Apple M1 Microarchitecture Research by Dougall Johnson

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

FCMEQ (vector, zero, 2S)

Test 1: uops

Code:

  fcmeq 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)acc3cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
1004203715216116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
1004203715156116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
1004203715156116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
1004203715126116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371536116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371506116862510001000100026452112018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038

Test 2: Latency 1->2

Code:

  fcmeq 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)9fa0a1a6a8a9acc2branch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204200371500000000061196862510100100100001001000050028475210200182003720037184213187451010020010000200100002003720037111020110099100100100001000000000071021622197910100001002003820038200382003820038
1020420037150000000176061196862510100100100001001000050028475210200182003720037184213187451010020010000200100002003720037111020110099100100100001000000000071021622197910100001002003820038200382003820038
10204200371500000000061196862510100100100001001000050028475210200182003720037184213187451010020010000200100002003720037111020110099100100100001000000000071021622197910100001002003820038200382003820038
10204200371500000000061196862510100100100001001000050028475210200182003720037184213187451010020010000204100002003720037111020110099100100100001000000000071021622197910100001002003820038200382003820038
10204200371500000000161196862510100100100001001000050028475210200182003720037184213187451010020010000200100002003720037111020110099100100100001000000000071021622197910100001002003820038200382003820038
102042003715000000000840196862510100100100001001000050028475210200182003720037184213187451010020010000200100002003720037111020110099100100100001000000000071021622197910100001002003820038200382003820038
102042003715000000000105196862510100100100001001000050028475210200182003720037184213187451010020010000200100002003720037111020110099100100100001000000000071021622197910100001002003820038200382003820038
10204200371500000000082196862510100100100001001000050028475210200182003720037184217187451010020010000200100002003720037111020110099100100100001000000000071021622197910100001002003820038200382003820038
102042003715000000750061196862510100100100001001000050028475210200182003720037184213187451010020010000200100002003720037111020110099100100100001000000000071021622197910100001002003820038200382003820038
10204200371500000000061196862510100100100001001000050028475210200182003720037184213187451010020010000200100002003720037111020110099100100100001000000000071021622197910100001002003820038200382003820038

1000 unrolls and 10 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)9fa0a6a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100242003715005710319686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100000640316331978610000102003820038200382003820038
10024200371500186119686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000104313913640316331978610000102003820038200382003820038
10024200371501126119686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100000640316331978610000102003820038200382003820038
10024200371500053619686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100000640316331978610000102003820038200382003820038
1002420037150006119686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100000640316331978610000102003820038200382003820038
100242003715001561196862510010101000010100005028475211200182003720037184433187671001020100002010000200372003711100211091010100001000250640316331978610000102003820038200382003820038
10024200371501156119686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100000640316331978610000102003820038200382003820038
1002420037150006119686251001010100001010000502847521020018200372003718443318767100102010000201000020037200371110021109101010000100000640316331978610000102003820038200382003820038
10024200371500186119686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100000640316331978610000102003820038200382003820038
1002420037150066119686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100000640316331978610000102003820038200382003820038

Test 3: throughput

Count: 8

Code:

  fcmeq v0.2s, v8.2s, #0
  fcmeq v1.2s, v8.2s, #0
  fcmeq v2.2s, v8.2s, #0
  fcmeq v3.2s, v8.2s, #0
  fcmeq v4.2s, v8.2s, #0
  fcmeq v5.2s, v8.2s, #0
  fcmeq v6.2s, v8.2s, #0
  fcmeq 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)03070a1e1f3f51inst 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)cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8020420049150114502925801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
80204200381501120102925801081008000810080020500640132120019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181162120035800001002003920039200392003920039
802042003815011002925801081008000810080112500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
802042003815011005225801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
802042003815011002925801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
802042003815011002925801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
8020420038150113302925801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
80204200381501128502925801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039
802042003815011902925801081008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010002011151181161120035800001002003920039200392003920039
8020420038150110069425801961008000810080020500640132020019200382003899776998980120200800322008003220038200381180201100991001008000010000011151181161120035800001002003920039200392003920039

1000 unrolls and 10 iterations

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

retire (01)cycle (02)0308091e3f51inst 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)cfd5d6daddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
800242005115000213925800101080000108000050640000020019200382003899963100188001020800002080000200382003811800211091010800001000005020916022112003580000102003920039200392003920039
80024200381500003925800101080000108000050640000020019200382003899963100188001020800002080000200382003811800211091010800001000005020111609222003580000102003920039200392003920039
800242003815000039258001010800001080000506400000200192003820038999631001880010208000020800002003820038118002110910108000010000050201016022102003580000102003920039200392003920039
800242003815000039258001010800001080000506400001200192003820038999631001880010208000020800002003820038118002110910108000010000050201116022102003580000102003920039200392003920039
800242003815000039258001010800001080000506400000200192003820038999631001880010208000020800002003820038118002110910108000010000050202216022222003580000102003920039200392003920039
800242003815000039258001010800001080000506400001200192003820038999631001880010208000020800002003820038118002110910108000010000050202216011222003580000102003920039200392003920039
800242003815000012325800101080000108000050640000020019200382003899963100188001020800002080000200382003811800211091010800001000005020916022112003580000102003920039200392003920039
80024200381500003925800101080000108000050640000020019200382003899963100188001020800002080000200382003811800211091010800001000005020111609222003580000102003920039200392003920039
800242003815000039258001010800001080000506400000200192003820038999631001880010208000020800002003820038118002110910108000010000050201116011222003580000102003920039200392003920039
800242003815000039258001010800001080000506400001200192003820038999631001880010208000020800002003820038118002110910108000010000050202216011222007480000102003920039200392003920039