Icestorm: FMLS (by element, 4H) measurements

Code:

  fmls v0.4h, v1.4h, v2.h[1]

Initialization

  movi v0.16b, 1
  movi v1.16b, 2
  movi v2.16b, 3

(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

Recorded non-zero counters (first 10 of 64 runs)

retire uop (01)	cycle (02)	schedule uop (52)	schedule int uop (53)	schedule simd uop (54)	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)
1004	4033	1001	1	1000	1000	101557	1000	1000	3000	1	1000
1004	4033	1001	1	1000	1000	101557	1000	1000	3000	1	1000
1004	4033	1001	1	1000	1000	101557	1000	1000	3000	1	1000
1004	4033	1001	1	1000	1000	101557	1000	1000	3000	1	1000
1004	4033	1001	1	1000	1000	101557	1000	1000	3000	1	1000
1004	4033	1001	1	1000	1000	101557	1000	1000	3000	1	1000
1004	4033	1001	1	1000	1000	101557	1000	1000	3000	1	1000
1004	4033	1001	1	1000	1000	101557	1000	1000	3000	1	1000
1004	4033	1001	1	1000	1000	101557	1000	1000	3000	1	1000
1004	4033	1001	1	1000	1000	101557	1000	1000	3000	1	1000

Test 2: Latency 1->1

Code:

  fmls v0.4h, v1.4h, v2.h[1]

Initialization

  movi v0.16b, 1
  movi v1.16b, 2
  movi v2.16b, 3

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 4.0033

Recorded non-zero counters (first 10 of 64 runs)

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 ldst uop inputs (80)	map simd uop inputs (81)	? int output thing (e9)	? ldst retires (ed)	? simd retires (ee)	? int retires (ef)
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	0	30012	1	0	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	0	30012	1	0	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	0	30012	1	0	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	0	30012	1	0	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	0	30012	1	0	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	0	30012	1	0	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	0	30012	1	0	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	0	30012	1	0	10000	100
10205	40066	10105	101	10004	100	10030	300	1028557	10100	200	10004	200	0	30012	1	0	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	0	30012	1	0	10000	100

1000 unrolls and 10 iterations

Result (median cycles for code): 4.0033

Recorded non-zero counters (first 10 of 64 runs)

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)	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 ldst uop inputs (80)	map simd uop inputs (81)	? int output thing (e9)	? ldst retires (ed)	? simd retires (ee)	? int retires (ef)
10024	40033	10021	21	10000	0	20	10000	70	1028557	10020	20	10004	20	0	30000	11	0	10000	10
10024	40033	10021	21	10000	0	20	10000	70	1028904	10050	20	10044	20	0	30000	11	0	10000	10
10024	40033	10021	21	10000	0	20	10000	70	1028557	10020	20	10000	20	0	30000	11	0	10000	10
10024	40033	10021	21	10000	0	20	10000	70	1028557	10020	20	10000	20	0	30000	11	0	10000	10
10024	40033	10021	21	10000	0	20	10000	70	1028557	10020	20	10000	20	0	30000	11	0	10000	10
10024	40033	10021	21	10000	0	20	10000	70	1028557	10020	20	10000	20	0	30000	11	0	10000	10
10024	40033	10021	21	10000	0	20	10000	70	1028557	10020	20	10000	20	0	30000	11	0	10000	10
10025	40066	10027	23	10004	0	22	10029	70	1028557	10020	20	10000	20	0	30000	11	0	10000	10
10024	40033	10021	21	10000	0	20	10000	70	1028557	10020	20	10000	20	0	30000	11	0	10000	10
10024	40033	10021	21	10000	0	20	10000	70	1028557	10020	20	10000	20	0	30000	11	0	10000	10

Test 3: Latency 1->2

Code:

  fmls v0.4h, v0.4h, v1.h[1]

Initialization

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

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 4.0033

Recorded non-zero counters (first 10 of 64 runs)

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)	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)
10204	40033	10101	101	10000	0	100	10000	300	1028557	10100	200	10006	200	30018	1	10000	100
10204	40033	10101	101	10000	0	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	0	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	0	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	0	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	0	100	10000	300	1028904	10130	200	10044	200	30132	1	10000	100
7083	28020	6998	112	6860	26	105	6886	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	0	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	0	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	0	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100

1000 unrolls and 10 iterations

Result (median cycles for code): 4.0033

Recorded non-zero counters (first 10 of 64 runs)

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)
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10

Test 4: Latency 1->3

Code:

  fmls v0.4h, v1.4h, v0.h[1]

Initialization

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

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 4.0033

Recorded non-zero counters (first 10 of 64 runs)

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)
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10006	200	30018	1	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100
10205	40066	10105	101	10004	100	10030	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100
10204	40033	10101	101	10000	100	10000	300	1028557	10100	200	10004	200	30012	1	10000	100

1000 unrolls and 10 iterations

Result (median cycles for code): 4.0033

Recorded non-zero counters (first 10 of 64 runs)

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)
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1030757	10144	20	10152	20	30000	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30012	11	10000	10
10024	40033	10021	21	10000	20	10000	70	1028557	10020	20	10000	20	30000	11	10000	10

Test 5: throughput

Count: 8

Code:

  movi v0.16b, 0
  fmls v0.4h, v8.4h, v9.h[1]
  movi v1.16b, 0
  fmls v1.4h, v8.4h, v9.h[1]
  movi v2.16b, 0
  fmls v2.4h, v8.4h, v9.h[1]
  movi v3.16b, 0
  fmls v3.4h, v8.4h, v9.h[1]
  movi v4.16b, 0
  fmls v4.4h, v8.4h, v9.h[1]
  movi v5.16b, 0
  fmls v5.4h, v8.4h, v9.h[1]
  movi v6.16b, 0
  fmls v6.4h, v8.4h, v9.h[1]
  movi v7.16b, 0
  fmls v7.4h, v8.4h, v9.h[1]

Initialization

  movi v8.16b, 9
  movi v9.16b, 10

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.5011

Recorded non-zero counters (first 10 of 64 runs)

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)
160204	40480	80110	101	80009	100	80013	300	320056	80113	200	80013	200	240039	1	160000	100
160204	40110	80110	101	80009	100	80013	300	320052	80112	200	80012	200	240036	1	160000	100
160204	40098	80110	101	80009	100	80013	300	320052	80112	200	80012	200	240036	1	160000	100
160204	40086	80109	101	80008	100	80012	300	320052	80112	200	80012	200	240036	1	160000	100
160204	40086	80109	101	80008	100	80012	300	320052	80112	200	80012	200	240036	1	160000	100
160205	40132	80143	101	80042	100	80046	300	320052	80112	200	80012	200	240036	1	160000	100
160204	40110	80110	101	80009	100	80013	300	320052	80112	200	80012	200	240036	1	160000	100
160204	40086	80109	101	80008	100	80012	300	320052	80112	200	80012	200	240036	1	160000	100
160204	40086	80109	101	80008	100	80012	300	320196	80148	200	80048	200	240036	1	160000	100
160204	40086	80109	101	80008	100	80012	300	320052	80112	200	80012	200	240036	1	160000	100

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.5053

Recorded non-zero counters (first 10 of 64 runs)

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)
160025	43981	80030	11	80019	10	80027	30	320000	80010	20	80000	20	240000	1	160000	10
160024	41317	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40450	80011	11	80000	10	80000	30	320200	80059	20	80049	20	240000	1	160000	10
160024	40431	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40409	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40438	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40426	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40437	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40427	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40435	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10

Test 6: throughput

Count: 12

Code:

  fmls v0.4h, v12.4h, v13.h[1]
  fmls v1.4h, v12.4h, v13.h[1]
  fmls v2.4h, v12.4h, v13.h[1]
  fmls v3.4h, v12.4h, v13.h[1]
  fmls v4.4h, v12.4h, v13.h[1]
  fmls v5.4h, v12.4h, v13.h[1]
  fmls v6.4h, v12.4h, v13.h[1]
  fmls v7.4h, v12.4h, v13.h[1]
  fmls v8.4h, v12.4h, v13.h[1]
  fmls v9.4h, v12.4h, v13.h[1]
  fmls v10.4h, v12.4h, v13.h[1]
  fmls v11.4h, v12.4h, v13.h[1]

Initialization

  movi v12.16b, 13
  movi v13.16b, 14

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.5003

Recorded non-zero counters (first 10 of 64 runs)

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)
120204	60086	120107	101	120006	100	120010	300	480044	120110	200	120014	200	360039	1	120000	100
120204	60036	120107	101	120006	100	120010	300	480224	120158	200	120066	200	360042	1	120000	100
120204	60036	120105	101	120004	100	120008	300	480036	120108	200	120012	200	360036	1	120000	100
120204	60036	120105	101	120004	100	120008	300	480036	120108	200	120012	200	360036	1	120000	100
120204	60036	120105	101	120004	100	120008	300	480036	120108	200	120012	200	360036	1	120000	100
120204	60036	120105	101	120004	100	120008	300	480036	120108	200	120012	200	360036	1	120000	100
120204	60036	120105	101	120004	100	120008	300	480036	120108	200	120012	200	360036	1	120000	100
120204	60036	120105	101	120004	100	120008	300	480036	120108	200	120012	200	360036	1	120000	100
120204	60036	120105	101	120004	100	120008	300	480036	120108	200	120012	200	360036	1	120000	100
120204	60036	120105	101	120004	100	120008	300	480036	120108	200	120012	200	360036	1	120000	100

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.5003

Recorded non-zero counters (first 10 of 64 runs)

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)
120024	60282	120015	11	120004	10	120008	30	480044	120020	20	120014	20	360000	1	120000	10
120024	60036	120011	11	120000	10	120000	30	480000	120010	20	120000	20	360000	1	120000	10
120024	60036	120011	11	120000	10	120000	30	480000	120010	20	120000	20	360000	1	120000	10
120024	60036	120011	11	120000	10	120000	30	480000	120010	20	120000	20	360000	1	120000	10
120024	60036	120011	11	120000	10	120000	30	480000	120010	20	120000	20	360000	1	120000	10
120024	60036	120011	11	120000	10	120000	30	480000	120010	20	120000	20	360000	1	120000	10
120024	60036	120011	11	120000	10	120000	30	480000	120010	20	120000	20	360000	1	120000	10
120024	60036	120011	11	120000	10	120000	30	480000	120010	20	120000	20	360000	1	120000	10
120024	60036	120011	11	120000	10	120000	30	480000	120010	20	120000	20	360192	1	120000	10
120024	60146	120017	11	120006	10	120010	30	480000	120010	20	120000	20	360000	1	120000	10

FMLS (by element, 4H)

Test 1: uops

1000 unrolls and 1 iteration

Test 2: Latency 1->1

100 unrolls and 100 iterations

1000 unrolls and 10 iterations

Test 3: Latency 1->2

100 unrolls and 100 iterations

1000 unrolls and 10 iterations

Test 4: Latency 1->3

100 unrolls and 100 iterations

1000 unrolls and 10 iterations

Test 5: throughput

100 unrolls and 100 iterations

1000 unrolls and 10 iterations

Test 6: throughput

100 unrolls and 100 iterations

1000 unrolls and 10 iterations