Icestorm: SDOT (by element, 16B) measurements

Code:

  sdot v0.4s, v1.16b, v2.4b[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	3033	1001	1	1000	1000	75905	1000	1000	3000	1	1000
1004	3033	1001	1	1000	1000	75905	1000	1000	3000	1	1000
1004	3033	1001	1	1000	1000	75905	1000	1000	3000	1	1000
1004	3033	1001	1	1000	1000	75905	1000	1000	3000	1	1000
1004	3033	1001	1	1000	1000	75905	1000	1000	3000	1	1000
1004	3033	1001	1	1000	1000	75905	1000	1000	3000	1	1000
1004	3033	1001	1	1000	1000	75905	1000	1000	3000	1	1000
1004	3033	1001	1	1000	1000	75905	1000	1000	3000	1	1000
1004	3033	1001	1	1000	1000	75905	1000	1000	3000	1	1000
1004	3033	1001	1	1000	1000	75905	1000	1000	3000	1	1000

Test 2: Latency 1->1

Code:

  sdot v0.4s, v1.16b, v2.4b[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): 3.0033

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	30033	10101	101	10000	100	10000	300	768905	10100	200	10006	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0033

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)
10024	30033	10021	21	10000	0	20	10000	70	768905	10020	20	10004	20	30000	11	10000	10
10024	30033	10021	21	10000	0	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	0	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	0	20	10000	70	768905	10020	20	10000	20	30132	11	10000	10
10024	30033	10021	21	10000	0	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	0	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	0	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10025	30066	10027	21	10006	0	20	10031	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	0	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	0	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10

Test 3: Latency 1->2

Code:

  sdot v0.4s, v0.16b, v1.4b[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): 3.0033

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	30033	10101	101	10000	100	10000	300	768905	10100	200	10006	200	30018	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0033

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	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10

Test 4: Latency 1->3

Code:

  sdot v0.4s, v1.16b, v0.4b[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): 3.0033

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	30033	10101	101	10000	100	10000	300	768905	10100	200	10006	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30018	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30015	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30129	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	30012	1	10000	100

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0033

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	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10025	30066	10027	21	10006	20	10031	69	768905	10020	20	10004	20	30249	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	65	769247	10051	20	10048	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10

Test 5: throughput

Count: 8

Code:

  movi v0.16b, 0
  sdot v0.4s, v8.16b, v9.4b[1]
  movi v1.16b, 0
  sdot v1.4s, v8.16b, v9.4b[1]
  movi v2.16b, 0
  sdot v2.4s, v8.16b, v9.4b[1]
  movi v3.16b, 0
  sdot v3.4s, v8.16b, v9.4b[1]
  movi v4.16b, 0
  sdot v4.4s, v8.16b, v9.4b[1]
  movi v5.16b, 0
  sdot v5.4s, v8.16b, v9.4b[1]
  movi v6.16b, 0
  sdot v6.4s, v8.16b, v9.4b[1]
  movi v7.16b, 0
  sdot v7.4s, v8.16b, v9.4b[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	40501	80109	101	80008	100	80012	300	320214	80154	200	80054	200	240039	1	160000	100
160204	40119	80110	101	80009	100	80013	300	320052	80112	200	80012	200	240039	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
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
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
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

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)
160024	43868	80019	11	80008	10	80012	30	320000	80010	20	80000	20	240000	1	160000	10
160024	41278	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40385	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40407	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40416	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40403	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40419	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40414	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
160025	40595	80055	11	80044	10	80048	30	320000	80010	20	80000	20	240000	1	160000	10

Test 6: throughput

Count: 16

Code:

  sdot v0.4s, v16.16b, v17.4b[1]
  sdot v1.4s, v16.16b, v17.4b[1]
  sdot v2.4s, v16.16b, v17.4b[1]
  sdot v3.4s, v16.16b, v17.4b[1]
  sdot v4.4s, v16.16b, v17.4b[1]
  sdot v5.4s, v16.16b, v17.4b[1]
  sdot v6.4s, v16.16b, v17.4b[1]
  sdot v7.4s, v16.16b, v17.4b[1]
  sdot v8.4s, v16.16b, v17.4b[1]
  sdot v9.4s, v16.16b, v17.4b[1]
  sdot v10.4s, v16.16b, v17.4b[1]
  sdot v11.4s, v16.16b, v17.4b[1]
  sdot v12.4s, v16.16b, v17.4b[1]
  sdot v13.4s, v16.16b, v17.4b[1]
  sdot v14.4s, v16.16b, v17.4b[1]
  sdot v15.4s, v16.16b, v17.4b[1]

Initialization

  movi v16.16b, 17
  movi v17.16b, 18

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.5002

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)
160204	80074	160108	101	160007	0	100	160011	300	640044	160110	200	160014	200	0	480039	1	0	160000	100
160204	80035	160107	101	160006	0	100	160010	300	640036	160108	200	160012	200	0	480036	1	0	160000	100
160204	80035	160105	101	160004	0	100	160008	300	640036	160108	200	160012	200	0	480036	1	0	160000	100
160204	80035	160105	101	160004	0	100	160008	300	640036	160108	200	160012	200	0	480036	1	0	160000	100
160204	80035	160105	101	160004	0	100	160008	300	640036	160108	200	160012	200	0	480036	1	0	160000	100
160204	80035	160105	101	160004	0	100	160008	300	640036	160108	200	160012	200	0	480036	1	0	160000	100
160204	80035	160105	101	160004	0	100	160008	300	640036	160108	200	160012	200	0	480186	1	0	160000	100
160204	80035	160107	101	160006	0	100	160010	300	640036	160108	200	160012	200	0	480036	1	0	160000	100
160204	80035	160105	101	160004	0	100	160008	300	640036	160108	200	160012	200	0	480036	1	0	160000	100
160204	80035	160105	101	160004	0	100	160008	300	640036	160108	200	160012	200	0	480036	1	0	160000	100

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.5002

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)
160024	80156	160017	11	160006	10	160010	30	640000	160010	20	160000	20	0	480000	1	0	160000	10
160024	80036	160011	11	160000	10	160000	30	640000	160010	20	160000	20	0	480000	1	0	160000	10
160024	80035	160011	11	160000	10	160000	30	640216	160066	20	160063	20	0	480000	1	0	160000	10
160024	80035	160011	11	160000	10	160000	30	640000	160010	20	160000	20	0	480000	1	0	160000	10
160024	80035	160011	11	160000	10	160000	30	640000	160010	20	160000	20	0	480000	1	0	160000	10
160024	80035	160011	11	160000	10	160000	30	640000	160010	20	160000	20	0	480000	1	0	160000	10
160024	80035	160011	11	160000	10	160000	30	640000	160010	20	160000	20	0	480000	1	0	160000	10
160024	80035	160011	11	160000	10	160000	30	640000	160010	20	160000	20	0	480000	1	0	160000	10
160024	80035	160011	11	160000	10	160000	30	640000	160010	20	160000	20	0	480000	1	0	160000	10
160024	80036	160011	11	160000	10	160000	30	640040	160019	20	160013	20	0	480192	1	0	160000	10

SDOT (by element, 16B)

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