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, 2S)

Test 1: uops

Code:

  cmlt v0.2s, v0.2s, #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
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110001073116111786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110001073116111786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000973116111786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371606116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371696116862510001000100026452102018203720371571318951000100010002037203711100110000373116111786100020382038203820382038

Test 2: Latency 1->2

Code:

  cmlt v0.2s, v0.2s, #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)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)9fa1a6a7a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020420037150000009006119686251010010010000100100005002847521120018200372003718421318745101002001000020010000200372003711102011009910010010000100000000071011612197910100001002003820038200382003820038
1020420037150000000006119686441010010010000100100005002847521020018200372003718421318745101002001000020010000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038
1020420037150000009006119686251010010010000100100005002847521120018200372003718421318745101002001000020010000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038
102042003715000000342006119686251010010010000100100005002847521020018200372003718421318745101002001000020010000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038
102042003715000000150061196862510100100100001001000050028475211200182003720037184213187451010020010000200100002003720037111020110099100100100001000003900071011611197910100001002003820038200382003820038
1020420037150000000006119686251010010010000100101525002847521120018200372003718421318745101002001000020010000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038
1020420037150000000006119686251010010010000100100005002847521120018200372003718421318745101002001000020010000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038
1020420037150000000006119686251010010010000100100005002847521120018200372003718421318745101002001000020010000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038
102042003715000000369006119686251010010010000100100005002847521120018200372003718421318745101002001000020010000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038
1020420037150000000006119686251010010010000100100005002847521020018200372003718421318745101002001000020010000200372003711102011009910010010000100000000071011611197910100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)03090b181e3f4e51inst 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
10024200371502000611968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010000640816221978610000102003820038200382003820038
10024200371500000611968625100101010000101000050284752112001820037200371844331876710010201000020100002003720037111002110910101000010000640316221978610000102003820038200382003820038
10024200371500000611968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010000668216221978610000102003820038200382003820038
100242013315000024611968625100101010000101000050284752102001820037200371844331882310010201000020100002003720037111002110910101000010000640216221978610000102003820038200382003820038
10024200371500000611968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010000640216221978610000102003820038200382003820038
10024200371500000611968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010000640216221978610000102003820038200382003820086
10024200371500000611968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010000640216221978610000102003820038200382003820038
10024200371500000611968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010000640216221978610000102003820038200382003820038
100242003715000048611968625100101010000101000050284752102001820037200371844331876710010201000020100002003720037111002110910101000010000640216221978610000102003820038200382003820038
1002420037150002399611968625100101010000101000050285004902001820037200371844331876710010201000020100002003720037111002110910101000010030640216221978610000102003820038200382003820038

Test 3: throughput

Count: 8

Code:

  cmlt v0.2s, v8.2s, #0
  cmlt v1.2s, v8.2s, #0
  cmlt v2.2s, v8.2s, #0
  cmlt v3.2s, v8.2s, #0
  cmlt v4.2s, v8.2s, #0
  cmlt v5.2s, v8.2s, #0
  cmlt v6.2s, v8.2s, #0
  cmlt v7.2s, v8.2s, #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)03070a18191e1f3f51inst 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)9fa8acc2c5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
80204200571501100002925801081008000810080020500640132120019200382003899776998980120200800322008003220038200381180201100991001008000010000011151182161120035800001002003920039200392003920039
802042003815011008102925801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
80204200381501100002925801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
80204200381501100002925801081008000810080020500640132120019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
80204200381501100002925801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
80204200381501100002925801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
80204200381501100002925801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
80204200381501100002925801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
802042003815011000069425801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
80204200381501100002925801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039

1000 unrolls and 10 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)9faccfd5d6daddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420039150405392580010108000010800005064000002001920038200389996310018800102080000208000020038200381180021109101080000100502013160632003580000102003920039200392003920039
8002420038150039258001010800001080000506400000200192003820038999631001880010208000020800002003820038118002110910108000010050203160692003580000102003920039200392003920039
800242003815003925800101080000108000050640000020019200382003899963100188001020800002080000200382003811800211091010800001005020101601052003580000102003920039200392003920039
80024200381500392580010108000010800005064000002001920038200389996310018800102080000208000020038200381180021109101080000100502051601072003580000102003920039200392003920039
800242003815003925800101080000108000050640000020019200382003899963100188001020800002080000200382003811800211091010800001005020121601162003580000102003920039200392003920039
80024200381500392580010108000010800005064000002001920038200389996310018800102080000208000020038200381180021109101080000100502051605112003580000102003920039200392003920039
8002420038150039258001010800001080000506400000200192003820038999631001880010208000020800002003820038118002110910108000010050206160452003580000102003920039200392003920039
8002420038150039258001010800001080000506400000200192003820038999631001880010208000020800002003820038118002110910108000010350206160542003580000102003920039200392003920039
80024200381500392580010108000010800005064000002001920038200389996310018800102080000208000020038200381180021109101080000100502041605112003580000102003920039200392003920039
8002420038150039258001010800001080000506400000200192003820038999631001880010208000020800002003820038118002110910108000010350206160592003580000102003920039200392003920039