6.8h, v8.8h, v9.8h
  zip2 v7.8h, 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)0308090b18191e1f3f51inst 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)9fa0a1a6a7a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
80204200381560000000822580100100800001008000050064000012001902003820038997339996801002008000020016000020038200381180201100991001008000010000000000511021611200350800001002003920039200392003920039
80204200381550000000402580100100800001008000050064000002001902003820038997339996801002008000020016000020038200381180201100991001008000010000000000511011611200350800001002003920039200392003920039
80204200381550000000402580100100800001008000050064000002001902003820038997339996801002008000020016000020038200381180201100991001008000010000000000511011611200350800001002003920039200392003920039
80204200381550000000402580100100800001008000050064000002001902003820038997339996801002008000020016000020038200381180201100991001008000010000000000511011611200350800001002003920039200392003920039
80204200381550000030402580100100800001008000050064000002001902003820038997339996801002008000020016000020038200381180201100991001008000010000000000516211611200350800001002003920039200392003920039
8020420038155000000132402580100100800001008000050064000002001902003820038997339996801002008000020016000020038200381180201100991001008000010000000000511011611200350800001002003920039200392003920039
80204200381550000000402580100100800001008000050064000002001902003820038997339996801002008000020016000020038200381180201100991001008000010000000030511011611200350800001002003920039200392003920039
802042003815500000007052580100100800001008000050064000002001902003820038997339996801002008000020016000020038200381180201100991001008000010000000000511011611200350800001002003920039200392003920039
80204200381550000000402580100100800001008000050064000002001902003820038997339996801002008000020016000020038200381180201100991001008000010000000000511011611200350800001002003920039200392003920039
80204200381550000000402580100100800001008000050064000002001902003820038997339996801002008000020016000020038200381180201100991001008000010000000000514511611200350800001002003920039200392003920039

1000 unrolls and 10 iterations

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

retire (01)cycle (02)03040708090a0b18191e1f3a3f51inst 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)9fa0a1a6a8a9acc2cfitlb miss (d4)d5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
800242004815600000000003104425800101080000108000050640000020019200382003899963100188001020800002016000020038200381180021109101080000100000030502001716171020035080000102003920039200392003920039
8002420038155000000000023925800101080000108000050640000020019200382003899963100188001020800002016000020038200381180021109101080000100000060502001616161720035080000102003920039200392003920039
80024200381550000000000014625800101080000108000050640000020019200382003899963100188001020800002016000020038200381180021109101080000100000000502001616161320035080000102003920039200392003920039
8002420038156000000000006225800101080000108000050640000020019200382003899963100188001020800002016000020038200381180021109101080000100000000502001616161720035080000102003920039200392003920039
80024200381550000000000016525800101080000108000050640000020019200382003899963100188001020800002016000020038200381180021109101080000100000000502001416161220035080000102003920039200392003920039
80024200381550000000000025425800101080000108000050640000020019200382003899963100188001020800002016000020038200381180021109101080000100000000502001516171320035080000102003920039200392003920039
80024200381550000100000012325800101080000108000050640000020019200382003899963100188001020800002016000020038200381180021109101080000100000000502001616141620035080000102003920039200392003920039
8002420038156000000000003925800101080000108000050640000020019200382003899963100188001020800002016000020038200381180021109101080000100000000502001416161420035080000102003920039200392003920039
8002420038155000000000003925800101080000108000050640000020019200382003899963100188001020800002016000020038200381180021109101080000100000000502111616161620035080000102003920039200392003920039
80024200381551000000000021025800101080000108000050640000020019200382003899963100188001020800002016000020038200381180021109101080000100000000502111416171420035080000102003920039200392003920039