Apple M1 Microarchitecture Research by Dougall Johnson

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

UQSHL (vector, 2S)

Test 1: uops

Code:

  uqshl v0.2s, v0.2s, v1.2s
  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)031e1f3f4e51inst 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
1004203715008216872510001000100026468012018203720371572318951000100020002037203711100110000073116111787100020382038203820382038
1004203715006116872510001000100026468012018203720371572318951000100020002037203711100110000073116111787100020382038203820382038
1004203715006116872510001000100026468002018203720371572318951000100020002037203711100110000073116111787100020382038203820382038
1004203715006116872510001000100026468012018203720371572318951000100020002037203711100110000073116111787100020382038203820382038
10042037153006116872510001000100026468012018203720371572318951000100020002037203711100110000073116111787100020382038203820382038
10042037151206116872510001000100026468002018203720371572318951000100020002037203711100110000073116111787100020382038203820382038
1004203715006116872510001000100026468012018203720371572318951000100020002037203711100110000073116111787100020382038203820382038
10042037150015616872510001000100026468012018203720371572318951000100020002037203711100110000373116111787100020382038203820382038
10042037150024216872510001000100026468002018203720371572318951000100020002037203711100110000073116111787100020382038203820382038
1004203715008216872510001000100026468002018203720371572318951000100020002037203711100110000073116111787100020382038203820382038

Test 2: Latency 1->2

Code:

  uqshl v0.2s, v0.2s, v1.2s
  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)030b1e3a3f4e51inst 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)9fa0a8acc2cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020420037186000611968725101001001000010010000500284768002001820037200371842971874010100200100082002001620037200371110201100991001001000010000007101161119791100001002003820038200382003820038
1020420037186000611968725101161001000010010000500284768012001820037200371842231874510100200100002002000020037200371110201100991001001000010000007101161119791100001002003820038200382003820038
10204200371860006221968725101001001000010010000500284768002001820037200371842231874510100200100002002000020037200371110201100991001001000010000007101161119791100001002003820038200382003820038
1020420037173000611968725101001001000010010000500284768002001820037200371842231874510100200100002002000020037200371110201100991001001000010000007101161119791100001002003820038200382003820038
1020420037173000611968725101001001000010010000500284768002001820037200371842231874510100200100002002000020037200371110201100991001001000010001007101161119791100001002003820038200382003820038
10204200371740003461968725101001001000010010000500284768002001820037200371842231874510100200100002002000020037200371110201100991001001000010000027101161119791100001002003820038200382003820038
1020420037173000611968725101001001000010010000500284768012001820037200371842231874510100200100002002000020037200371110201100991001001000010001307101161119791100001002003820038200382003820038
10204200371610120611968725101001001000010010000500284768002001820037200371842231874510100200100002002000020037200371110201100991001001000010001007101161119791100001002003820038200382003820038
1020420037160090611968725101001001000010010000500284768002001820037200371842231874510100200100002002000020037200371110201100991001001000010001307101161119791100001002003820038200382003820038
10204200371610121611968725101001141003611310152500284768012001820037200371842281874510100200100002002000020037200371110201100991001001000010000007101161119791100001002003820038200382003820038

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)9fa0a1a6a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100242003715000002400631196872510010101000010100005028476801200182003720037184443187671001020100002020000200372003711100211091010100001000000006402162219785010000102003820038200382003820038
10024200371500000450061196872510010101000010100005028476801200182003720037184443187671001020100002020000200372003711100211091010100001000000006402162219785010000102003820038200382003820038
1002420037150000000061196872510010101000010100005028476801200182003720037184443187671001020100002020000200372003711100211091010100001000000006402162219785010000102003820038201332008620038
1002420037150000000061196872510010101000010100005028476801200182003720132184443187671001020100002020000200372003711100211091010100001000000006402162219785010000102003820038200382003820038
100242003715000201770061196872510010101000010100005028476801200182003720037184443187671001020100002020000200372003711100211091010100001000000006402163219785010000102003820038200382003820038
1002420037150000000061196872510010101000010100005028476801200182003720037184443187671001020100002020000200372003711100211091010100001000000006402162219785010000102003820038200382003820038
1002420037150000000061196872510010101000010100005028476801200182003720037184443187671001020100002020000200372003711100211091010100001000000306403163219785110000102003820038200852003820038
10024200371500100000103196872510010101000010100006028488871200182008520037184447187671001020100002220000200372003711100211091010100001000000006402162219824010000102003820038200382003820038
1002420037150000000061196872510010101000010100005028476801200182003720037184443187671001020100002020000200372003711100211091010100001000000306842162219785010000102003820038200382003820038
1002420037150003090061196872510010101000010100005028476801200182003720037184443187671001020100002020000200372003711100211091010100001000010006402162219785010000102003820038200382003820038

Test 3: Latency 1->3

Code:

  uqshl v0.2s, v1.2s, v0.2s
  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)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
1020420037150000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038
1020420037150000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038
1020420037150001000006119687251010010010000100100005002847680120018200372003718422718745101002001000020020000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038
10204200371500000000025119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038
1020420037150000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038
1020420037150000000008219687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000100071011611197910100001002003820038200382003820038
1020420037150000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038
102042003715000000000150019654251010010010024100100005332848924120018200372003718422318745101002001000020020000200372003711102011009910010010000100000202010071011611197910100001002003820038200382003820038
1020420037150000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000103071011611197910100001002003820038200382003820038
1020420037150000000006119687251010010010000100100005002847680120018200372003718422318745101002001000020020000200372003711102011009910010010000100000100071011611197910100001002003820038200382003820038

1000 unrolls and 10 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)9fcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100242003715000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010640716331978510000102003820038200382003820038
100242003715000611968725100101010000101000050284768002001820037200371844431876710010201000020200002003720037111002110910101000010640316331978510000102003820038200382003820038
100242003715000611968725100101010000101000050284768002001820037200371844431876710010201000020200002003720037111002110910101000010640316331978510000102003820038200382003820038
10024200371490210611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010640316331978510000102003820038200382003820038
100242003715000611968725100101010000101000050284768002001820037200371844431876710010201000020200002003720037111002110910101000010640316331978510000102003820038200382003820038
100242003715000611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010640316331978510000102003820038200382003820038
100242003715000611968725100101010000101000050284768002001820037200371844431876710010201000020200002003720037111002110910101000010640316331978510000102003820038200382003820038
100242003715500611968725100101010000101000050284768012001820037200371844431876710010201000020200002003720037111002110910101000010640316331978510000102003820038200382003820038
100242003715000611968725100101010000101000050284768002001820037200371844431876710010201000020200002003720037111002110910101000010640316331978510000102003820038200382003820038
100242003715000611968725100101010000101000050284768002001820037200371844431876710010201000020200002003720037111002110910101000010640316331978510000102003820038200382003820038

Test 4: throughput

Count: 8

Code:

  uqshl v0.2s, v8.2s, v9.2s
  uqshl v1.2s, v8.2s, v9.2s
  uqshl v2.2s, v8.2s, v9.2s
  uqshl v3.2s, v8.2s, v9.2s
  uqshl v4.2s, v8.2s, v9.2s
  uqshl v5.2s, v8.2s, v9.2s
  uqshl v6.2s, v8.2s, v9.2s
  uqshl v7.2s, v8.2s, v9.2s
  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)03081e3f51inst 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)9fa8acbranch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
80204200591500040258010010080000100800005006400002001920038200389973039996801002008000020016000020038200381180201100991001008000010000051104162220035800001002003920039200392003920039
80204200381500040258010010080000100800005006400002001920038200389973039996801002008000020016000020038200381180201100991001008000010000051102162220035800001002003920039200392003920039
80204200381500040258010010080000100800005006400002001920038200389973039996801002008000020016000020038200381180201100991001008000010020051102162220035800001002003920039200392003920039
80204200381500040258010010080000100800005006400002001920038200389973039996801002008000020016000020038200381180201100991001008000010000051102162220035800001002023920039200392003920039
80204200381500040258010010080000100800005006400002001920038200389973039996801002008000020016000020038200382180201100991001008000010000051102162220035800001002003920039200392003920039
80204200381500040258010010080000100800005006400002001920038200389973039996801002008000020016000020038200381180201100991001008000010000051102162220035800001002003920039200392003920039
80204200381500040258010010080000100800005006400002001920038200389973039996801002008000020016000020038200381180201100991001008000010000051102162220035800001002003920039200392003920039
80204200381500040258010010080000100800005006400002001920038200389973039996801002008000020016000020038200381180201100991001008000010010051102162220035800001002003920039200392003920039
80204200381500040258010010080000100800005006400002001920038200389973039996801002008000020016000020038200381180201100991001008000010000051102162220035800001002003920039200392003920039
80204200381500040258010010080000100800005006400002001920038200389973079996801002008000020016000020038200381180201100991001008000010050051102162220035800001002003920039200392003920039

1000 unrolls and 10 iterations

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

retire (01)cycle (02)0304080b18191e1f3a3f51inst 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)9fa6accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420039150100000003925800101080000108000050640000020019200382003899963100188001020800002016000020038200381180021109101080000100050202816252520035080000102003920039200392003920039
8002420038151000000003925800101080000128000050640000120019200382003899963100188001020800002016000020038200381180021109101080000100950202716202720035080000102003920039200392003920039
80024200381500000000039258001010800001080000506400000200192003820038999631001880010208000020160000200382003811800211091010800001001850372916142720333080000102003920146200972009620099
8002420038150100455520193710380484158037713803886064232202025020296202901003926101538049420803892016058420188201915180021109101080000102209850932467273520229480000102018920188202402029320141
800242023615101155669352012861028048311802801080000506400000200192003820038999631001880010208000020160000200382003811800211091010800001005150202216272120035080000102003920039200392003920039
8002420038150000000003925800101080000108000050640000020019200382003899963100188001020800002016000020038200381180021109101080000100650202716222620035080000102003920039200392003920039
80024200381500000000039258001010800001080000506400000200192003820038999631001880010208000020160000200382003811800211091010800001001250201416291520035080000102003920039200392003920039
8002420038150000002100392580010108000010800005064000012001920038200389996310018800102080000201600002003820038118002110910108000010010250201616281220035080000102003920039200392003920039
8002420038150010000003925800101080000108000050640000020019200382003899963100188001020800002016000020038200381180021109101080000100950202716152620035080000102003920039200392003920039
80024200381500000000039258001010800001080000506400000200192003820038999631001880010208000020160000200382003811800211091010800001001850201416261920035080000102003920039200392003920039