Apple M1 Microarchitecture Research by Dougall Johnson

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

CMLT (zero, 4H)

Test 1: uops

Code:

  cmlt v0.4h, v0.4h, #0
  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
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371506116862510001000100026452112018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371506116862510001000100026452112018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371506116862510001000100026452112018203720371571318951000100010002037203711100110001073116111786100020382038203820382038
1004203715127016862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371606116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
1004203715186116862510001000100026452112018203720371571318951000100010002037203711100110000373116111786100020382038203820382038

Test 2: Latency 1->2

Code:

  cmlt v0.4h, v0.4h, #0
  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)03191e1f3f4e51inst 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)9fa1a8a9acc2cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102042003715000021219686251010010010000100100005002847521020018200372003718421318745101002001000020010000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
102042003715000061019686251010010010000100100005282847521020018200372003718421318745101002001000020010000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
10204200371500006119686251010010010000100100005002847521120018200372003718421318745101002001000020010000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
10204200371500006119686251010010010000100100005002847521120018200372003718421318745101002001000020010000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
10204200371500006119686251010010010000100100005002847521020018200372003718421318745101002001000020010000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
10204200371500006119686251010010010000100100005002847521120018200372003718421318745101002001000020010000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
102042003714900048819686251010010010000100100005002847521120018200372003718421318745101002001000020010000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
10204200371500006119686251010010010000100100005002847521120018200372003718421318745102782001000020010000200372003711102011009910010010000100010007101161119791100001002003820038200382003820038
10204200371500006119686251010010010000100100005002847521020018200372003718421318745101002001000020010000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038
10204200371500006119686251010010010000100100005002847521120018200372003718421318745101002001000020010000200372003711102011009910010010000100000007101161119791100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)03043f4e51inst 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)9fa8accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024200371500611968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010006404162219786010000102003820038200382003820038
100242003715007961968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010006402162219786010000102003820038200382003820038
100242003715001701968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010006402162219786010000102003820038200382003820038
100242003715001891968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010006402163219786010000102003820038200382003820038
100242003715002211968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010321116402162219786010000102003820038200382003820038
10024200371500611968625100101010000101000050284752112001820037200371844331876710010201000020100002003720037111002110910101000010006402162219786010000102003820038200382003820038
10024200371500611968625100101010000101000050284752112001820037200371844331876710010201000020100002003720037111002110910101000010006402162219786010000102003820038200382003820038
10024200371500611968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010006402162219786010000102003820038200382003820038
100242013315007261968625100101410000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010006402162219786010000102003820038200382003820038
10024200371500821968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010006402162219786010000102003820038200382003820038

Test 3: throughput

Count: 8

Code:

  cmlt v0.4h, v8.4h, #0
  cmlt v1.4h, v8.4h, #0
  cmlt v2.4h, v8.4h, #0
  cmlt v3.4h, v8.4h, #0
  cmlt v4.4h, v8.4h, #0
  cmlt v5.4h, v8.4h, #0
  cmlt v6.4h, v8.4h, #0
  cmlt v7.4h, v8.4h, #0
  movi v8.16b, 9

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

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

retire (01)cycle (02)031e3f51inst 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)cdcfd6e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
802042005715007325801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000111511816200350800001002003920039200392003920039
802042003815002925801081008000810080020500640132120019200382003899776998980120200800322008003220038200381180201100991001008000010000111511816200350800001002003920039200392003920039
802042003815002925801081008000810080020500640132120068200382003899776998980120200800322008003220038200381180201100991001008000010000111511816200350800001002003920039200392003920039
802042003815002925801081008000810080020500640132120019200382003899776998980120200800322008003220038200381180201100991001008000010000111511816200350800001002003920039200392003920039
802042003815002925802041008000810080020500640132120019200382003899776998980120200800322008003220038200381180201100991001008000010000111511816200350800001002003920039200392003920039
802042003815002925801081008000810080020500640132120019200382003899776998980120200800322008003220038200381180201100991001008000010000111511816200350800001002003920039200392003920039
80204200381500119425801081008000810080020500640132120019200382003899776998980120200800322008003220038200381180201100991001008000010000111511816200350800001002003920039200392003920039
8020420038150050425801081008000810080020500640132120019200382003899776998980120200800322008003220038200381180201100991001008000010000111511816200350800001002003920039200392003920039
802042003815005025801081008000810080020500640132120019200382003899776998980120200800322008003220038200381180201100991001008000010000111511816200350800001002003920039200392003920039
802042003815002925801081008000810080020500640132120019200382003899776998980120200800322008003220038200381180201100991001008000010000111511816200350800001002003920039200392003920085

1000 unrolls and 10 iterations

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

retire (01)cycle (02)030708090a0b18191e1f3a3f51inst 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)9fa0a1a6a8a9acc2cfd5d6dbddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420051150000000000760258001010800001080000506400001200192003820038999631001880010208000020800002003820038118002110910108000010000000050201616012122003580000102003920039200392003920039
80024200381500000000003622580010108000010800005064000012001920038200389996310018800102080000208000020038200381180021109101080000100000000502071607112003580000102003920039200392003920039
80024200381500000000003499258001010800001080000506400001200192003820038999631001880010208000020800002003820038118002110910108000010000000050201116013132003580000102003920039200392003920039
8002420038150000000000339528001010800001080000506400001200192003820038999631001880010208000020800002003820038118002110910108000010000000050201116010102003580000102003920039200392003920039
8002420038150000000000239258001010800001080000506400001200192003820038999631001880010208000020800002003820038118002110910108000010000000050201316012132003580000102003920039200392003920039
80024200381500000000002392580010108000010800005064000002001920038200389996310018801092080000208000020038200381180021109101080000100002500050201316011122003580000102003920039200392003920039
800242003815000000000031042580010108000010800005064000012001920038200389996310018800102080000208000020038200381180021109101080000100005103050201216011122003580000102003920039200392003920039
800242003815000000000033925800101080000108000050640000120019200382003899963100188001020800002080000200382003811800211091010800001000039000502011161582003580000102003920039200392003920039
8002420038150000000000383258001010800001080000506400000200192003820038999631001880010208000020800002003820038118002110910108000010000000050201016010102003580000102003920039200392003920039
80024200381500000000003392580010108000010800005064000012001920038200389996310018800102080000208000020038200381180021109101080000100000000502010160972003580000102003920039200392003920039