Icestorm: SMLAL2 (vector, 8H) measurements

Code:

  smlal2 v0.8h, v1.16b, v2.16b

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:

  smlal2 v0.8h, v1.16b, v2.16b

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	10004	200	30012	1	10000	100
10205	30066	10109	103	10006	102	10031	300	768905	10100	200	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	206	30486	4	10000	100
10205	30066	10107	101	10006	100	10031	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	202	30138	2	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
10205	30066	10109	103	10006	102	10031	300	768905	10100	200	10004	200	30018	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)
10025	30066	10027	21	10006	20	10031	70	768905	10020	20	10004	20	30012	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 3: Latency 1->2

Code:

  smlal2 v0.8h, v0.16b, v1.16b

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)	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)
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10006	200	30018	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10006	200	30012	1	10000	100
10204	30084	10109	101	10008	100	10033	0	300	0	768905	10100	200	0	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	769451	10133	200	0	10044	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	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	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
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
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	30012	11	10000	10
10024	30033	10021	21	10000	0	20	10000	70	768905	10020	20	10000	20	30000	11	10000	10

Test 4: Latency 1->3

Code:

  smlal2 v0.8h, v1.16b, v0.16b

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)	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)
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10006	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	769247	10131	200	0	10048	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	1555	63109	18580	768126	14411	3869	1837	10058	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10004	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	10005	200	30012	1	10000	100
10204	30033	10101	101	10000	100	10000	0	300	0	768905	10100	200	0	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	10006	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	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 5: throughput

Count: 8

Code:

  movi v0.16b, 0
  smlal2 v0.8h, v8.16b, v9.16b
  movi v1.16b, 0
  smlal2 v1.8h, v8.16b, v9.16b
  movi v2.16b, 0
  smlal2 v2.8h, v8.16b, v9.16b
  movi v3.16b, 0
  smlal2 v3.8h, v8.16b, v9.16b
  movi v4.16b, 0
  smlal2 v4.8h, v8.16b, v9.16b
  movi v5.16b, 0
  smlal2 v5.8h, v8.16b, v9.16b
  movi v6.16b, 0
  smlal2 v6.8h, v8.16b, v9.16b
  movi v7.16b, 0
  smlal2 v7.8h, v8.16b, v9.16b

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	40542	80109	101	80008	100	80012	300	320056	80113	200	80013	200	240039	1	160000	100
160204	40096	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
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	43844	80020	11	80009	10	80013	30	320192	80057	20	80047	20	240039	1	160000	10
160024	41271	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40423	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40423	80011	11	80000	10	80000	30	320056	80023	20	80013	20	240000	1	160000	10
160024	40522	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40398	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40400	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40507	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40429	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10
160024	40463	80011	11	80000	10	80000	30	320000	80010	20	80000	20	240000	1	160000	10

Test 6: throughput

Count: 16

Code:

  smlal2 v0.8h, v16.16b, v17.16b
  smlal2 v1.8h, v16.16b, v17.16b
  smlal2 v2.8h, v16.16b, v17.16b
  smlal2 v3.8h, v16.16b, v17.16b
  smlal2 v4.8h, v16.16b, v17.16b
  smlal2 v5.8h, v16.16b, v17.16b
  smlal2 v6.8h, v16.16b, v17.16b
  smlal2 v7.8h, v16.16b, v17.16b
  smlal2 v8.8h, v16.16b, v17.16b
  smlal2 v9.8h, v16.16b, v17.16b
  smlal2 v10.8h, v16.16b, v17.16b
  smlal2 v11.8h, v16.16b, v17.16b
  smlal2 v12.8h, v16.16b, v17.16b
  smlal2 v13.8h, v16.16b, v17.16b
  smlal2 v14.8h, v16.16b, v17.16b
  smlal2 v15.8h, v16.16b, v17.16b

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)	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)
160204	80085	160105	101	160004	100	160008	0	300	0	640044	160110	200	0	160014	200	480042	1	160000	100
160204	80035	160107	101	160006	100	160010	0	300	0	640036	160108	200	0	160012	200	480036	1	160000	100
160204	80035	160105	101	160004	100	160008	0	300	0	640036	160108	200	0	160012	200	480036	1	160000	100
160204	80035	160105	101	160004	100	160008	0	300	0	640036	160108	200	0	160012	200	480036	1	160000	100
160204	80035	160105	101	160004	100	160008	0	300	0	640044	160110	200	0	160013	200	480039	1	160000	100
160204	80035	160105	101	160004	100	160008	0	300	0	640036	160108	200	0	160012	200	480036	1	160000	100
160204	80035	160105	101	160004	100	160008	0	300	0	640036	160108	200	0	160012	200	480036	1	160000	100
160204	80035	160105	101	160004	100	160008	0	300	0	640036	160108	200	0	160012	200	480036	1	160000	100
160204	80035	160105	101	160004	100	160008	0	300	0	640036	160108	200	0	160012	200	480180	1	160000	100
160204	80035	160105	101	160004	100	160008	0	300	0	640036	160108	200	0	160012	200	480039	1	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)	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)
160024	80264	160015	11	160004	0	10	160008	30	640044	160020	20	160014	20	480000	1	160000	10
160024	80035	160011	11	160000	0	10	160000	30	640000	160010	20	160000	20	480000	1	160000	10
160024	80035	160011	11	160000	0	10	160000	30	640000	160010	20	160000	20	480195	1	160000	10
160024	80035	160011	11	160000	0	10	160000	30	640000	160010	20	160000	20	480000	1	160000	10
160024	80035	160011	11	160000	0	10	160000	30	640000	160010	20	160000	20	480000	1	160000	10
160024	80035	160011	11	160000	0	10	160000	30	640000	160010	20	160000	20	480195	1	160000	10
160024	80035	160011	11	160000	0	10	160000	30	640000	160010	20	160000	20	480000	1	160000	10
207339	230498	201291	24910	160345	16036	24005	160343	30	640000	160010	20	160000	20	480000	1	160000	10
160024	80035	160011	11	160000	0	10	160000	30	640000	160010	20	160000	20	480000	1	160000	10
160024	80035	160011	11	160000	0	10	160000	30	640000	160010	20	160000	20	480000	1	160000	10

SMLAL2 (vector, 8H)

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