Firestorm: SQDMLAL (by element, 4S) measurements

Code:

  sqdmlal v0.4s, 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: 0.000

Load/store unit issues: 0.000

SIMD/FP unit issues: 1.000

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

retire uop (01)	cycle (02)	03	09	1e	3f	4e	51	schedule uop (52)	schedule simd uop (54)	dispatch simd uop (57)	ldst uops in schedulers (5b)	69	6d	6e	map stall dispatch (70)	map rewind (75)	map stall (76)	dispatch uop (78)	map simd uop (7e)	map simd uop inputs (81)	82	83	flush restart other nonspec (84)	85	inst all (8c)	inst simd alu (9a)	l1d cache writeback (a8)	ac	cf	d5	map dispatch bubble (d6)	dd	fetch restart (de)	e0	? simd retires (ee)	f5	f6	f7	f8	fd
1004	3037	22	0	0	61	2548	25	1000	1000	1000	398313	3018	3037	3037	2415	3	2895	1000	1000	3000	3037	3037	1	1	1001	1000	0	0	73	2	16	1	1	2630	1000	3038	3038	3038	3038	3038
1004	3037	23	0	0	61	2548	25	1000	1000	1000	398313	3018	3037	3037	2415	3	2895	1000	1000	3000	3037	3037	1	1	1001	1000	0	0	73	1	16	1	1	2630	1000	3038	3038	3038	3038	3038
1004	3037	22	0	0	61	2548	25	1000	1000	1000	398313	3018	3037	3037	2415	3	2895	1000	1000	3000	3037	3037	1	1	1001	1000	0	0	73	1	16	1	1	2630	1000	3038	3038	3038	3038	3038
1004	3037	22	0	0	61	2548	25	1000	1000	1000	398313	3018	3037	3037	2415	3	2895	1000	1000	3000	3037	3037	1	1	1001	1000	0	6	73	1	16	1	1	2630	1000	3038	3038	3038	3038	3038
1004	3037	23	0	9	61	2548	25	1000	1000	1000	398313	3018	3037	3037	2415	3	2895	1000	1000	3000	3037	3037	1	1	1001	1000	0	0	73	1	16	1	1	2630	1000	3038	3038	3038	3038	3038
1004	3037	22	0	0	61	2548	25	1000	1000	1000	398313	3018	3037	3037	2415	3	2895	1000	1000	3000	3037	3037	1	1	1001	1000	2	0	73	1	16	1	1	2630	1000	3038	3038	3038	3038	3038
1004	3037	23	0	18	61	2548	25	1000	1000	1000	398313	3018	3037	3037	2415	3	2895	1000	1000	3000	3037	3037	1	1	1001	1000	0	0	73	1	16	1	1	2630	1000	3038	3038	3038	3038	3038
1004	3037	24	0	0	61	2548	25	1000	1000	1000	398313	3018	3037	3037	2415	3	2895	1000	1000	3000	3037	3037	1	1	1001	1000	3	0	73	1	16	1	1	2630	1000	3038	3038	3038	3038	3038
1004	3037	23	0	0	61	2548	25	1000	1000	1000	398313	3018	3037	3037	2415	3	2895	1000	1000	3000	3037	3037	1	1	1001	1000	0	0	73	1	16	1	1	2630	1000	3038	3038	3038	3038	3038
1004	3037	23	0	0	61	2548	25	1000	1000	1000	398313	3018	3037	3037	2415	3	2895	1000	1000	3000	3037	3037	1	1	1001	1000	11	0	73	1	16	1	1	2630	1000	3038	3038	3038	3038	3038

Test 2: Latency 1->1

Code:

  sqdmlal v0.4s, 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): 3.0037

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

retire uop (01)	cycle (02)	03	mmu table walk data (08)	09	l2 tlb miss data (0b)	18	19	1e	1f	3a	3f	4e	51	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)	60	69	6d	6e	map stall dispatch (70)	simd prf full (72)	map rewind (75)	map stall (76)	dispatch uop (78)	map int uop (7c)	map simd uop (7e)	map int uop inputs (7f)	map simd uop inputs (81)	82	83	flush restart other nonspec (84)	85	inst all (8c)	inst branch (8d)	inst branch taken (90)	inst branch cond (94)	inst int alu (97)	inst simd alu (9a)	9f	l1d tlb access (a0)	ld unit uop (a6)	l1d cache writeback (a8)	ac	c2	cf	d0	d5	map dispatch bubble (d6)	dd	fetch restart (de)	e0	? int output thing (e9)	? simd retires (ee)	? int retires (ef)	f5	f6	f7	f8	fd
10204	30037	225	0	0	0	0	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	0	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	2	1	10201	100	99	100	100	10000	100	0	0	3	21	2	844	1	3	57	2	4	29776	9	10000	100	30326	30182	30362	30373	30321
10204	30321	227	1	1	0	4	6	924	264	0	251	29548	25	10100	100	10000	100	10000	500	4277313	0	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	710	1	2	16	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	0	0	0	0	350	29548	25	10100	100	10000	100	10000	500	4277313	0	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	1	0	0	710	1	2	16	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	0	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	710	1	2	16	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	0	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	710	1	2	16	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	0	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	710	1	2	16	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	0	18	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	0	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	2	0	0	710	1	2	16	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	0	0	0	0	61	29548	25	10100	100	10008	100	10000	500	4277313	0	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	710	1	2	16	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	0	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	712	1	2	16	2	3	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	0	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	710	1	2	16	2	2	29634	0	10000	100	30038	30038	30038	30038	30038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

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

retire uop (01)	cycle (02)	03	mmu table walk data (08)	18	19	1e	3f	4e	51	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)	60	69	6d	6e	map stall dispatch (70)	map rewind (75)	map stall (76)	dispatch uop (78)	map int uop (7c)	map simd uop (7e)	map int uop inputs (7f)	map simd uop inputs (81)	82	83	flush restart other nonspec (84)	85	inst all (8c)	inst branch (8d)	inst branch taken (90)	inst branch cond (94)	inst int alu (97)	inst simd alu (9a)	9f	l1d tlb miss (a1)	ld unit uop (a6)	l1d cache writeback (a8)	ac	cf	d5	map dispatch bubble (d6)	dd	fetch restart (de)	e0	? int output thing (e9)	? simd retires (ee)	? int retires (ef)	f5	f6	f7	f8	fd
10024	30037	225	0	0	0	0	61	29548	25	10010	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	3	640	3	16	3	3	29630	0	10000	10	30038	30038	30038	30038	30038
10024	30037	224	0	0	0	0	61	29548	25	10010	10	10024	10	10000	50	4284285	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30483	30085	30227	5	1	10021	10	9	10	10	10000	10	0	0	1	0	640	3	16	3	3	29630	0	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	0	0	0	61	29548	25	10010	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	0	640	3	16	3	3	29630	0	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	0	0	0	61	29548	25	10010	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	0	640	3	16	4	3	29630	0	10000	10	30038	30038	30038	30038	30038
10024	30037	224	0	0	0	0	61	29548	25	10010	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	2	1	10021	10	9	10	10	10000	10	0	0	5	3	704	3	24	3	3	29630	1	10000	10	30038	30132	30086	30180	30038
10024	30037	225	0	0	0	0	61	29548	25	10010	10	10000	10	10000	61	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10162	22	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	1	8395	640	3	16	3	4	29630	1	10000	10	30038	30038	30181	30038	30133
10024	30037	225	2	0	1	132	742	29548	45	10066	12	10000	10	10149	50	4281384	1	30018	30037	30037	28287	3	28767	10010	22	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	2	2	4	0	640	3	16	3	3	29630	0	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	1	1	21	732	29548	25	10010	10	10000	10	10447	50	4278670	1	30018	30037	30084	28287	3	28840	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	0	640	3	16	3	3	29630	0	10000	10	30038	30038	30038	30038	30038
10024	30037	225	1	0	0	0	61	29548	25	10018	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	2	1	10021	10	9	10	10	10000	10	0	0	3	0	640	3	16	3	3	29630	0	10000	10	30038	30038	30038	30038	30038
10024	30037	224	0	0	0	0	61	29548	25	10018	10	10000	10	10447	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	6	8374	640	5	16	4	4	29630	0	10000	10	30038	30038	30181	30086	30038

Test 3: Latency 1->2

Code:

  sqdmlal v0.4s, 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): 3.0037

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

retire uop (01)	cycle (02)	03	l2 tlb miss data (0b)	18	19	1e	3f	4e	51	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)	69	6d	6e	map stall dispatch (70)	simd prf full (72)	map rewind (75)	map stall (76)	dispatch uop (78)	map int uop (7c)	map simd uop (7e)	map int uop inputs (7f)	map simd uop inputs (81)	82	83	flush restart other nonspec (84)	85	inst all (8c)	inst branch (8d)	inst branch taken (90)	inst branch cond (94)	inst int alu (97)	inst simd alu (9a)	9f	ld unit uop (a6)	l1d cache writeback (a8)	a9	ac	c2	cd	cf	d0	d5	map dispatch bubble (d6)	db	dd	fetch restart (de)	e0	? int output thing (e9)	? simd retires (ee)	? int retires (ef)	f5	f6	f7	f8	fd
10204	30037	225	0	0	0	0	82	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	0	3	28745	10100	200	10332	204	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	0	710	0	2	16	0	2	2	29634	0	10000	100	30038	30071	30038	30038	30038
10204	30037	225	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	0	710	0	2	16	0	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	0	710	0	2	16	0	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	0	710	0	2	16	0	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	61	29530	25	10100	122	10000	100	10000	500	4277313	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	0	710	0	2	16	0	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	224	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	0	710	0	2	16	0	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	15	0	0	710	0	2	16	0	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	0	710	0	2	16	0	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30084	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	0	710	0	2	16	0	2	2	29634	0	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	0	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	0	0	741	1	2	16	0	2	2	29634	0	10000	100	30038	30038	30038	30038	30038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

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

retire uop (01)	cycle (02)	03	1e	3f	4e	51	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)	60	69	6d	6e	map stall dispatch (70)	simd prf full (72)	map rewind (75)	map stall (76)	dispatch uop (78)	map int uop (7c)	map simd uop (7e)	map int uop inputs (7f)	map simd uop inputs (81)	82	83	flush restart other nonspec (84)	85	inst all (8c)	inst branch (8d)	inst branch taken (90)	inst branch cond (94)	inst int alu (97)	inst simd alu (9a)	9f	l1d cache writeback (a8)	ac	cf	d5	map dispatch bubble (d6)	dd	fetch restart (de)	e0	? simd retires (ee)	? int retires (ef)	f5	f6	f7	f8	fd
10024	30037	224	0	61	29548	25	10010	10	10000	10	10000	50	4277313	0	30018	30037	30037	28287	0	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	61	29548	25	10010	10	10008	10	10000	50	4277313	0	30018	30037	30037	28287	0	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	61	29548	25	10010	10	10000	10	10000	50	4277313	0	30018	30037	30037	28287	0	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	190	29548	25	10010	10	10000	10	10000	50	4277313	0	30018	30037	30037	28287	0	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	61	29548	25	10010	10	10000	10	10000	50	4277313	0	30018	30037	30037	28287	0	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	82	29548	25	10010	10	10000	10	10000	50	4277313	0	30018	30037	30037	28287	0	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	61	29548	25	10010	10	10000	10	10000	50	4277313	0	30018	30037	30037	28287	0	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	6	61	29548	25	10010	10	10000	10	10000	50	4277313	0	30018	30037	30037	28287	0	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	640	2	16	2	2	29630	10000	10	30086	30038	30038	30038	30038
10024	30037	225	0	61	29548	25	10010	10	10000	10	10000	50	4277313	0	30018	30037	30037	28287	0	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	61	29548	25	10010	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	0	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038

Test 4: Latency 1->3

Code:

  sqdmlal v0.4s, 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): 3.0037

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

retire uop (01)	cycle (02)	03	18	1e	1f	3f	4e	51	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)	69	6d	6e	map stall dispatch (70)	map rewind (75)	map stall (76)	dispatch uop (78)	map int uop (7c)	map simd uop (7e)	map int uop inputs (7f)	map simd uop inputs (81)	82	83	flush restart other nonspec (84)	85	inst all (8c)	inst branch (8d)	inst branch taken (90)	inst branch cond (94)	inst int alu (97)	inst simd alu (9a)	9f	ld unit uop (a6)	l1d cache writeback (a8)	a9	ac	cf	d5	map dispatch bubble (d6)	dd	fetch restart (de)	e0	? simd retires (ee)	? int retires (ef)	f5	f6	f7	f8	fd
10204	30037	225	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	710	1	16	1	1	29634	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	3	28745	10100	200	10000	200	30000	30084	30037	2	1	10201	100	99	100	100	10000	100	0	0	0	0	710	1	16	1	1	29634	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	710	1	16	1	1	29634	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	943	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	710	1	16	1	1	29634	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	710	1	16	1	1	29683	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	710	1	16	1	1	29634	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	536	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	710	1	16	1	1	29634	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	150	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	710	1	16	1	1	29634	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	61	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	710	1	16	1	1	29634	10000	100	30038	30038	30038	30038	30038
10204	30037	225	0	0	0	726	29548	25	10100	100	10000	100	10000	500	4277313	30018	30037	30037	28265	3	28745	10100	200	10000	200	30000	30037	30037	1	1	10201	100	99	100	100	10000	100	0	0	0	0	710	1	16	1	1	29634	10000	100	30038	30038	30038	30038	30038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

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

retire uop (01)	cycle (02)	03	1e	3f	4e	51	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)	60	69	6d	6e	map stall dispatch (70)	map rewind (75)	map stall (76)	dispatch uop (78)	map int uop (7c)	map simd uop (7e)	map int uop inputs (7f)	map simd uop inputs (81)	82	83	flush restart other nonspec (84)	85	inst all (8c)	inst branch (8d)	inst branch taken (90)	inst branch cond (94)	inst int alu (97)	inst simd alu (9a)	9f	l1d cache writeback (a8)	a9	ac	cd	cf	d5	map dispatch bubble (d6)	dd	fetch restart (de)	e0	? simd retires (ee)	? int retires (ef)	f5	f6	f7	f8	fd
10024	30037	225	0	363	29548	25	10010	10	10024	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	0	640	3	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	61	29548	25	10010	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	61	29548	25	10010	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	63	61	29548	25	10010	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	61	29548	25	10010	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	61	29548	25	10010	12	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	251	29548	25	10010	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	61	29548	25	10010	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	435	94	29548	25	10010	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	1	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038
10024	30037	225	0	61	29548	25	10010	10	10000	10	10000	50	4277313	1	30018	30037	30037	28287	3	28767	10010	20	10000	20	30000	30037	30037	1	1	10021	10	9	10	10	10000	10	0	0	0	0	640	2	16	2	2	29630	10000	10	30038	30038	30038	30038	30038

Test 5: throughput

Count: 8

Code:

  movi v0.16b, 0
  sqdmlal v0.4s, v8.4h, v9.h[1]
  movi v1.16b, 0
  sqdmlal v1.4s, v8.4h, v9.h[1]
  movi v2.16b, 0
  sqdmlal v2.4s, v8.4h, v9.h[1]
  movi v3.16b, 0
  sqdmlal v3.4s, v8.4h, v9.h[1]
  movi v4.16b, 0
  sqdmlal v4.4s, v8.4h, v9.h[1]
  movi v5.16b, 0
  sqdmlal v5.4s, v8.4h, v9.h[1]
  movi v6.16b, 0
  sqdmlal v6.4s, v8.4h, v9.h[1]
  movi v7.16b, 0
  sqdmlal v7.4s, 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): 0.2508

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

retire uop (01)	cycle (02)	03	1e	3a	3f	51	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)	60	61	69	6d	6e	rob full (74)	map rewind (75)	map stall (76)	dispatch uop (78)	map int uop (7c)	map simd uop (7e)	map int uop inputs (7f)	map simd uop inputs (81)	82	83	flush restart other nonspec (84)	85	inst all (8c)	inst branch (8d)	inst branch taken (90)	inst branch cond (94)	inst int alu (97)	inst simd alu (9a)	9f	l1d tlb miss (a1)	l1d cache writeback (a8)	a9	ac	branch cond mispred nonspec (c5)	cf	d0	d2	d5	map dispatch bubble (d6)	dd	fetch restart (de)	e0	? simd retires (ee)	? int retires (ef)	f5	f6	f7	f8	fd
160204	20089	150	0	0	1199	25	80100	100	80000	100	80000	500	640000	0	0	20045	20064	20064	0	3	22	80100	200	80000	200	240000	20064	20064	1	1	160201	100	99	100	100	160000	100	0	0	0	0	0	10112	0	0	6	16	2	2	20061	160000	100	20065	20065	20065	20065	20065
160204	20064	150	0	0	996	25	80100	100	80000	100	80000	500	640000	0	0	20045	20064	20064	0	3	22	80100	200	80000	200	240000	20064	20064	1	1	160201	100	99	100	100	160000	100	0	0	0	0	0	10112	0	0	2	16	2	2	20061	160000	100	20065	20065	20065	20065	20065
160204	20064	150	0	1	81	25	80100	100	80000	100	80000	500	640000	0	0	20045	20064	20064	0	3	22	80100	200	80000	200	240000	20064	20064	1	1	160201	100	99	100	100	160000	100	0	0	0	0	0	10112	0	0	2	16	2	2	20061	160000	100	20065	20065	20065	20065	20065
160204	20064	150	0	0	764	25	80100	100	80000	100	80000	500	640000	0	0	20045	20064	20064	0	3	22	80100	200	80000	200	240000	20064	20064	1	1	160201	100	99	100	100	160000	100	0	0	0	0	0	10112	0	0	2	58	2	2	20061	160000	100	20065	20065	20065	20065	20065
160204	20064	150	0	0	944	65	80100	100	80000	100	80000	500	640000	1	0	20045	20064	20064	0	3	22	80100	200	80000	200	240000	20064	20064	1	1	160201	100	99	100	100	160000	100	0	17	0	0	0	10112	5	0	2	16	2	2	20061	160000	100	20065	20065	20065	20065	20065
160204	20064	150	0	0	39	25	80100	100	80000	100	80000	500	640000	1	5	20045	20064	20064	0	3	22	80100	200	80000	200	240000	20064	20064	1	1	160201	100	99	100	100	160000	100	0	0	0	0	0	10112	0	0	2	16	2	2	20061	160000	100	20065	20065	20065	20065	20065
160204	20064	151	0	0	39	25	80100	100	80000	100	80000	500	640000	0	5	20045	20064	20064	0	3	22	80100	200	80000	200	240000	20064	20064	1	1	160201	100	99	100	100	160000	100	0	0	0	0	0	10112	0	0	2	16	2	2	20061	160000	100	20065	20065	20065	20065	20065
160204	20064	150	0	0	39	25	80100	100	80000	100	80000	500	640000	0	0	20045	20064	20064	0	3	22	80100	200	80000	200	240000	20064	20064	1	1	160201	100	99	100	100	160000	100	0	0	0	0	0	10112	5	0	2	16	2	2	20061	160000	100	20065	20065	20065	20065	20065
160204	20064	150	0	0	711	25	80100	100	80000	100	80000	500	640000	0	5	20045	20064	20064	0	3	22	80100	200	80000	200	240000	20064	20064	1	1	160201	100	99	100	100	160000	100	0	0	1	0	0	10112	5	0	2	16	2	2	20061	160000	100	20065	20065	20065	20065	20065
160204	20064	151	0	0	39	25	80100	100	80000	100	80000	500	640000	0	0	20045	20064	20064	0	3	22	80100	200	80000	200	240000	20064	20064	1	1	160201	100	99	100	100	160000	100	0	0	0	0	0	10112	0	0	2	16	2	2	20061	160000	100	20065	20065	20065	20065	20065

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2506

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

retire uop (01)	cycle (02)	03	l2 tlb miss data (0b)	18	19	1e	3f	51	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)	5f	60	69	6b	6d	6e	map rewind (75)	map stall (76)	dispatch uop (78)	map int uop (7c)	map simd uop (7e)	map int uop inputs (7f)	map simd uop inputs (81)	82	83	flush restart other nonspec (84)	85	inst all (8c)	inst branch (8d)	inst branch taken (90)	inst branch cond (94)	inst int alu (97)	inst simd alu (9a)	9f	l1d tlb access (a0)	ld unit uop (a6)	st unit uop (a7)	l1d cache writeback (a8)	ac	c2	cd	cf	d0	d2	l1i cache miss demand (d3)	d5	map dispatch bubble (d6)	d9	da	db	dd	fetch restart (de)	e0	? int output thing (e9)	ea	eb	ec	? simd retires (ee)	? int retires (ef)	f5	f6	f7	f8	fd
160024	20084	150	0	0	0	0	45	29	80012	13	80000	12	80000	62	640000	1	1	20032	0	20051	20051	3	22	80012	20	80000	20	240000	20051	20051	1	1	160021	10	9	10	10	160000	10	0	0	0	0	0	0	0	10039	6	2	2	20	25	2	1	1	19	14	20048	2	20	76	5	160000	10	20052	20052	20052	20061	20052
160024	20051	150	0	0	0	0	45	29	80012	12	80000	12	80000	62	640000	0	1	20032	0	20060	20051	3	22	80012	20	80000	20	240000	20051	20051	1	1	160021	10	9	10	10	160000	10	0	0	0	0	0	0	0	10039	6	2	2	13	34	4	2	2	13	13	20057	2	40	68	5	160000	10	20061	20061	20061	20061	20061
160024	20060	151	0	0	0	0	51	29	80012	12	80000	12	80000	62	640000	0	1	20041	0	20060	20060	3	22	80012	20	80000	20	240000	20060	20060	1	1	160021	10	9	10	10	160000	10	0	0	0	0	0	0	0	10044	6	2	2	19	34	4	2	2	12	13	20057	2	40	79	4	160000	10	20061	20061	20061	20061	20061
160024	20060	150	0	0	0	0	45	29	80012	12	80000	12	80000	62	640000	0	1	20041	0	20060	20060	3	22	80012	20	80000	20	240000	20060	20060	1	1	160021	10	9	10	10	160000	10	0	0	0	0	0	0	0	10037	6	2	2	11	34	4	2	1	11	13	20057	2	40	75	5	160000	10	20061	20061	20061	20061	20061
160024	20060	150	0	0	0	0	51	29	80012	12	80000	12	80000	62	640000	0	1	20041	0	20060	20060	3	22	80012	20	80000	20	240000	20060	20060	1	1	160021	10	9	10	10	160000	10	0	0	0	0	0	0	0	10039	6	2	2	13	34	4	2	2	11	13	20057	2	40	81	4	160000	10	20061	20061	20061	20061	20061
160024	20060	150	0	0	0	0	51	29	80012	12	80000	12	80000	62	640000	0	0	20032	0	20060	20060	3	22	80012	20	80000	20	240000	20060	20060	1	1	160021	10	9	10	10	160000	10	0	0	0	0	0	0	0	10040	6	2	2	13	34	4	2	1	19	14	20057	2	40	74	4	160000	10	20061	20061	20061	20061	20061
160024	20060	150	0	0	0	0	51	29	80012	12	80000	12	80000	62	640000	0	0	20041	0	20060	20060	3	22	80012	20	80000	20	240000	20060	20051	1	1	160021	10	9	10	10	160000	10	0	0	0	0	0	0	0	10037	6	2	1	11	34	4	2	2	12	19	20057	2	40	72	4	160000	10	20052	20052	20052	20052	20052
160024	20051	150	0	0	0	0	45	29	80012	12	80000	12	80000	62	640000	0	1	20041	0	20060	20060	3	22	80012	20	80000	20	240000	20060	20060	1	1	160021	10	9	10	10	160000	10	0	0	0	1	0	0	0	10037	6	1	1	14	25	2	1	1	14	13	20048	2	20	79	5	160000	10	20052	20052	20052	20052	20052
160024	20051	151	0	0	0	0	45	29	80012	12	80000	12	80000	62	640000	0	1	20041	0	20060	20060	3	22	80012	20	80000	20	240000	20060	20060	1	1	160021	10	9	10	10	160000	10	0	0	0	0	0	0	0	10036	3	1	1	13	25	2	2	1	12	18	20057	2	20	65	5	160000	10	20052	20052	20052	20052	20052
160024	20051	150	0	0	0	0	45	29	80012	12	80000	12	80000	62	640000	0	1	20041	0	20060	20060	3	22	80012	20	80000	20	240000	20060	20060	1	1	160021	10	9	10	10	160000	10	0	0	0	0	0	0	0	10037	6	2	1	11	34	4	2	2	14	12	20057	2	40	63	5	160000	10	20061	20061	20061	20061	20061

Test 6: throughput

Count: 12

Code:

  sqdmlal v0.4s, v12.4h, v13.h[1]
  sqdmlal v1.4s, v12.4h, v13.h[1]
  sqdmlal v2.4s, v12.4h, v13.h[1]
  sqdmlal v3.4s, v12.4h, v13.h[1]
  sqdmlal v4.4s, v12.4h, v13.h[1]
  sqdmlal v5.4s, v12.4h, v13.h[1]
  sqdmlal v6.4s, v12.4h, v13.h[1]
  sqdmlal v7.4s, v12.4h, v13.h[1]
  sqdmlal v8.4s, v12.4h, v13.h[1]
  sqdmlal v9.4s, v12.4h, v13.h[1]
  sqdmlal v10.4s, v12.4h, v13.h[1]
  sqdmlal v11.4s, 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): 0.2503

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

retire uop (01)	cycle (02)	03	mmu table walk data (08)	09	l2 tlb miss data (0b)	18	19	1e	1f	37	3f	4e	51	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)	60	69	6d	6e	map stall dispatch (70)	map rewind (75)	map stall (76)	dispatch uop (78)	map int uop (7c)	map simd uop (7e)	map int uop inputs (7f)	map simd uop inputs (81)	82	83	flush restart other nonspec (84)	85	inst all (8c)	inst branch (8d)	inst branch taken (90)	inst branch cond (94)	inst int alu (97)	inst simd alu (9a)	9f	l1d tlb access (a0)	l1d tlb miss (a1)	ld unit uop (a6)	l1d cache writeback (a8)	a9	ac	cf	d5	map dispatch bubble (d6)	dd	fetch restart (de)	e0	? int output thing (e9)	? simd retires (ee)	? int retires (ef)	f5	f6	f7	f8	fd
120204	30060	225	0	0	0	0	0	0	0	35	767	0	25	120204	100	120036	100	120000	500	960000	0	30020	30039	31748	16653	3	14997	120100	200	120000	200	360000	31748	30039	1	1	120201	100	99	100	100	120000	100	0	0	0	0	0	0	7610	2	16	2	2	30036	0	120000	100	30040	31749	30040	31749	30040
120205	30039	225	0	0	0	0	0	0	0	18	444	0	25	120101	100	120018	100	120000	500	960000	1	30020	30039	31748	16653	3	16706	120100	200	120000	200	360000	31748	30039	1	1	120201	100	99	100	100	120000	100	0	0	0	0	0	0	7610	2	16	2	2	31745	0	120000	100	31749	30040	30041	30040	30040
120204	30039	238	0	0	0	0	0	0	0	1	167	6713	25	120118	100	120018	100	120000	500	960000	1	30020	31748	30039	14973	3	14997	120100	200	120000	200	360630	30039	30039	1	1	120201	100	99	100	100	120000	100	0	0	0	0	0	0	7610	2	16	2	2	31745	0	120000	100	30040	31749	30040	31749	31749
120204	31748	224	0	0	0	0	0	0	0	0	941	6713	25	120100	100	120000	100	120000	500	990000	1	31729	31748	30039	14973	3	14997	120100	200	120000	200	360000	30039	30040	1	1	120201	100	99	100	100	120000	100	0	0	0	0	0	0	7610	2	16	2	2	31745	0	120000	100	30040	31749	30040	31749	30041
120204	30040	225	0	0	0	0	0	0	0	0	1573	6713	25	120118	100	120001	100	120000	500	960000	1	30020	30039	30039	14973	3	14997	120100	200	120000	200	360000	31748	30039	1	1	120201	100	99	100	100	120000	100	0	0	0	0	0	0	7610	2	16	2	2	30036	0	120000	100	30040	31749	30040	30041	30040
120204	30039	237	0	0	0	0	0	0	0	0	876	6713	25	120118	100	120018	100	120000	500	960000	1	30020	30039	30040	14973	3	16706	120100	200	120000	200	360000	30040	30039	1	1	120201	100	99	100	100	120000	100	0	0	0	0	0	0	7610	2	16	2	2	30036	0	120000	100	31749	30040	30041	30040	30041
120204	31748	225	0	0	0	0	0	0	0	18	3658	6713	25	120118	100	120018	100	120000	500	960000	1	31729	31748	30039	14973	3	14997	120100	200	120000	200	360000	30039	31748	1	1	120201	100	99	100	100	120000	100	0	0	0	0	0	0	7610	2	16	2	2	30037	0	120000	100	30040	30041	30040	31749	31749
120204	31748	225	0	0	0	0	0	0	0	1	932	0	25	120100	100	120000	100	120000	500	990000	0	31729	30039	31748	16653	3	16706	120100	200	120000	200	360000	30039	31748	1	1	120201	100	99	100	100	120000	100	0	0	0	0	0	0	7610	2	16	2	2	30036	0	120000	100	31749	30040	31749	30040	31749
120204	30040	225	0	0	0	0	0	0	0	18	835	6713	25	120118	100	120018	100	120000	500	960000	0	30020	30039	31748	14973	3	16706	120100	200	120000	200	360000	30039	31748	1	1	120201	100	99	100	100	120000	100	0	0	0	0	0	0	7610	2	16	2	2	30036	0	120000	100	30040	30040	30040	30040	30040
120204	30039	225	0	0	0	0	0	0	0	0	180	0	25	120118	100	120018	100	120000	500	960000	0	31729	30040	30039	14973	3	14997	120100	200	120000	200	360000	30039	31010	1	1	120201	100	99	100	100	120000	100	0	0	0	0	0	0	7610	2	16	2	2	30036	0	120000	100	30043	30041	30040	31749	30040

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2503

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

retire uop (01)	cycle (02)	03	l1i tlb fill (04)	mmu table walk instruction (07)	09	l2 tlb miss instruction (0a)	l2 tlb miss data (0b)	18	1e	3a	3f	51	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)	5f	60	69	6d	6e	map stall dispatch (70)	map rewind (75)	map stall (76)	dispatch uop (78)	map int uop (7c)	map simd uop (7e)	map int uop inputs (7f)	map simd uop inputs (81)	82	83	flush restart other nonspec (84)	85	inst all (8c)	inst branch (8d)	inst branch taken (90)	inst branch cond (94)	inst int alu (97)	inst simd alu (9a)	9f	ac	cf	d0	d2	l1i cache miss demand (d3)	l1i tlb miss demand (d4)	d5	map dispatch bubble (d6)	d9	da	db	dd	fetch restart (de)	e0	ea	eb	? simd retires (ee)	? int retires (ef)	f5	f6	f7	f8	fd
120024	30922	225	0	1	0	1	0	0	0	0	226	25	120010	10	120000	10	120000	50	960000	1	1	30020	30039	30039	14996	3	15019	120127	20	120000	20	360000	30039	30039	1	1	120021	10	9	10	10	120000	10	0	7523	3	1	1	1	30	16	1	1	1	14	27	30036	16	5	120000	10	30040	30040	30040	30040	30040
120024	30039	224	1	1	1	0	0	0	0	0	95	25	120010	10	120000	10	120000	50	960000	1	1	30020	30039	30039	14996	29	15017	120010	20	120000	20	360000	30039	30039	1	1	120021	10	9	10	10	120000	10	0	7523	3	1	1	1	26	16	1	1	1	26	26	30036	16	5	120000	10	30040	30040	30040	30040	30040
120024	30039	225	1	1	1	1	0	0	0	1	1268	25	120010	10	120000	10	120000	50	960000	1	1	30020	30039	30039	14996	3	15019	120010	20	120000	20	360000	30039	30039	1	1	120021	10	9	10	10	120000	10	0	7523	3	1	1	1	18	16	1	1	1	26	19	30036	16	5	120000	10	30040	30040	30040	30040	30040
120024	30039	225	1	0	1	1	0	0	0	0	938	25	120010	10	120000	10	120000	50	960000	1	1	30020	30039	30039	14996	3	15019	120010	20	120000	20	360000	30039	30039	1	1	120021	10	9	10	10	120000	10	0	7524	3	1	1	1	18	16	1	1	1	27	16	30036	16	5	120000	10	30040	30040	30040	30040	30040
120024	30039	225	1	0	1	0	0	0	0	1	923	25	120010	10	120000	10	120000	50	960000	1	1	30020	30039	30039	14996	3	15019	120010	20	120000	20	360000	30039	30039	1	1	120021	10	9	10	10	120000	10	0	7524	3	1	1	1	17	16	1	1	1	26	19	30036	16	5	120000	10	30040	30040	30040	30040	30040
120024	30039	224	1	0	1	0	0	0	0	1	953	25	120010	10	120000	10	120000	50	960000	1	1	30020	30039	30039	14996	3	15019	120010	20	120000	20	360000	30039	30039	1	1	120021	10	9	10	10	120000	10	0	7523	3	1	1	1	26	16	1	1	1	17	26	30036	16	5	120000	10	30040	30040	30040	30040	30040
120024	30039	224	1	0	1	0	0	0	0	1	818	25	120010	10	120000	10	120000	50	960000	1	1	30020	30039	30039	14996	3	15019	120010	20	120000	20	360000	30039	30039	1	1	120021	10	9	10	10	120000	10	0	7523	3	1	1	1	19	16	1	1	1	27	16	30036	16	5	120000	10	30040	30040	30040	30040	30040
120024	30039	225	1	1	1	1	0	0	0	0	170	25	120010	10	120000	10	120000	50	960000	1	1	30020	30039	30039	14996	3	15019	120010	20	120000	20	360000	30039	30039	1	1	120021	10	9	10	10	120000	10	0	7523	3	1	1	1	27	16	1	1	1	29	27	30036	16	5	120000	10	30040	30040	30040	30040	30040
120024	30039	225	1	1	1	1	0	0	0	0	945	25	120010	10	120000	10	120000	50	960000	1	1	30020	30039	30039	14996	3	15019	120010	20	120000	20	360000	30039	30039	1	1	120021	10	9	10	10	120000	10	0	7524	3	1	1	1	27	16	1	1	1	28	16	30036	16	5	120000	10	30040	30040	30040	30040	30040
120024	30039	225	1	0	1	1	0	0	0	1	987	25	120010	10	120000	10	120000	50	960000	1	1	30020	30039	30039	14996	3	15019	120010	20	120000	20	360000	30039	30039	1	1	120021	10	9	10	10	120000	10	0	7524	3	1	1	1	26	16	1	1	1	17	25	30036	16	5	120000	10	30040	30040	30040	30040	30040

SQDMLAL (by element, 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: 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