Apple Microarchitecture Research by Dougall Johnson

M1/A14 P-core (Firestorm): Overview | Base Instructions | SIMD and FP Instructions
M1/A14 E-core (Icestorm):  Overview | Base Instructions | SIMD and FP Instructions

EON (register, lsr, 32-bit)

Test 1: uops

Code:

  eon w0, w0, w1, lsr #17
  mov x0, 1
  mov x1, 2

(no loop instructions)

1000 unrolls and 1 iteration

Retires: 1.000

Issues: 2.000

Integer unit issues: 2.001

Load/store unit issues: 0.000

SIMD/FP unit issues: 0.000

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)dispatch int uop (56)int uops in schedulers (59)dispatch uop (78)map int uop (7c)map int uop inputs (7f)? int output thing (e9)? int retires (ef)
100420302001200110005226510001000200020011000
100420302001200110005226510001000200020011000
100420302001200110005226510001000200020011000
100420302001200110005226510001000200020011000
100420302001200110005226510001000200020011000
100420302001200110005226510001000200020011000
100420302001200110005226510001000200020011000
100420302001200110005226510001000200020011000
100420302001200110005226510001000200020011000
100420302001200110005226510001000200020011000

Test 2: Latency 1->2

Code:

  eon w0, w0, w1, lsr #17
  mov x0, 1
  mov x1, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 2.0030

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)dispatch int uop (56)int uops in schedulers (59)dispatch uop (78)map int uop (7c)map int uop inputs (7f)? int output thing (e9)? int retires (ef)
10204200302010120101101045291001010410210202202000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0030

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)dispatch int uop (56)int uops in schedulers (59)dispatch uop (78)map int uop (7c)map int uop inputs (7f)? int output thing (e9)? int retires (ef)
10025200602003520035100585292201002510034200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010

Test 3: Latency 1->3

Code:

  eon w0, w1, w0, lsr #17
  mov x0, 1
  mov x1, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 2.0030

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)dispatch int uop (56)int uops in schedulers (59)dispatch uop (78)map int uop (7c)map int uop inputs (7f)? int output thing (e9)? int retires (ef)
10204200302010120101101045291001010410210202202000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100
10204200302010120101101045291861010410212202242000110100

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0030

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)dispatch int uop (56)int uops in schedulers (59)dispatch uop (78)map int uop (7c)map int uop inputs (7f)? int output thing (e9)? int retires (ef)
10024200302002120021100255291971002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010
10024200302002120021100205292531002010020200202001110010

Test 4: throughput

Count: 8

Code:

  eon w0, w8, w9, lsr #17
  eon w1, w8, w9, lsr #17
  eon w2, w8, w9, lsr #17
  eon w3, w8, w9, lsr #17
  eon w4, w8, w9, lsr #17
  eon w5, w8, w9, lsr #17
  eon w6, w8, w9, lsr #17
  eon w7, w8, w9, lsr #17
  mov x8, 9
  mov x9, 10

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.6675

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)dispatch int uop (56)int uops in schedulers (59)dispatch uop (78)map int uop (7c)map int uop inputs (7f)? int output thing (e9)? int retires (ef)
8020453429160118160118801311360297801298023416027216001780100
8020453404160117160117801301360838801308023616027216001780100
8020453404160117160117801301360838801308023616027216001780100
8020453404160117160117801301360838801308023616027216001780100
8020453404160117160117801301360838801308023616027216001780100
8020553436160165160165801781360838801308023616027216001780100
8020453404160117160117801301360838801308023616027216001780100
8020453404160117160117801301360838801308023616027216001780100
8020553433160159160159801751360838801308023616027216001780100
8020453404160117160117801301360838801308023616027216001780100

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.6672

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)dispatch int uop (56)int uops in schedulers (59)dispatch uop (78)map int uop (7c)map int uop inputs (7f)? int output thing (e9)? int retires (ef)
8002453378160038160038800501359975800518005816009616002980010
8002453378160039160039800511360232800518005616009216002980010
8002553409160084160084800981360232800518005616009216002980010
8002453378160039160039800511360232800518005616019616007480010
8002453378160039160039800511360232800518005616009216002980010
8002453378160039160039800511360232800518005616009216002980010
8002453378160039160039800511360232800518005616009216002980010
8002453378160039160039800511360232800518005616009216002980010
8002453378160039160039800511360232800518005616009216002980010
8002453378160039160039800511360232800518005616009216002980010