Apple M1 Microarchitecture Research by Dougall Johnson

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

TRN2 (vector, 4H)

Test 1: uops

Code:

  trn2 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
100420371506116872510001000100026468012018203720371572318951000100020002037203711100110000073216111787100020382038203820382038
100420371506116872510001000100026468012018203720371572318951000100020002037203711100110000073116111787100020382038203820382038
100420371506116872510001000100026468012018203720371572318951000100020002037203711100110000373116111787100020382038203820382038
100420371596116872510001000100026468012018203720371572318951000100020002037203711100110000073116111787100020382038203820382038
100420371506116872510001000100026468012018203720371572318951000100020002037203711100110000373116111787100020382038203820382038
100420371506116872510001000100026468012018203720371572318951000100020002037203711100110000373116111787100020382038203820382038
100420371506116872510001000100026468012018203720371572318951000100020002037203711100110000073116111787100020382038203820382038
100420371696116872510001000100026468012018203720371572318951000100020002037203711100110000073116111787100020382038203820382038
100420371506116872510001000100026468012018203720371572318951000100020002037203711100110000073116111787100020382038203820382038
100420371606116872510001000100026468012018203720371572318951000100020002037203711100110000073116111787100020382038203820382038

Test 2: Latency 1->2

Code:

  trn2 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): 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)9fa1a8a9accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204203701521076936528038031960815610229107100841411091269128565811202342037120361184423518854111982201083822021998203262037271102011009910010010000100211212073844216312000732100001002032420323203272032420324
102042032215201666635280124196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000230671011611197910100001002003820038200382003820038
102042003715000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000071011611197910100001002003820038200382003820038
102042003715000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000071011611197910100001002003820038200382003820038
102042003715000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000071011611197910100001002003820038200382003820038
102042003715000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000071011611197910100001002003820038200382003820038
102042003715000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000071011611197910100001002003820038200382003820038
102042003715000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000371011611197910100001002003820038200382003820038
102042003715000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000071011611197910100001002003820038200382003820038
102042003715000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000071011611197910100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)0308090b18191e1f243f4e51inst 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
100242003715000000000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000000006402162219785010000102003820038200382003820038
1002420037150000000006119687251001010100001010000502847680120018200372003718444318767100102010000202000020037200371110021109101010000101000001206402162219785010000102003820038200382003820038
100242017815001000000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000000006402162219785010000102003820038200382003820038
100242003715000000000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000010306401162219785010000102003820038200382003820038
100242003715000000000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000000006402162219785010000102003820038200382003820038
10024200371500000000061196872510010101000010100005028476801200182003720037184443187671001020100002020000200372003711100211091010100001000000015006402162219785010000102003820038200382003820038
1002420037150000000006119687251001010100001010000502847680120018200372003718444318767100102010000202000020037200371110021109101010000100000208406402162219785010000102003820038200382003820038
100242003715000000000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000000006402162219785010000102003820038200382003820038
1002420037150000000006119687251001010100001010000502847680120018200372003718444318767100102010000202000020037200371110021109101010000100000500306402162219785010000102003820038200382003820038
100242003715000000000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010000010306402162219785010000102003820038200382003820038

Test 3: Latency 1->3

Code:

  trn2 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): 2.0037

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)9fa0a1a6a7a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020420037150000000611968725101001001000010010000500284768012001820037200371842231874510100200100002002000020037200371110201100991001001000010000003500071011611197910100001002003820038200382003820038
1020420037150000000611968725101001001000010010000500284768012001820037200371842231874510100200100002002000020037200371110201100991001001000010000003003071011611197910100001002003820038200382003820038
1020420037150000000611968725101001001000010010000500284768012001820037200371842231874510100200100002002000020037200371110201100991001001000010000003303071011611197910100001002003820038200382003820038
102042003715000000061196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000000103963482712512200050100001002003820038200382003820038
102042003715000000061196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000000100071011611197910100001002003820038200382003820038
102042003715000000061196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000000000071011611197910100001002003820038200382003820038
102042003715000000084196872510100100100001001000050028476801200182003720037184223187451010020010000200200002003720037111020110099100100100001000000000071011611197910100001002003820038200382003820038
1020420037150000030611968725101001001000010010000500284768012001820037200371842231874510100200100002002000020037200371110201100991001001000010000002703071011611197910100001002003820084200382003820038
1020420037149000000611968725101001001000010010000500284768012001820037200371842231874510100200100002002000020037200371110201100991001001000010000003303071011611197910100001002003820038200382003820038
1020420037150000000611968725101001001000010010000500284768012001820037200371842231874510100200100002002000020037200371110201100991001001000010000002503071011611197910100001002003820038200382003820086

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)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)9fa8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100242003715001031968725100101010000101000050284768002001802003720037184443187671061920100002020000200372003711100211091010100001000640316331978510000102003820038200382003820038
10024200371500611968725100101010000101000050284768002001802003720037184443187671001020100002020000200372003711100211091010100001000640316341978510000102003820038200382003820038
10024200371500611968725100101010000101000050284768002001802003720037184443187671001020100002020000200372003711100211091010100001010640416431978510000102003820038200382003820038
10024200371500611968725100101010000101000050284768002001832003720037184443187671001020100002020000200372003711100211091010100001000640316441978510000102003820038200382003820038
10024200371500611967644100101010000101000050284768002001802003720037184443187671001020100002020000200372003711100211091010100001000640316431978510000102003820038200382003820038
10024200371500611968725100101010000101000050284768002001802003720037184443187671001020100002020000200372003711100211091010100001000640416341978510000102003820038200382003820038
100242003715006119687251001010100001010000502847680020018020037200371844431876710010201000020200002003720037111002110910101000010310640416441978510000102003820038200382003820038
10024200371500611968725100101010000101000050284768002001802003720037184443187671001020100002020000200372003711100211091010100001000640316431978510000102003820038200382003820038
10024200371500611968725100101010000101000050284768002001802003720037184443187671001020100002020000200372003711100211091010100001000640416441978510000102003820038200382003820038
10024200371500611968725100101010000101000050284768002001802003720037184443187671001020100002020000200372003711100211091010100001010640316341978510000102003820038200382003820038

Test 4: throughput

Count: 8

Code:

  trn2 v0.4h, v8.4h, v9.4h
  trn2 v1.4h, v8.4h, v9.4h
  trn2 v2.4h, v8.4h, v9.4h
  trn2 v3.4h, v8.4h, v9.4h
  trn2 v4.4h, v8.4h, v9.4h
  trn2 v5.4h, v8.4h, v9.4h
  trn2 v6.4h, v8.4h, v9.4h
  trn2 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)031e1f3f4f51inst 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)9fa6a8acc5branch mispredict (cb)cdcficache miss (d3)d5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
80204200691500071025801081008000810080020500640132020019200382003899776998980120200800322001600642003820038118020110099100100800001000001115118011600200350800001002003920039200392003920039
80204200381500029025801081008000810080020500640132020019200382003899776998980120200800322001600642003820038118020110099100100800001000001115118001600200350800001002003920039200392003920039
80204200381500040025801001008000010080000500640000020019200382003899733999680100200800002001600002003820038118020110099100100800001000030005110011611200350800001002003920039200392003920039
8020420038150010863044801001008000010080000615640000020019200382003899813999680100200800002001600002003820038118020110099100100800001000000005110011611200350800001002003920039200392003920039
802042003815090437025801951108009310080000500640000020019200382003899733999680100200800002001600002003820038118020110099100100800001000100005110011611200350800001002003920039200392003920039
802042003815000888025801001008000010080000500640000020019200382003899733999680100200800002001600002003820038118020110099100100800001000000005110011611200350800001002003920039200392003920039
8020420038150120736025801001008000010080000500640000020019200382003899733999680100200800002001600002003820038118020110099100100800001000100005110011611200350800001002003920039200392003920039
80204200381600040025801001008000010080000500640000020019200382003899733999680100200800002001600002003820038118020110099100100800001000030015110011611200350800001002003920039200392003920039
802042003815000145025801001008000010080000500640000020019200382003899733999680100200800002001600002003820038118020110099100100800001002000005110011611200750800001002003920039200872003920039
802042003815000610258010010080000100800005456400000200192009520038997371004980100200800002001600002003820038118020110099100100800001000030005110011611200350800001002003920039200392003920039

1000 unrolls and 10 iterations

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

retire (01)cycle (02)0309191e3f51inst 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? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420048150000812580010108000010800005064000002001920038200389996310018800102080000201600002003820038118002110910108000010000502021168820035080000102003920039200392003920039
8002420038150000392580010108000010800005064000002001920038200389996310046800102080000201600002003820038118002110910108000010000502061661220035080000102003920039200392003920039
80025200381500001892580010108000010800005064000002001920038200389996310018800102080000201600002003820038118002110910108000010000502061661220035080000102003920039200392003920039
80024200381500001042580010108000010800005064000002001920038200389996310018800102080000201600002003820038118002110910108000010001502081612820035080000102003920039200392003920039
8002420038150000392580010108000010800005064000002001920038200389996310018800102080000201600002003820038118002110910108000010000502061661220035080000102003920039200392003920039
800242003815000039258001010800001080000506400000200192003820038999631001880010208000020160000200382003811800211091010800001000050208166820035080000102003920039200392003920039
80024200381500001022580010108000010800005064077202001920038200389996310018800102080000201600002003820038118002110910108000010000502012168620035080000102003920039200392003920039
8002420038149000148258001010800001080000506400000200192003820038999631001880010208000020160000200382003811800211091010800001000050206168620035080000102003920039200392003920039
800242003815000039258001010800001080000506400000200192003820038999631001880010208000020160000200382003811800211091010800001000050208168620035080000102003920039200392003920039
8002420038150000148258001010800001080000506400000200192003820038999631001880010208000020160000200382003811800211091010800001010050208168620035080000102003920039200392003920039