Icestorm: FCCMP (scalar, H) measurements

Code:

  fccmp h0, h1, #0, lt

Initialization

  mov x0, 1
  mov x1, 2
  mov x2, 3
  mov x3, 4

(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)
1004	2033	1001	1	1000	1000	24769	1000	1000	3000	1
1004	2033	1001	1	1000	1000	24769	1000	1000	3000	1
1004	2033	1001	1	1000	1000	24769	1000	1000	3000	1
1004	2033	1001	1	1000	1000	24769	1000	1000	3000	1
1004	2033	1001	1	1000	1000	24769	1000	1000	3000	1
1004	2033	1001	1	1000	1000	24769	1000	1000	3000	1
1004	2033	1001	1	1000	1000	24769	1000	1000	3000	1
1004	2033	1001	1	1000	1000	24769	1000	1000	3000	1
1004	2033	1001	1	1000	1000	24769	1000	1000	3000	1
1004	2033	1001	1	1000	1000	24769	1000	1000	3000	1

Test 2: Latency 3->1

Chain cycles: 2

Code:

  fccmp h0, h1, #0, lt
  fcsel d0, d2, d3, eq

Initialization

  mov x0, 1
  mov x1, 2
  mov x2, 3
  mov x3, 4

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code, minus 2 chain cycles): 2.0033

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)
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20006	200	60018	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	202	60387	2	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100
20205	40066	20111	103	20008	102	20034	300	1019248	20100	200	20004	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100

1000 unrolls and 10 iterations

Result (median cycles for code, minus 2 chain cycles): 2.0033

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)
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20025	40066	20019	11	20008	10	20034	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019580	20044	20	20048	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10

Test 3: Latency 3->2

Chain cycles: 2

Code:

  fccmp h0, h1, #0, lt
  fcsel d1, d2, d3, eq

Initialization

  mov x0, 1
  mov x1, 2
  mov x2, 3
  mov x3, 4

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code, minus 2 chain cycles): 2.0033

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)
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20006	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60144	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100
20204	40033	20101	101	20000	100	20000	300	1019248	20100	200	20004	200	60012	1	10000	100

1000 unrolls and 10 iterations

Result (median cycles for code, minus 2 chain cycles): 2.0033

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)
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20006	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10
20024	40033	20011	11	20000	10	20000	30	1019248	20010	20	20000	20	60000	1	10000	10

Test 4: Latency 3->3

Code:

  fccmp h0, h1, #0, lt

Initialization

  mov x0, 1
  mov x1, 2
  mov x2, 3
  mov x3, 4

(non-fused SUB/CBNZ loop)

100 unrolls and 100 iterations

Result (median cycles for code): 2.0033

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)	? int retires (ef)
10204	20033	10201	201	10000	200	10000	700	249768	10200	200	10006	200	30012	101	100
10204	20033	10201	201	10000	200	10000	700	249769	10200	200	10004	200	30012	101	100
10204	20033	10201	201	10000	200	10000	700	249769	10200	200	10004	200	30012	101	100
10204	20033	10201	201	10000	200	10000	700	249769	10200	200	10004	200	30012	101	100
10204	20033	10201	201	10000	200	10000	700	249769	10200	200	10004	200	30012	101	100
10204	20033	10201	201	10000	200	10000	700	249769	10200	200	10004	200	30012	101	100
10204	20033	10201	201	10000	200	10000	700	249769	10200	200	10004	200	30012	101	100
10204	20033	10201	201	10000	200	10000	700	249769	10200	200	10004	200	30012	101	100
10204	20033	10201	201	10000	200	10000	700	249769	10200	200	10004	200	30012	101	100
10204	20033	10201	201	10000	200	10000	700	249769	10200	200	10004	200	30012	101	100

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0033

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)	? int retires (ef)
10024	20033	10021	21	10000	20	10000	67	249906	10041	20	10034	20	30018	11	10
10024	20033	10021	21	10000	20	10000	70	249769	10020	20	10000	20	30000	11	10
10024	20033	10021	21	10000	20	10000	70	249769	10020	20	10000	20	30000	11	10
10024	20033	10021	21	10000	20	10000	70	249769	10020	20	10000	20	30000	11	10
10024	20033	10021	21	10000	20	10000	70	249769	10020	20	10000	20	30000	11	10
10024	20033	10021	21	10000	20	10000	70	249769	10020	20	10000	20	30000	11	10
10024	20033	10021	21	10000	20	10000	70	249769	10020	20	10000	20	30000	11	10
10024	20033	10021	21	10000	20	10000	70	249769	10020	20	10000	20	30000	11	10
10024	20033	10021	21	10000	20	10000	70	249769	10020	20	10000	20	30000	11	10
10024	20033	10021	21	10000	20	10000	70	249769	10020	20	10000	20	30000	11	10

Test 5: throughput

Count: 8

Code:

  ands xzr, xzr, xzr
  fccmp h0, h1, #0, lt
  ands xzr, xzr, xzr
  fccmp h0, h1, #0, lt
  ands xzr, xzr, xzr
  fccmp h0, h1, #0, lt
  ands xzr, xzr, xzr
  fccmp h0, h1, #0, lt
  ands xzr, xzr, xzr
  fccmp h0, h1, #0, lt
  ands xzr, xzr, xzr
  fccmp h0, h1, #0, lt
  ands xzr, xzr, xzr
  fccmp h0, h1, #0, lt
  ands xzr, xzr, xzr
  fccmp h0, h1, #0, lt

Initialization

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

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 1.0736

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)
160204	85868	160111	80108	80003	80107	80007	240321	560037	160114	80207	80007	200	0	240021	80008	0	0	100
160204	85896	160111	80108	80003	80107	80007	240318	560031	160111	80206	80006	200	0	240018	80005	0	0	100
160205	84446	160146	80127	80019	80126	80025	240321	477252	160113	80207	80006	200	0	240021	80008	0	0	100
160204	85886	160107	80105	80002	80106	80005	240318	560031	160111	80206	80006	200	0	240018	80005	0	0	100
160204	85886	160107	80105	80002	80106	80005	240318	560031	160111	80206	80006	200	0	240018	80008	0	0	100
160204	85886	160107	80105	80002	80106	80005	240318	560031	160111	80206	80006	200	0	240018	80005	0	0	100
160204	85886	160107	80105	80002	80106	80005	240318	560031	160111	80206	80006	200	0	240018	80005	0	0	100
160204	85886	160107	80105	80002	80106	80005	240318	560031	160111	80206	80006	200	0	240018	80005	0	0	100
160205	85424	160146	80127	80019	80126	80025	240318	560031	160111	80206	80006	200	0	240018	80005	0	0	100
160204	83938	160111	80108	80003	80107	80006	240318	560031	160111	80206	80006	200	0	240018	80005	0	0	100

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 1.0482

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	85711	160017	80015	80002	80016	80005	240045	551734	160020	80025	80005	20	0	240000	80001	0	0	10
160024	85576	160011	80011	80000	80010	80000	240030	476419	160010	80020	80000	20	0	240000	80001	0	0	10
160024	83856	160011	80011	80000	80010	80000	240030	476419	160010	80020	80000	20	0	240000	80001	0	0	10
160024	83856	160011	80011	80000	80010	80000	240030	476419	160010	80020	80000	20	0	240000	80001	0	0	10
160024	83857	160011	80011	80000	80010	80000	240030	551697	160010	80020	80000	20	0	240000	80001	0	0	10
160024	86078	160017	80015	80002	80016	80005	240030	535099	160010	80020	80000	20	0	240000	80001	0	0	10
160024	83856	160011	80011	80000	80010	80000	240030	476429	160010	80020	80000	20	0	240000	80001	0	0	10
160024	83856	160011	80011	80000	80010	80000	240030	476429	160010	80020	80000	20	0	240000	80001	0	0	10
160024	83862	160011	80011	80000	80010	80000	240108	478644	160059	80046	80024	1670	1257	240126	80744	596	3	940
160024	83853	160011	80011	80000	80010	80000	240030	476428	160010	80020	80000	20	0	240000	80001	0	0	10

FCCMP (scalar, H)

Test 1: uops

1000 unrolls and 1 iteration

Test 2: Latency 3->1

100 unrolls and 100 iterations

1000 unrolls and 10 iterations

Test 3: Latency 3->2

100 unrolls and 100 iterations

1000 unrolls and 10 iterations

Test 4: Latency 3->3

100 unrolls and 100 iterations

1000 unrolls and 10 iterations

Test 5: throughput

100 unrolls and 100 iterations

1000 unrolls and 10 iterations