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

FRINTZ (vector, 2D)

Test 1: uops

Code:

  frintz v0.2d, v0.2d
  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: 0.001

Load/store unit issues: 0.000

SIMD/FP unit issues: 1.000

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)schedule simd uop (54)schedule ldst uop (55)dispatch int uop (56)dispatch simd uop (57)ldst uops in schedulers (5b)dispatch uop (78)map simd uop (7e)map simd uop inputs (81)? int output thing (e9)? simd retires (ee)
92983172180724771565273645906157590510001000100011000
100430331001110000010007590510001000100011000
100430331001110000010007590510001000100011000
100430331001110000010007590510001000100011000
100430331001110000010007590510001000100011000
100430331001110000010007590510001000100011000
100430331001110000010007590510001000100011000
100430331001110000010007590510001000100011000
100430331001110000010007590510001000100011000
100430331001110000010007590510001000100011000

Test 2: Latency 1->2

Code:

  frintz v0.2d, v0.2d
  movi v0.16b, 1
  movi v1.16b, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 3.0033

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)schedule simd uop (54)dispatch int uop (56)dispatch simd uop (57)int uops in schedulers (59)ldst uops in schedulers (5b)dispatch uop (78)map int uop (7c)map simd uop (7e)map int uop inputs (7f)map simd uop inputs (81)? int output thing (e9)? simd retires (ee)? int retires (ef)
1020430033101011011000010010000300768905101002001000620210046210000100
1020430033101011011000010010000300768905101002001000420010004110000100
1020430033101011011000010010000300768905101002001000420010004110000100
1020430033101011011000010010000300768905101002001000420010004110000100
1020430033101011011000010010000300768905101002001000420010004110000100
1020430033101011011000010010000300768905101002001000420010004110000100
1020430033101011011000010010000300768905101002001000420010004110000100
1020430033101011011000010010000300768905101002001000420010004110000100
1020430033101011011000010010000300768905101002001000420010004110000100
1020430033101011011000010010000300768905101002001000420010004110000100

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0033

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)schedule simd uop (54)dispatch int uop (56)dispatch simd uop (57)int uops in schedulers (59)ldst uops in schedulers (5b)dispatch uop (78)map int uop (7c)map simd uop (7e)map int uop inputs (7f)map simd uop inputs (81)? int output thing (e9)? simd retires (ee)? int retires (ef)
10024300331002121100002010000707689051002020100042010000111000010
10024300331002121100002010000707689051002020100002010000111000010
10024300331002121100002010000707689051002020100002010000111000010
10024300331002121100002010000707689051002020100002010000111000010
10024300331002121100002010000707689051002020100002010000111000010
10024300331002121100002010000707689051002020100002010000111000010
10024300331002121100002010000707689051002020100002010000111000010
10024300331002121100002010000707689051002020100002010000111000010
10024300331002121100002010000707689051002020100002010000111000010
10024300331002121100002010000707689051002020100002010000111000010

Test 3: throughput

Count: 8

Code:

  frintz v0.2d, v8.2d
  frintz v1.2d, v8.2d
  frintz v2.2d, v8.2d
  frintz v3.2d, v8.2d
  frintz v4.2d, v8.2d
  frintz v5.2d, v8.2d
  frintz v6.2d, v8.2d
  frintz v7.2d, v8.2d
  movi v8.16b, 9

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.5004

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)schedule simd uop (54)dispatch int uop (56)dispatch simd uop (57)int uops in schedulers (59)ldst uops in schedulers (5b)dispatch uop (78)map int uop (7c)map simd uop (7e)map int uop inputs (7f)map simd uop inputs (81)? int output thing (e9)? simd retires (ee)? int retires (ef)
8020440045801071018000610080010300320036801082008001220080012180000100
8020440045801051018000410080008300320044801102008001420080014180000100
8020440035801051018000410080008300320036801082008001220080012180000100
8020440035801051018000410080008300320036801082008001220080012180000100
8020440035801051018000410080008300320036801082008001220080012180000100
8020440035801051018000410080008300320036801082008001220080012180000100
8020440036801071018000610080010300320044801102008001420080014180000100
8020440035801071018000610080010300320036801082008001220080014180000100
8020440035801051018000410080008300320036801082008001220080012180000100
8020440035801051018000410080008300320036801082008001220080012180000100

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.5004

retire uop (01)cycle (02)schedule uop (52)schedule int uop (53)schedule simd uop (54)dispatch int uop (56)dispatch simd uop (57)dispatch ldst uop (58)int uops in schedulers (59)simd uops in schedulers (5a)ldst uops in schedulers (5b)dispatch uop (78)map int uop (7c)map ldst uop (7d)map simd uop (7e)map int uop inputs (7f)map simd uop inputs (81)? int output thing (e9)? simd retires (ee)? int retires (ef)
80024401768002721800062080010070032000080020200800002080000118000010
80024400358002121800002080000070032000080020200800002080000118000010
80024400358002121800002080000070032000080020200800002080000118000010
80024400358002121800002080000070032000080020200800002080000118000010
80024400358002121800002080000070032000080020200800002080000118000010
80024400358002121800002080000070032000080020200800002080000118000010
80024400358002121800002080000070032000080020200800002080000118000010
80024400418002921800082080012070032000080020200800002080000118000010
80024400358002121800002080000070032000080020200800002080000118000010
80024400358002121800002080000070032000080020200800002080000118000010