Icestorm: SHRN2 (4S) measurements

Code:

  shrn2 v0.8h, v1.4s, #3

Initialization

  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

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 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)
1004	3033	1001	1	1000	1000	75905	1000	1000	0	0	2000	1	0	1000	0
1004	3033	1001	1	1000	1000	75905	1000	1000	0	0	2000	1	0	1000	0
1004	3033	1001	1	1000	1000	75905	1000	1000	0	0	2000	1	0	1000	0
1004	3033	1001	1	1000	1000	75905	1000	1000	0	0	2000	1	0	1000	0
1004	3033	1001	1	1000	1000	75905	1000	1000	0	0	2000	1	0	1000	0
1004	3033	1001	1	1000	1000	75905	1000	1000	0	0	2000	1	0	1000	0
1004	3033	1001	1	1000	1000	75905	1000	1000	0	0	2000	1	0	1000	0
1004	3033	1001	1	1000	1000	75905	1000	1000	0	0	2000	1	0	1000	0
1004	3033	1001	1	1000	1000	75905	1000	1000	0	0	2000	1	0	1000	0
1004	3033	1001	1	1000	1000	75905	1000	1000	0	0	2000	1	0	1000	0

Test 2: Latency 1->1

Code:

  shrn2 v0.8h, v1.4s, #3

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 ldst uop inputs (80)	map simd uop inputs (81)	? int output thing (e9)	? ldst retires (ed)	? simd retires (ee)	? int retires (ef)
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10006	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20090	1	0	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	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10

Test 3: Latency 1->2

Code:

  shrn2 v0.8h, v0.4s, #3

Initialization

  movi v0.16b, 1

(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 ldst uop inputs (80)	map simd uop inputs (81)	? int output thing (e9)	? ldst retires (ed)	? simd retires (ee)	? int retires (ef)
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20088	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	10000	100
10204	30033	10101	101	10000	100	10000	300	768905	10100	200	10004	200	0	20008	1	0	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	10004	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	769045	10051	20	10042	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10
10024	30033	10021	21	10000	20	10000	70	768905	10020	20	10000	20	20000	11	10000	10

Test 4: throughput

Count: 8

Code:

  movi v0.16b, 0
  shrn2 v0.8h, v8.4s, #3
  movi v1.16b, 0
  shrn2 v1.8h, v8.4s, #3
  movi v2.16b, 0
  shrn2 v2.8h, v8.4s, #3
  movi v3.16b, 0
  shrn2 v3.8h, v8.4s, #3
  movi v4.16b, 0
  shrn2 v4.8h, v8.4s, #3
  movi v5.16b, 0
  shrn2 v5.8h, v8.4s, #3
  movi v6.16b, 0
  shrn2 v6.8h, v8.4s, #3
  movi v7.16b, 0
  shrn2 v7.8h, v8.4s, #3

Initialization

  movi v8.16b, 9

(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	40476	80109	101	80008	100	80012	300	320052	80112	200	80012	200	160092	1	160000	100
160205	40498	80145	101	80044	100	80048	300	320052	80112	200	80012	200	160096	1	160000	100
160204	40086	80109	101	80008	100	80012	300	320052	80112	200	80012	200	160024	1	160000	100
160204	40086	80109	101	80008	100	80012	300	320052	80112	200	80012	200	160024	1	160000	100
160204	40086	80109	101	80008	100	80012	300	320056	80113	200	80013	200	160024	1	160000	100
160204	40086	80109	101	80008	100	80012	300	320052	80112	200	80012	200	160024	1	160000	100
160204	40086	80109	101	80008	100	80012	300	320052	80112	200	80012	200	160024	1	160000	100
160204	40086	80109	101	80008	100	80012	300	320052	80112	200	80012	200	160024	1	160000	100
160204	40086	80109	101	80008	100	80012	300	320052	80112	200	80012	200	160024	1	160000	100
160204	40086	80109	101	80008	100	80012	300	320052	80112	200	80012	200	160024	1	160000	100

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.5052

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	43852	80019	11	80008	10	80012	30	320000	80010	20	80000	20	160000	1	160000	10
160024	41292	80011	11	80000	10	80000	30	320000	80010	20	80000	20	160000	1	160000	10
160024	40395	80011	11	80000	10	80000	30	320000	80010	20	80000	20	160000	1	160000	10
160024	40416	80011	11	80000	10	80000	30	320204	80060	20	80050	20	160000	1	160000	10
160024	40397	80011	11	80000	10	80000	30	320000	80010	20	80000	20	160000	1	160000	10
160024	40439	80011	11	80000	10	80000	30	320000	80010	20	80000	20	160000	1	160000	10
160024	40416	80011	11	80000	10	80000	30	320000	80010	20	80000	20	160000	1	160000	10
160024	40416	80011	11	80000	10	80000	30	320000	80010	20	80000	20	160000	1	160000	10
160024	40416	80011	11	80000	10	80000	30	320000	80010	20	80000	20	160000	1	160000	10
160024	40416	80011	11	80000	10	80000	30	320000	80010	20	80000	20	160000	1	160000	10

Test 5: throughput

Count: 16

Code:

  shrn2 v0.8h, v16.4s, #3
  shrn2 v1.8h, v16.4s, #3
  shrn2 v2.8h, v16.4s, #3
  shrn2 v3.8h, v16.4s, #3
  shrn2 v4.8h, v16.4s, #3
  shrn2 v5.8h, v16.4s, #3
  shrn2 v6.8h, v16.4s, #3
  shrn2 v7.8h, v16.4s, #3
  shrn2 v8.8h, v16.4s, #3
  shrn2 v9.8h, v16.4s, #3
  shrn2 v10.8h, v16.4s, #3
  shrn2 v11.8h, v16.4s, #3
  shrn2 v12.8h, v16.4s, #3
  shrn2 v13.8h, v16.4s, #3
  shrn2 v14.8h, v16.4s, #3
  shrn2 v15.8h, v16.4s, #3

Initialization

  movi v16.16b, 17

(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	80056	160105	101	160004	100	160008	0	300	0	640036	160108	200	0	160012	200	320028	1	160000	100
160204	80035	160107	101	160006	100	160010	0	300	0	640044	160110	200	0	160014	200	320124	1	160000	100
160204	80035	160105	101	160004	100	160008	0	300	0	640036	160108	200	0	160012	200	320024	1	160000	100
160204	80035	160105	101	160004	100	160008	0	300	0	640036	160108	200	0	160012	200	320024	1	160000	100
160204	80112	160159	101	160058	100	160062	0	300	0	640228	160156	200	0	160060	200	320024	1	160000	100
160204	80035	160105	101	160004	100	160008	0	300	0	640036	160108	200	0	160012	200	320024	1	160000	100
160204	80035	160107	101	160006	100	160010	0	300	0	640036	160108	200	0	160012	200	320024	1	160000	100
160204	80108	160153	101	160052	100	160056	0	300	0	640440	160209	200	0	160113	200	320024	1	160000	100
160205	80223	160241	101	160140	100	160152	0	300	0	640036	160108	200	0	160012	200	320312	1	160000	100
160204	80035	160105	101	160004	100	160008	0	300	0	640036	160108	200	0	160012	200	320024	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)	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	80259	160015	11	160004	10	160008	30	640000	160010	20	160000	20	320000	1	160000	10
160024	80035	160011	11	160000	10	160000	30	640208	160064	20	160064	20	320000	1	160000	10
160024	80035	160011	11	160000	10	160000	30	640000	160010	20	160000	20	320000	1	160000	10
160024	80035	160011	11	160000	10	160000	30	640000	160010	20	160000	20	320000	1	160000	10
160024	80035	160011	11	160000	10	160000	30	640000	160010	20	160000	20	320000	1	160000	10
160024	80035	160011	11	160000	10	160000	30	640000	160010	20	160000	20	320000	1	160000	10
160024	80035	160011	11	160000	10	160000	30	640208	160064	20	160064	20	320000	1	160000	10
160024	80035	160011	11	160000	10	160000	30	640000	160010	20	160000	20	320000	1	160000	10
160024	80035	160011	11	160000	10	160000	30	640000	160010	20	160000	20	320000	1	160000	10
160024	80035	160011	11	160000	10	160000	30	640000	160010	20	160000	20	320000	1	160000	10

SHRN2 (4S)

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: throughput

100 unrolls and 100 iterations

1000 unrolls and 10 iterations

Test 5: throughput

100 unrolls and 100 iterations

1000 unrolls and 10 iterations