all commits for a clean git history

This commit is contained in:
2024-09-18 14:22:27 +01:00
commit 18306c66fc
6128 changed files with 3186840 additions and 0 deletions

View File

@@ -0,0 +1,396 @@
import matplotlib.pyplot as plt
import numpy as np
dim3_physically_contigous = np.array([92265])
dmin_3_physically_contigous = sum(dim3_physically_contigous)
dim3_regular = np.array([94650])
dmin_3_regular = sum(dim3_regular)
dim6_contigous = np.array(np.array([int(x) for x in """0
0
202420
108748
7366843
12350564
13586069
6779734
56950""".replace(' ',',').replace('\n','').split(",")]))
dim_6_contigous = sum(dim6_contigous)
dim6_regular = np.array(np.array([int(x) for x in """105883
0
135218
0
8322082
12700590
12017926
6160362
134527""".replace(' ',',').replace('\n','').split(",")]))
dim_6_regular = sum(dim6_regular)
dim40_contigous = np.array([int(x) for x in """0
0
4417755799
0
2587126745
0
1907452696
0
4470211896
1490535856
1488268722
1492406027
0
0
3020774732
1502998186
1495229685
1513519294
2943026264
1500269058
52149259
1496657426
2937658244
1494297672
74030192
1501455826
1495883230
2888428091
1482110806
150844250
2836859365
0
2995913912
0
2986402746
0
1637257243
2848563352
336366333
0
2955346561
1582901866
1494805041
1486848958
1575414883
0
3006540619
1651961484
1495211562
1418961073
2413046393
0
2310831892
1545685148
1494591413
1494394190
1495772460
1497140001
1506995279
1502064624
1503212470
1494161247
1496442614
1494353921
1496910315
1490640984
2120990326
1496964537
1503480472
0
2487680099
1511772203
1990506585
1085
1495006176
2981033808
1497807398
1499905019
0
2992809338
1499678220
1503422919
1432738350
1572013300
1488687745
1491956406
1493056793
1490487590
1494885596
0
2981312602
1495995022
1494473531
91287100
2756251323
230008646
1491190180
1430538651
1583249575
1576908184
1332064486
0
3175178358
0
3080574208
2649916724
1275899646
0
2878955633""".replace(' ',',').replace('\n','').split(",")])
dim_40_contigous = sum(dim40_contigous)
dim40_regular = np.array([int(x) for x in """0
2873345808
0
0
4568364039
0
2962578326
1482026881
1482959662
1283314470
1696338408
1493064117
1496278352
1493690192
1490808431
1482241102
1485073059
1493488736
1493734395
1494051944
1491925626
1490400073
1492918850
0
2989795422
0
2981131303
1490971079
1490457922
0
2972098981
148614774
0
0
4865632909
1257951927
1588161938
1785512571
0
0
0
6770405012
0
2348949300
0
0
4570645310
0
2824636217
1762646952
0
2891891542
1579360612
0
0
4915926092
1496744037
1495870886
1497206544
1287107800
0
3203518565
0
2992849786
0
2975114733
1481666721
686803
2988542677
1496054096
1496578264
1491746836
1492851059
1497143785
1495192565
1500194130
0
1637919784
2854243731
0
2366222238
2128844017
0
1818736204
2676149826
1497854865
962431460
0
3539793686
1024018613
1489816277
0
3158434789
1369912046
1491311601
1575013089
0
0
4446661541
1416884482
1710611180
1757241245
1233889541
1437873524
0
3280246404
0
2995489571
1541069762
3930514690
3863607919
3597368896
3734892809
3944232184
3709601722
3642560210
3625597517
3920215743
3927141046
4000023820
3916655686
3951797607
3952120224
3689180680
3587372425
3615616989
3726388814
3929746604
3929801334
3942019114
3648646073
3944755905
3929255746
3668055728
3672894376
3844013718
3910779810
3940544740
4006830187
3928198267
3531412165
3818215173
3738150454
3604332393
3650311511
3807369864
3826302554
3775371186
3710423568
3934761097
3925844952
3602607914
3741857686
3732482931
3729783593
3635250094
3706648118
3853167484
3916845229
3983629794
3943670196
3867725730
3548946522
3696888472
3729664873
3750609657
3702688827
3994465571
3666995886
3881628312
3720334387
3933568515
3693178186
3888954122
3961471376
3937112215
3936691623
3915403037
3693727226
3917170944
3930555435
3933184551
3931662845
3940295083
1033014609""".replace(' ',',').replace('\n','').split(",")])
dim_40_regular = sum(dim40_regular)
# dimentions = ("3-dementions", "6-dementions", "40-dementions")
# comparitors = {
# 'FAT-Pointer based range address': (dmin_3_physically_contigous, dim_6_contigous, dim_40_contigous),
# 'System Allocator': (dmin_3_regular, dim_6_regular, dim_40_regular),
# }
# x = np.arange(len(dimentions)) # the label locations
# width = 0.25 # the width of the bars
# multiplier = 0
# fig, ax = plt.subplots(layout='constrained')
# for attribute, measurement in comparitors.items():
# offset = width * multiplier
# rects = ax.bar(x + offset, measurement, width, label=attribute)
# ax.bar_label(rects, padding=3)
# multiplier += 1
# # Add some text for labels, title and custom x-axis tick labels, etc.
# ax.set_ylabel('DTLB L1 reads')
# ax.set_title('L1D_TLB')
# ax.set_xticks(x + width, dimentions)
# ax.legend(loc='upper left', ncols=2)
# ax.set_ylim(0, 250)
# plt.show()
# Sample data
categories = ['Size 200', 'Size 10000']
group_1 = [dmin_3_physically_contigous, dim_6_contigous]
group_2 = [dmin_3_regular, dim_6_regular]
# Number of categories
n = len(categories)
# Create a bar width
bar_width = 0.25
# Create an array with the positions of the bars on the x-axis
r1 = np.arange(n)
r2 = [x + bar_width for x in r1]
# r3 = [x + bar_width for x in r2]
# Create the grouped bar graph
plt.bar(r1, group_1, color='b', width=bar_width, edgecolor='grey', label='FAT-Pointer based range based addresses')
plt.bar(r2, group_2, color='g', width=bar_width, edgecolor='grey', label='System memory allocator')
# plt.bar(r3, group_3, color='r', width=bar_width, edgecolor='grey', label='Group 3')
# Add xticks on the middle of the grouped bars
plt.xlabel('Size of Matrix COZ MatrixMultiply', fontweight='bold')
plt.xticks([r + bar_width for r in range(n)], categories)
# Add labels and title
plt.ylabel('DTLB L1 reads', fontweight='bold')
plt.title('Sum of DTLB L1 reads')
# Add a legend
plt.legend()
# Show the plot
# plt.show()
plt.savefig('l1-miss-matrixmultiply.png')

View File

@@ -0,0 +1,395 @@
import matplotlib.pyplot as plt
import numpy as np
dim3_physically_contigous = np.array([399876441])
dmin_3_physically_contigous = sum(dim3_physically_contigous)
dim3_regular = np.array([380500198])
dmin_3_regular = sum(dim3_regular)
dim6_contigous = np.array([int(x) for x in """0
0
3182857841
2676269451
3632836262
5086931936
4921595689
2380537223""".replace(' ',',').replace('\n','').split(",")])
dim_6_contigous = sum(dim6_contigous)
dim6_regular = np.array([int(x) for x in """0
1662045387
2568704269
0
5404906944
4946152426
5097512016
1934071481""".replace(' ',',').replace('\n','').split(",")])
dim_6_regular = sum(dim6_regular)
dim40_contigous = np.array([int(x) for x in """0
0
4417755799
0
2587126745
0
1907452696
0
4470211896
1490535856
1488268722
1492406027
0
0
3020774732
1502998186
1495229685
1513519294
2943026264
1500269058
52149259
1496657426
2937658244
1494297672
74030192
1501455826
1495883230
2888428091
1482110806
150844250
2836859365
0
2995913912
0
2986402746
0
1637257243
2848563352
336366333
0
2955346561
1582901866
1494805041
1486848958
1575414883
0
3006540619
1651961484
1495211562
1418961073
2413046393
0
2310831892
1545685148
1494591413
1494394190
1495772460
1497140001
1506995279
1502064624
1503212470
1494161247
1496442614
1494353921
1496910315
1490640984
2120990326
1496964537
1503480472
0
2487680099
1511772203
1990506585
1085
1495006176
2981033808
1497807398
1499905019
0
2992809338
1499678220
1503422919
1432738350
1572013300
1488687745
1491956406
1493056793
1490487590
1494885596
0
2981312602
1495995022
1494473531
91287100
2756251323
230008646
1491190180
1430538651
1583249575
1576908184
1332064486
0
3175178358
0
3080574208
2649916724
1275899646
0
2878955633""".replace(' ',',').replace('\n','').split(",")])
dim_40_contigous = sum(dim40_contigous)
dim40_regular = np.array([int(x) for x in """0
2873345808
0
0
4568364039
0
2962578326
1482026881
1482959662
1283314470
1696338408
1493064117
1496278352
1493690192
1490808431
1482241102
1485073059
1493488736
1493734395
1494051944
1491925626
1490400073
1492918850
0
2989795422
0
2981131303
1490971079
1490457922
0
2972098981
148614774
0
0
4865632909
1257951927
1588161938
1785512571
0
0
0
6770405012
0
2348949300
0
0
4570645310
0
2824636217
1762646952
0
2891891542
1579360612
0
0
4915926092
1496744037
1495870886
1497206544
1287107800
0
3203518565
0
2992849786
0
2975114733
1481666721
686803
2988542677
1496054096
1496578264
1491746836
1492851059
1497143785
1495192565
1500194130
0
1637919784
2854243731
0
2366222238
2128844017
0
1818736204
2676149826
1497854865
962431460
0
3539793686
1024018613
1489816277
0
3158434789
1369912046
1491311601
1575013089
0
0
4446661541
1416884482
1710611180
1757241245
1233889541
1437873524
0
3280246404
0
2995489571
1541069762
3930514690
3863607919
3597368896
3734892809
3944232184
3709601722
3642560210
3625597517
3920215743
3927141046
4000023820
3916655686
3951797607
3952120224
3689180680
3587372425
3615616989
3726388814
3929746604
3929801334
3942019114
3648646073
3944755905
3929255746
3668055728
3672894376
3844013718
3910779810
3940544740
4006830187
3928198267
3531412165
3818215173
3738150454
3604332393
3650311511
3807369864
3826302554
3775371186
3710423568
3934761097
3925844952
3602607914
3741857686
3732482931
3729783593
3635250094
3706648118
3853167484
3916845229
3983629794
3943670196
3867725730
3548946522
3696888472
3729664873
3750609657
3702688827
3994465571
3666995886
3881628312
3720334387
3933568515
3693178186
3888954122
3961471376
3937112215
3936691623
3915403037
3693727226
3917170944
3930555435
3933184551
3931662845
3940295083
1033014609""".replace(' ',',').replace('\n','').split(",")])
dim_40_regular = sum(dim40_regular)
# dimentions = ("3-dementions", "6-dementions", "40-dementions")
# comparitors = {
# 'FAT-Pointer based range address': (dmin_3_physically_contigous, dim_6_contigous, dim_40_contigous),
# 'System Allocator': (dmin_3_regular, dim_6_regular, dim_40_regular),
# }
# x = np.arange(len(dimentions)) # the label locations
# width = 0.25 # the width of the bars
# multiplier = 0
# fig, ax = plt.subplots(layout='constrained')
# for attribute, measurement in comparitors.items():
# offset = width * multiplier
# rects = ax.bar(x + offset, measurement, width, label=attribute)
# ax.bar_label(rects, padding=3)
# multiplier += 1
# # Add some text for labels, title and custom x-axis tick labels, etc.
# ax.set_ylabel('DTLB L1 reads')
# ax.set_title('L1D_TLB')
# ax.set_xticks(x + width, dimentions)
# ax.legend(loc='upper left', ncols=2)
# ax.set_ylim(0, 250)
# plt.show()
# Sample data
categories = ['Size 200', 'Size 10000']
group_1 = [dmin_3_physically_contigous, dim_6_contigous]
group_2 = [dmin_3_regular, dim_6_regular]
# Number of categories
n = len(categories)
# Create a bar width
bar_width = 0.25
# Create an array with the positions of the bars on the x-axis
r1 = np.arange(n)
r2 = [x + bar_width for x in r1]
# r3 = [x + bar_width for x in r2]
# Create the grouped bar graph
plt.bar(r1, group_1, color='b', width=bar_width, edgecolor='grey', label='FAT-Pointer based range based addresses')
plt.bar(r2, group_2, color='g', width=bar_width, edgecolor='grey', label='System memory allocator')
# plt.bar(r3, group_3, color='r', width=bar_width, edgecolor='grey', label='Group 3')
# Add xticks on the middle of the grouped bars
plt.xlabel('Size of Matrix COZ MatrixMultiply', fontweight='bold')
plt.xticks([r + bar_width for r in range(n)], categories)
# Add labels and title
plt.ylabel('DTLB L1 reads', fontweight='bold')
plt.title('Sum of DTLB L1 reads')
# Add a legend
plt.legend()
# Show the plot
# plt.show()
plt.savefig('l1-tlb-matrixmultiply.png')

View File

@@ -0,0 +1,397 @@
import matplotlib.pyplot as plt
import numpy as np
dim3_physically_contigous = np.array([3013349])
dmin_3_physically_contigous = sum(dim3_physically_contigous)
dim3_regular = np.array([2946541])
dmin_3_regular = sum(dim3_regular)
dim6_contigous = np.array([int(x) for x in """13933616
0
0
55855105
177840659
380285140
292719568
163746827""".replace(' ',',').replace('\n','').split(",")])
dim_6_contigous = sum(dim6_contigous)
dim6_regular = np.array([int(x) for x in """0
0
48672313
0
243876172
332240431
283300132
151566198""".replace(' ',',').replace('\n','').split(",")])
dim_6_regular = sum(dim6_regular)
dim40_contigous = np.array([int(x) for x in """11074868
17796846
0
42335753
0
42578037
17369088
0
0
0
82499577
0
0
0
82928349
19903822
20322217
20196113
20575304
19769508
0
40692176
0
35388820
0
46400139
20884653
0
40479127
20938047
20020758
21590212
17955844
1895112
24707564
35247723
19854386
19994871
0
0
39958384
39619071
19821971
0
39754647
0
39873651
19521815
20209862
20048932
20231348
20214678
0
41152963
6223168
16818928
37877594
12893970
21214360
0
32124557
20718531
0
40755758
35334046
0
41293476
20648826
0
20479335
21645864
28444433
0
34204784
36576871
0
0
0
76911224
20723929
14021956
28392009
26330087
0
41195363
9598203
28499682
0
34404203
29859217
20539283
20714755
20553408
0
20889230
32963850
7949901
33765967
28019151
13062566
26967792
0
0
62723768
15030483
20277594
0
41160435
0
28127160""".replace(' ',',').replace('\n','').split(",")])
dim_40_contigous = sum(dim40_contigous)
dim40_regular = np.array([int(x) for x in """2780755
0
40650126
19973207
0
45212438
14734736
21512132
20024828
17015744
21012868
24889120
0
40307315
0
40543393
19447333
16752820
19802564
19717349
0
32838418
7313016
36828882
19421922
19221311
0
36504040
19221867
19111822
4391041
32495439
0
23803968
18651515
0
37881224
29985095
4176134
18086488
18628963
0
0
0
77724802
0
37326656
0
37118376
17904262
19199261
19094655
18162476
19841950
18493958
19955899
15237090
18940111
16777309
19653443
17616389
1417668
23258907
0
38260471
22608071
13460148
19002183
16931931
19518969
12655691
18821392
19084457
0
38850414
0
38673127
0
39578640
0
38841607
0
38340444
0
0
59014396
13524903
19465492
25539086
19637297
5254742
27355676
24079140
19975996
19226004
16878651
0
0
62518013
0
0
39742980
39872968
18698312
19633681
19652322
18131608
0
39807132
84805863
205063169
207026601
207353408
208077229
223112773
232033395
210898122
200158459
203639347
230644673
237184228
230544844
234041882
225077016
208652470
206710725
207688549
208921986
237145817
216789158
207248978
206867158
226521698
243656316
236022389
216914633
207163647
207752568
206269550
207422862
240605554
220855937
207420877
210468984
207768015
250267485
212878277
209949240
207411078
208223656
210492256
224683374
242489309
231567088
211246599
208066123
208277683
228206741
229508886
207161109
206156489
214927236
231786596
232625811
230571602
209516467
219150327
228216552
206438524
220447351
219045513
230838177
216657765
205321014
207699856
219191131
236171517
225724156
208329895
210489462
211206457
207353005
230789569
231425162
43897100""".replace(' ',',').replace('\n','').split(",")])
dim_40_regular = sum(dim40_regular)
# dimentions = ("3-dementions", "6-dementions", "40-dementions")
# comparitors = {
# 'FAT-Pointer based range address': (dmin_3_physically_contigous, dim_6_contigous, dim_40_contigous),
# 'System Allocator': (dmin_3_regular, dim_6_regular, dim_40_regular),
# }
# x = np.arange(len(dimentions)) # the label locations
# width = 0.25 # the width of the bars
# multiplier = 0
# fig, ax = plt.subplots(layout='constrained')
# for attribute, measurement in comparitors.items():
# offset = width * multiplier
# rects = ax.bar(x + offset, measurement, width, label=attribute)
# ax.bar_label(rects, padding=3)
# multiplier += 1
# # Add some text for labels, title and custom x-axis tick labels, etc.
# ax.set_ylabel('DTLB L1 reads')
# ax.set_title('L1D_TLB')
# ax.set_xticks(x + width, dimentions)
# ax.legend(loc='upper left', ncols=2)
# ax.set_ylim(0, 250)
# plt.show()
# Sample data
categories = ['Size 200', 'Size 10000']
group_1 = [dmin_3_physically_contigous, dim_6_contigous]
group_2 = [dmin_3_regular, dim_6_regular]
# Number of categories
n = len(categories)
# Create a bar width
bar_width = 0.25
# Create an array with the positions of the bars on the x-axis
r1 = np.arange(n)
r2 = [x + bar_width for x in r1]
# r3 = [x + bar_width for x in r2]
# Create the grouped bar graph
plt.bar(r1, group_1, color='b', width=bar_width, edgecolor='grey', label='FAT-Pointer based range based addresses')
plt.bar(r2, group_2, color='g', width=bar_width, edgecolor='grey', label='System memory allocator')
# plt.bar(r3, group_3, color='r', width=bar_width, edgecolor='grey', label='Group 3')
# Add xticks on the middle of the grouped bars
plt.xlabel('Size of Matrix COZ MatrixMultiply', fontweight='bold')
plt.xticks([r + bar_width for r in range(n)], categories)
# Add labels and title
plt.ylabel('DTLB L2 reads', fontweight='bold')
plt.title('Sum of DTLB L2 reads')
# Add a legend
plt.legend()
# Show the plot
# plt.show()
plt.savefig('l2-tlb-matrixmultiply.png')

View File

@@ -0,0 +1,393 @@
import matplotlib.pyplot as plt
import numpy as np
dim3_physically_contigous = np.array([26005])
dmin_3_physically_contigous = sum(dim3_physically_contigous)
dim3_regular = np.array([8517])
dmin_3_regular = sum(dim3_regular)
dim6_contigous = np.array([int(x) for x in """0
2722
0
0
84290
185921
251521
356452""".replace(' ',',').replace('\n','').split(",")])
dim_6_contigous = sum(dim6_contigous)
dim6_regular = np.array([int(x) for x in """1310
1658
86
0
73097
171472
237158
161478""".replace(' ',',').replace('\n','').split(",")])
dim_6_regular = sum(dim6_regular)
dim40_contigous = np.array([int(x) for x in """1072
1566
0
775
564
574
0
0
1842
397
832
0
1143
0
1132
0
0
3478
335
1181
0
1451
0
0
2456
914
566
208
1323
0
1341
0
0
0
2800
869
0
1624
852
0
1444
78
0
1165
242
0
0
1436
0
1656
1295
672
0
2124
689
247
974
196
664
0
0
0
0
4571
411
1108
0
1855
779
0
0
0
3973
0
0
2126
1090
0
0
0
1455
868
834
0
2252
743
0
0
0
2867
576
528
737
1279
753
0
0
1469
709
0
951
1011
560
923
389
681
887
0
1091""".replace(' ',',').replace('\n','').split(",")])
dim_40_contigous = sum(dim40_contigous)
dim40_regular = np.array([int(x) for x in """2077
0
2240
974
400
0
1949
0
353
0
2136
1292
1150
1080
1029
188
16
9401
3108
0
12850
2972
0
13417
4015
0
0
12179
708
4434
5515
105
0
0
52251
8327
0
0
0
7628
0
6047
2796
2505
1949
0
0
0
0
14301
0
5171
3362
3772
0
27162
15974
5284
11553
0
24698
0
14835
0
21593
664
5721
12945
0
13774
0
14722
1465
0
27655
2332
4084
33813
0
9790
0
50326
5316
0
0
0
0
9760
4192
1552
7241
3216
4888
8667
28500
8415
4523
1449
4652
0
7625
0
13775
0
1897
4708
0
12225
4482759
191711306
235410343
173845602
242711520
154862574
192747518
220667551
275602208
184073481
198073425
227482742
187971748
183689903
187178845
186674013
207741123
233177167
221368814
183933427
181175654
204068107
228269124
205027178
209400883
187840131
188293011
221927087
232807458
188215220
180783387
216711542
208036998
214266374
186414281
180185139
185237868
191677316
183911791
183803795
224964126
204113022
210381502
225585571
220223453
226214857
204932647
184864540
217262003
202414930
191948700
193400960
181280109
180323521
184002482
207777617
222940234
189422853
186182495
186261124
181304280
192816961
184117119
209667712
240951856
219942014
221564848
192926621
181690338
183097907
186087016
184347616
180377265
199009737
182802614""".replace(' ',',').replace('\n','').split(",")])
dim_40_regular = sum(dim40_regular)
# dimentions = ("3-dementions", "6-dementions", "40-dementions")
# comparitors = {
# 'FAT-Pointer based range address': (dmin_3_physically_contigous, dim_6_contigous, dim_40_contigous),
# 'System Allocator': (dmin_3_regular, dim_6_regular, dim_40_regular),
# }
# x = np.arange(len(dimentions)) # the label locations
# width = 0.25 # the width of the bars
# multiplier = 0
# fig, ax = plt.subplots(layout='constrained')
# for attribute, measurement in comparitors.items():
# offset = width * multiplier
# rects = ax.bar(x + offset, measurement, width, label=attribute)
# ax.bar_label(rects, padding=3)
# multiplier += 1
# # Add some text for labels, title and custom x-axis tick labels, etc.
# ax.set_ylabel('DTLB L1 reads')
# ax.set_title('L1D_TLB')
# ax.set_xticks(x + width, dimentions)
# ax.legend(loc='upper left', ncols=2)
# ax.set_ylim(0, 250)
# plt.show()
# Sample data
categories = ['Size 200', 'Size 10000']
group_1 = [dmin_3_physically_contigous, dim_6_contigous]
group_2 = [dmin_3_regular, dim_6_regular]
# Number of categories
n = len(categories)
# Create a bar width
bar_width = 0.25
# Create an array with the positions of the bars on the x-axis
r1 = np.arange(n)
r2 = [x + bar_width for x in r1]
# r3 = [x + bar_width for x in r2]
# Create the grouped bar graph
plt.bar(r1, group_1, color='b', width=bar_width, edgecolor='grey', label='FAT-Pointer based range based addresses')
plt.bar(r2, group_2, color='g', width=bar_width, edgecolor='grey', label='System memory allocator')
# plt.bar(r3, group_3, color='r', width=bar_width, edgecolor='grey', label='Group 3')
# Add xticks on the middle of the grouped bars
plt.xlabel('Size of Matrix COZ MatrixMultiply', fontweight='bold')
plt.xticks([r + bar_width for r in range(n)], categories)
# Add labels and title
plt.ylabel('DTLB L2 reads', fontweight='bold')
plt.title('Sum of DTLB L2 reads')
# Add a legend
plt.legend()
# Show the plot
# plt.show()
plt.savefig('tlb-walk-matrixmultiply.png')