-
Notifications
You must be signed in to change notification settings - Fork 79
/
Copy path4rWn.txt
801 lines (677 loc) · 39.8 KB
/
4rWn.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
sbc-bench v0.9.39 Hardkernel ODROID-N2Plus (Mon, 27 Mar 2023 08:49:31 +0200)
Distributor ID: Debian
Description: Debian GNU/Linux 11 (bullseye)
Release: 11
Codename: bullseye
Build system: https://github.com/armbian/build, 23.02.2, Odroid N2, meson-g12b, meson64
/usr/bin/gcc (Debian 10.2.1-6) 10.2.1 20210110
Uptime: 08:49:31 up 1:00, 1 user, load average: 0.31, 0.36, 0.19, 30.8°C, 96143579
Linux 6.1.11-meson64 (odroidn2) 03/27/23 _aarch64_ (6 CPU)
avg-cpu: %user %nice %system %iowait %steal %idle
0.76 0.00 0.29 0.03 0.00 98.92
Device tps kB_read/s kB_wrtn/s kB_dscd/s kB_read kB_wrtn kB_dscd
mmcblk1 5.95 140.91 148.26 0.00 512880 539641 0
zram0 0.16 0.65 0.00 0.00 2384 4 0
zram1 0.13 0.13 0.64 0.00 484 2324 0
total used free shared buff/cache available
Mem: 3.7Gi 183Mi 3.4Gi 5.0Mi 126Mi 3.4Gi
Swap: 1.8Gi 0B 1.8Gi
Filename Type Size Used Priority
/dev/zram0 partition 1938044 0 5
##########################################################################
Checking cpufreq OPP for cpu0-cpu1 (Cortex-A53):
Cpufreq OPP: 2016 Measured: 2013 (2013.409/2013.360/2013.164)
Cpufreq OPP: 1908 Measured: 1905 (1905.828/1905.388/1905.344)
Cpufreq OPP: 1800 Measured: 1797 (1797.629/1797.629/1797.629)
Cpufreq OPP: 1704 Measured: 1701 (1701.686/1701.651/1701.546)
Cpufreq OPP: 1608 Measured: 1605 (1605.800/1605.761/1605.644)
Cpufreq OPP: 1512 Measured: 1497 (1497.859/1497.486/1497.452)
Cpufreq OPP: 1398 Measured: 1395 (1395.900/1395.782/1395.723)
Cpufreq OPP: 1200 Measured: 1197 (1197.826/1197.772/1197.772)
Cpufreq OPP: 1000 Measured: 997 (997.878/997.760/997.689)
Checking cpufreq OPP for cpu2-cpu5 (Cortex-A73):
Cpufreq OPP: 2400 Measured: 2398 (2398.769/2398.714/2398.714)
Cpufreq OPP: 2304 Measured: 2302 (2302.735/2302.684/2302.684)
Cpufreq OPP: 2208 Measured: 2206 (2206.858/2206.858/2206.763)
Cpufreq OPP: 2108 Measured: 2098 (2098.844/2098.801/2098.759)
Cpufreq OPP: 2016 Measured: 2014 (2014.735/2014.735/2014.735)
Cpufreq OPP: 1908 Measured: 1906 (1906.883/1906.839/1906.839)
Cpufreq OPP: 1800 Measured: 1798 (1798.920/1798.803/1798.764)
Cpufreq OPP: 1704 Measured: 1702 (1702.879/1702.843/1702.808)
Cpufreq OPP: 1608 Measured: 1606 (1606.854/1606.815/1606.776)
Cpufreq OPP: 1512 Measured: 1498 (1498.912/1498.810/1498.810)
Cpufreq OPP: 1398 Measured: 1396 (1396.962/1396.962/1396.873)
Cpufreq OPP: 1200 Measured: 1198 (1198.966/1198.939/1198.939)
Cpufreq OPP: 1000 Measured: 998 (998.938/998.914/998.890)
##########################################################################
Hardware sensors:
gpio_fan-isa-0000
fan1: 0 RPM (min = 0 RPM, max = 1600 RPM)
cpu_thermal-virtual-0
temp1: +28.4 C (crit = +110.0 C)
ddr_thermal-virtual-0
temp1: +30.7 C (crit = +110.0 C)
##########################################################################
Executing benchmark on cpu0 (Cortex-A53):
tinymembench v0.4.9-nuumio (simple benchmark for memory throughput and latency)
CFLAGS:
bandwidth test min repeats (-b): 2
bandwidth test max repeats (-B): 3
bandwidth test mem realloc (-M): no (-m for realloc)
latency test repeats (-l): 3
latency test count (-c): 1000000
==========================================================================
== Memory bandwidth tests ==
== ==
== Note 1: 1MB = 1000000 bytes ==
== Note 2: Test result is the best of repeated runs. Number of repeats ==
== is shown in brackets ==
== Note 3: Results for 'copy' tests show how many bytes can be ==
== copied per second (adding together read and writen ==
== bytes would have provided twice higher numbers) ==
== Note 4: 2-pass copy means that we are using a small temporary buffer ==
== to first fetch data into it, and only then write it to the ==
== destination (source -> L1 cache, L1 cache -> destination) ==
== Note 5: If sample standard deviation exceeds 0.1%, it is shown in ==
== brackets ==
==========================================================================
C copy backwards : 2051.9 MB/s (3, 2.7%)
C copy backwards (32 byte blocks) : 2109.6 MB/s (3, 1.6%)
C copy backwards (64 byte blocks) : 2061.9 MB/s (3, 1.4%)
C copy : 2189.3 MB/s (3, 0.5%)
C copy prefetched (32 bytes step) : 1623.7 MB/s (3, 0.2%)
C copy prefetched (64 bytes step) : 1919.7 MB/s (3, 0.2%)
C 2-pass copy : 1888.4 MB/s (2)
C 2-pass copy prefetched (32 bytes step) : 1355.7 MB/s (2)
C 2-pass copy prefetched (64 bytes step) : 1201.6 MB/s (2)
C scan 8 : 390.8 MB/s (3)
C scan 16 : 773.1 MB/s (3, 0.3%)
C scan 32 : 1426.5 MB/s (2)
C scan 64 : 2539.6 MB/s (3, 0.7%)
C fill : 7714.9 MB/s (2)
C fill (shuffle within 16 byte blocks) : 7715.6 MB/s (2)
C fill (shuffle within 32 byte blocks) : 7716.4 MB/s (2)
C fill (shuffle within 64 byte blocks) : 7716.2 MB/s (2)
---
libc memcpy copy : 2213.1 MB/s (3, 0.5%)
libc memchr scan : 2567.7 MB/s (3, 1.8%)
libc memset fill : 7721.9 MB/s (2)
---
NEON LDP/STP copy : 2246.1 MB/s (2)
NEON LDP/STP copy pldl2strm (32 bytes step) : 1410.6 MB/s (3, 0.8%)
NEON LDP/STP copy pldl2strm (64 bytes step) : 1860.5 MB/s (3, 0.1%)
NEON LDP/STP copy pldl1keep (32 bytes step) : 2468.2 MB/s (2)
NEON LDP/STP copy pldl1keep (64 bytes step) : 2464.5 MB/s (2)
NEON LD1/ST1 copy : 2205.0 MB/s (3, 0.7%)
NEON LDP load : 3504.0 MB/s (2)
NEON LDNP load : 2553.0 MB/s (3, 0.8%)
NEON STP fill : 7722.6 MB/s (2)
NEON STNP fill : 6327.2 MB/s (2)
ARM LDP/STP copy : 2241.7 MB/s (2)
ARM LDP load : 3509.6 MB/s (2)
ARM LDNP load : 2538.2 MB/s (3, 0.7%)
ARM STP fill : 7720.1 MB/s (2)
ARM STNP fill : 6327.2 MB/s (2)
==========================================================================
== Framebuffer read tests. ==
== ==
== Many ARM devices use a part of the system memory as the framebuffer, ==
== typically mapped as uncached but with write-combining enabled. ==
== Writes to such framebuffers are quite fast, but reads are much ==
== slower and very sensitive to the alignment and the selection of ==
== CPU instructions which are used for accessing memory. ==
== ==
== Many x86 systems allocate the framebuffer in the GPU memory, ==
== accessible for the CPU via a relatively slow PCI-E bus. Moreover, ==
== PCI-E is asymmetric and handles reads a lot worse than writes. ==
== ==
== If uncached framebuffer reads are reasonably fast (at least 100 MB/s ==
== or preferably >300 MB/s), then using the shadow framebuffer layer ==
== is not necessary in Xorg DDX drivers, resulting in a nice overall ==
== performance improvement. For example, the xf86-video-fbturbo DDX ==
== uses this trick. ==
==========================================================================
NEON LDP/STP copy (from framebuffer) : 233.9 MB/s (2)
NEON LDP/STP 2-pass copy (from framebuffer) : 233.0 MB/s (2)
NEON LD1/ST1 copy (from framebuffer) : 64.6 MB/s (2)
NEON LD1/ST1 2-pass copy (from framebuffer) : 64.6 MB/s (2)
ARM LDP/STP copy (from framebuffer) : 123.7 MB/s (2)
ARM LDP/STP 2-pass copy (from framebuffer) : 124.5 MB/s (2)
==========================================================================
== Memory latency test ==
== ==
== Average time is measured for random memory accesses in the buffers ==
== of different sizes. The larger is the buffer, the more significant ==
== are relative contributions of TLB, L1/L2 cache misses and SDRAM ==
== accesses. For extremely large buffer sizes we are expecting to see ==
== page table walk with several requests to SDRAM for almost every ==
== memory access (though 64MiB is not nearly large enough to experience ==
== this effect to its fullest). ==
== ==
== Note 1: All the numbers are representing extra time, which needs to ==
== be added to L1 cache latency. The cycle timings for L1 cache ==
== latency can be usually found in the processor documentation. ==
== Note 2: Dual random read means that we are simultaneously performing ==
== two independent memory accesses at a time. In the case if ==
== the memory subsystem can't handle multiple outstanding ==
== requests, dual random read has the same timings as two ==
== single reads performed one after another. ==
==========================================================================
block size : single random read / dual random read, [MADV_NOHUGEPAGE]
1024 : 0.0 ns / 0.0 ns
2048 : 0.0 ns / 0.0 ns
4096 : 0.0 ns / 0.0 ns
8192 : 0.0 ns / 0.0 ns
16384 : 0.0 ns / 0.0 ns
32768 : 0.1 ns / 0.0 ns
65536 : 3.5 ns / 5.8 ns
131072 : 5.5 ns / 8.0 ns
262144 : 6.9 ns / 9.4 ns
524288 : 60.0 ns / 93.8 ns
1048576 : 92.2 ns / 125.2 ns
2097152 : 108.8 ns / 136.3 ns
4194304 : 121.5 ns / 144.5 ns
8388608 : 128.2 ns / 148.7 ns
16777216 : 132.2 ns / 151.6 ns
33554432 : 135.0 ns / 153.8 ns
67108864 : 148.5 ns / 179.4 ns
block size : single random read / dual random read, [MADV_HUGEPAGE]
1024 : 0.0 ns / 0.0 ns
2048 : 0.0 ns / 0.0 ns
4096 : 0.0 ns / 0.0 ns
8192 : 0.0 ns / 0.0 ns
16384 : 0.0 ns / 0.0 ns
32768 : 0.1 ns / 0.0 ns
65536 : 3.5 ns / 5.8 ns
131072 : 5.5 ns / 8.0 ns
262144 : 7.0 ns / 9.4 ns
524288 : 60.8 ns / 96.2 ns
1048576 : 93.5 ns / 130.0 ns
2097152 : 108.2 ns / 135.5 ns
4194304 : 116.4 ns / 139.3 ns
8388608 : 120.5 ns / 140.9 ns
16777216 : 122.5 ns / 141.4 ns
33554432 : 123.5 ns / 141.8 ns
67108864 : 124.0 ns / 141.9 ns
Executing benchmark on cpu2 (Cortex-A73):
tinymembench v0.4.9-nuumio (simple benchmark for memory throughput and latency)
CFLAGS:
bandwidth test min repeats (-b): 2
bandwidth test max repeats (-B): 3
bandwidth test mem realloc (-M): no (-m for realloc)
latency test repeats (-l): 3
latency test count (-c): 1000000
==========================================================================
== Memory bandwidth tests ==
== ==
== Note 1: 1MB = 1000000 bytes ==
== Note 2: Test result is the best of repeated runs. Number of repeats ==
== is shown in brackets ==
== Note 3: Results for 'copy' tests show how many bytes can be ==
== copied per second (adding together read and writen ==
== bytes would have provided twice higher numbers) ==
== Note 4: 2-pass copy means that we are using a small temporary buffer ==
== to first fetch data into it, and only then write it to the ==
== destination (source -> L1 cache, L1 cache -> destination) ==
== Note 5: If sample standard deviation exceeds 0.1%, it is shown in ==
== brackets ==
==========================================================================
C copy backwards : 4118.1 MB/s (2)
C copy backwards (32 byte blocks) : 4117.5 MB/s (2)
C copy backwards (64 byte blocks) : 4118.2 MB/s (2)
C copy : 4202.0 MB/s (2)
C copy prefetched (32 bytes step) : 4175.3 MB/s (2)
C copy prefetched (64 bytes step) : 4202.7 MB/s (2)
C 2-pass copy : 3000.3 MB/s (3, 0.1%)
C 2-pass copy prefetched (32 bytes step) : 2917.3 MB/s (2)
C 2-pass copy prefetched (64 bytes step) : 2931.7 MB/s (3, 0.1%)
C scan 8 : 798.0 MB/s (2)
C scan 16 : 1596.0 MB/s (2)
C scan 32 : 3158.0 MB/s (2)
C scan 64 : 6057.9 MB/s (2)
C fill : 7510.0 MB/s (2)
C fill (shuffle within 16 byte blocks) : 7512.5 MB/s (2)
C fill (shuffle within 32 byte blocks) : 7502.8 MB/s (2)
C fill (shuffle within 64 byte blocks) : 7512.7 MB/s (2)
---
libc memcpy copy : 4216.4 MB/s (3)
libc memchr scan : 9367.6 MB/s (2)
libc memset fill : 7509.3 MB/s (2)
---
NEON LDP/STP copy : 4220.5 MB/s (2)
NEON LDP/STP copy pldl2strm (32 bytes step) : 4221.6 MB/s (3, 0.1%)
NEON LDP/STP copy pldl2strm (64 bytes step) : 4220.5 MB/s (2)
NEON LDP/STP copy pldl1keep (32 bytes step) : 4009.6 MB/s (2)
NEON LDP/STP copy pldl1keep (64 bytes step) : 4164.4 MB/s (3, 0.1%)
NEON LD1/ST1 copy : 4220.2 MB/s (3, 1.8%)
NEON LDP load : 9397.4 MB/s (2)
NEON LDNP load : 9398.4 MB/s (3, 0.2%)
NEON STP fill : 7510.1 MB/s (2)
NEON STNP fill : 7508.9 MB/s (2)
ARM LDP/STP copy : 4219.6 MB/s (2)
ARM LDP load : 9284.7 MB/s (3, 0.5%)
ARM LDNP load : 9305.6 MB/s (2)
ARM STP fill : 7484.5 MB/s (2)
ARM STNP fill : 7505.1 MB/s (3, 0.5%)
==========================================================================
== Framebuffer read tests. ==
== ==
== Many ARM devices use a part of the system memory as the framebuffer, ==
== typically mapped as uncached but with write-combining enabled. ==
== Writes to such framebuffers are quite fast, but reads are much ==
== slower and very sensitive to the alignment and the selection of ==
== CPU instructions which are used for accessing memory. ==
== ==
== Many x86 systems allocate the framebuffer in the GPU memory, ==
== accessible for the CPU via a relatively slow PCI-E bus. Moreover, ==
== PCI-E is asymmetric and handles reads a lot worse than writes. ==
== ==
== If uncached framebuffer reads are reasonably fast (at least 100 MB/s ==
== or preferably >300 MB/s), then using the shadow framebuffer layer ==
== is not necessary in Xorg DDX drivers, resulting in a nice overall ==
== performance improvement. For example, the xf86-video-fbturbo DDX ==
== uses this trick. ==
==========================================================================
NEON LDP/STP copy (from framebuffer) : 401.4 MB/s (3, 0.1%)
NEON LDP/STP 2-pass copy (from framebuffer) : 388.6 MB/s (3, 0.3%)
NEON LD1/ST1 copy (from framebuffer) : 400.6 MB/s (2)
NEON LD1/ST1 2-pass copy (from framebuffer) : 388.4 MB/s (2)
ARM LDP/STP copy (from framebuffer) : 404.8 MB/s (2)
ARM LDP/STP 2-pass copy (from framebuffer) : 380.5 MB/s (2)
==========================================================================
== Memory latency test ==
== ==
== Average time is measured for random memory accesses in the buffers ==
== of different sizes. The larger is the buffer, the more significant ==
== are relative contributions of TLB, L1/L2 cache misses and SDRAM ==
== accesses. For extremely large buffer sizes we are expecting to see ==
== page table walk with several requests to SDRAM for almost every ==
== memory access (though 64MiB is not nearly large enough to experience ==
== this effect to its fullest). ==
== ==
== Note 1: All the numbers are representing extra time, which needs to ==
== be added to L1 cache latency. The cycle timings for L1 cache ==
== latency can be usually found in the processor documentation. ==
== Note 2: Dual random read means that we are simultaneously performing ==
== two independent memory accesses at a time. In the case if ==
== the memory subsystem can't handle multiple outstanding ==
== requests, dual random read has the same timings as two ==
== single reads performed one after another. ==
==========================================================================
block size : single random read / dual random read, [MADV_NOHUGEPAGE]
1024 : 0.0 ns / 0.0 ns
2048 : 0.0 ns / 0.0 ns
4096 : 0.0 ns / 0.0 ns
8192 : 0.0 ns / 0.0 ns
16384 : 0.0 ns / 0.0 ns
32768 : 0.0 ns / 0.0 ns
65536 : 4.7 ns / 7.7 ns
131072 : 7.0 ns / 10.8 ns
262144 : 8.8 ns / 12.1 ns
524288 : 10.5 ns / 12.6 ns
1048576 : 19.7 ns / 27.8 ns
2097152 : 71.9 ns / 106.9 ns
4194304 : 101.5 ns / 135.1 ns
8388608 : 119.4 ns / 150.3 ns
16777216 : 128.8 ns / 156.4 ns
33554432 : 134.6 ns / 159.1 ns
67108864 : 137.6 ns / 161.5 ns
block size : single random read / dual random read, [MADV_HUGEPAGE]
1024 : 0.0 ns / 0.0 ns
2048 : 0.0 ns / 0.0 ns
4096 : 0.0 ns / 0.0 ns
8192 : 0.0 ns / 0.0 ns
16384 : 0.0 ns / 0.0 ns
32768 : 0.0 ns / 0.0 ns
65536 : 4.7 ns / 7.7 ns
131072 : 7.0 ns / 10.8 ns
262144 : 8.3 ns / 12.1 ns
524288 : 9.2 ns / 12.7 ns
1048576 : 10.3 ns / 13.6 ns
2097152 : 70.1 ns / 104.9 ns
4194304 : 99.5 ns / 132.3 ns
8388608 : 114.0 ns / 141.6 ns
16777216 : 120.8 ns / 145.4 ns
33554432 : 124.3 ns / 146.9 ns
67108864 : 126.2 ns / 148.2 ns
##########################################################################
Executing ramlat on cpu0 (Cortex-A53), results in ns:
size: 1x32 2x32 1x64 2x64 1xPTR 2xPTR 4xPTR 8xPTR
4k: 2.006 1.988 1.490 1.490 1.490 1.490 2.049 4.161
8k: 1.987 1.987 1.490 1.490 1.490 1.490 2.049 4.160
16k: 1.988 1.987 1.490 1.490 1.490 1.490 2.049 4.160
32k: 1.992 1.991 1.493 1.492 1.493 1.493 2.055 4.166
64k: 16.77 15.52 13.49 15.07 13.49 15.19 18.50 33.43
128k: 17.34 17.10 15.97 16.80 15.97 16.88 19.85 38.14
256k: 17.99 17.93 17.48 17.84 17.46 17.92 20.56 39.26
512k: 115.8 131.8 123.9 128.8 123.7 129.7 164.7 304.4
1024k: 137.8 138.6 137.9 137.7 138.0 138.4 168.8 323.0
2048k: 122.7 123.0 122.7 122.4 122.7 122.8 161.7 312.3
4096k: 131.2 131.9 131.3 131.0 131.2 131.4 162.0 317.9
8192k: 134.5 134.7 134.5 134.2 134.5 134.7 165.7 319.1
16384k: 136.9 137.0 137.0 137.7 137.0 136.9 166.6 320.9
32768k: 137.2 137.6 137.2 136.8 137.2 137.2 168.3 319.5
65536k: 137.8 137.9 137.7 137.1 137.8 137.8 168.5 312.2
131072k: 138.3 138.3 138.2 137.8 138.2 138.2 165.9 311.3
Executing ramlat on cpu2 (Cortex-A73), results in ns:
size: 1x32 2x32 1x64 2x64 1xPTR 2xPTR 4xPTR 8xPTR
4k: 1.669 1.668 1.668 1.668 1.251 1.251 2.374 3.306
8k: 1.668 1.668 1.668 1.669 1.251 1.251 2.316 3.296
16k: 1.668 1.668 1.668 1.668 1.251 1.251 1.736 3.297
32k: 1.670 1.669 1.670 1.669 1.252 1.252 1.748 3.298
64k: 8.870 9.172 8.870 9.173 8.941 9.085 10.72 19.48
128k: 10.39 10.39 10.37 10.39 10.38 10.38 11.86 23.28
256k: 10.42 10.43 10.42 10.42 10.42 10.42 11.68 23.49
512k: 10.45 10.42 10.42 10.42 10.42 10.42 11.68 23.50
1024k: 11.32 10.96 11.26 10.90 11.27 13.66 12.31 24.57
2048k: 123.7 123.8 124.0 123.8 122.7 121.8 127.5 152.7
4096k: 132.4 132.4 131.9 132.3 131.4 130.8 134.4 163.7
8192k: 136.5 136.4 136.4 137.2 135.6 134.4 135.1 173.8
16384k: 138.7 138.6 138.7 138.7 138.6 136.7 136.6 178.6
32768k: 139.3 139.3 139.2 139.3 139.2 137.1 137.7 183.2
65536k: 141.3 141.4 141.3 141.4 141.4 140.2 140.0 183.7
131072k: 141.6 141.8 141.7 141.8 141.7 140.5 139.6 185.8
##########################################################################
Executing benchmark on each cluster individually
OpenSSL 1.1.1n, built on 15 Mar 2022
type 16 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes 16384 bytes
aes-128-cbc 130948.77k 416337.28k 890559.32k 1299143.68k 1499187.88k 1514897.41k (Cortex-A53)
aes-128-cbc 365161.19k 949810.65k 1515313.15k 1777747.97k 1902089.56k 1910489.09k (Cortex-A73)
aes-192-cbc 125298.14k 377637.70k 743570.26k 1011984.73k 1130378.58k 1127246.51k (Cortex-A53)
aes-192-cbc 340667.90k 850499.50k 1273256.45k 1503838.55k 1586883.24k 1593043.63k (Cortex-A73)
aes-256-cbc 122455.06k 352694.29k 653012.22k 851485.70k 933934.42k 938633.90k (Cortex-A53)
aes-256-cbc 326408.62k 777245.55k 1146589.18k 1295141.89k 1361668.78k 1366185.30k (Cortex-A73)
##########################################################################
Executing benchmark single-threaded on cpu0 (Cortex-A53)
7-Zip (a) [64] 16.02 : Copyright (c) 1999-2016 Igor Pavlov : 2016-05-21
p7zip Version 16.02 (locale=C,Utf16=off,HugeFiles=on,64 bits,6 CPUs LE)
LE
CPU Freq: 64000000 - - - 128000000 256000000 - - -
RAM size: 3785 MB, # CPU hardware threads: 6
RAM usage: 435 MB, # Benchmark threads: 1
Compressing | Decompressing
Dict Speed Usage R/U Rating | Speed Usage R/U Rating
KiB/s % MIPS MIPS | KiB/s % MIPS MIPS
22: 1170 99 1146 1139 | 21539 100 1847 1839
23: 1150 99 1180 1172 | 21052 100 1831 1822
24: 1135 99 1230 1221 | 20626 100 1820 1811
25: 1133 99 1303 1295 | 20257 100 1812 1803
---------------------------------- | ------------------------------
Avr: 99 1215 1207 | 100 1827 1819
Tot: 99 1521 1513
Executing benchmark single-threaded on cpu2 (Cortex-A73)
7-Zip (a) [64] 16.02 : Copyright (c) 1999-2016 Igor Pavlov : 2016-05-21
p7zip Version 16.02 (locale=C,Utf16=off,HugeFiles=on,64 bits,6 CPUs LE)
LE
CPU Freq: 64000000 - 64000000 - - - - - -
RAM size: 3785 MB, # CPU hardware threads: 6
RAM usage: 435 MB, # Benchmark threads: 1
Compressing | Decompressing
Dict Speed Usage R/U Rating | Speed Usage R/U Rating
KiB/s % MIPS MIPS | KiB/s % MIPS MIPS
22: 2049 100 1999 1993 | 31428 100 2686 2683
23: 2004 100 2048 2043 | 30713 100 2662 2659
24: 1984 100 2139 2134 | 29963 100 2634 2630
25: 1956 100 2239 2234 | 29289 100 2616 2607
---------------------------------- | ------------------------------
Avr: 100 2106 2101 | 100 2649 2645
Tot: 100 2378 2373
##########################################################################
Executing benchmark 3 times multi-threaded on CPUs 0-5
7-Zip (a) [64] 16.02 : Copyright (c) 1999-2016 Igor Pavlov : 2016-05-21
p7zip Version 16.02 (locale=C,Utf16=off,HugeFiles=on,64 bits,6 CPUs LE)
LE
CPU Freq: 64000000 64000000 - - - - - - -
RAM size: 3785 MB, # CPU hardware threads: 6
RAM usage: 1323 MB, # Benchmark threads: 6
Compressing | Decompressing
Dict Speed Usage R/U Rating | Speed Usage R/U Rating
KiB/s % MIPS MIPS | KiB/s % MIPS MIPS
22: 7419 552 1308 7218 | 140344 508 2357 11969
23: 7311 562 1325 7450 | 136778 507 2335 11835
24: 7175 563 1371 7715 | 128413 471 2395 11271
25: 7078 565 1431 8082 | 131084 507 2303 11666
---------------------------------- | ------------------------------
Avr: 560 1359 7616 | 498 2347 11685
Tot: 529 1853 9651
7-Zip (a) [64] 16.02 : Copyright (c) 1999-2016 Igor Pavlov : 2016-05-21
p7zip Version 16.02 (locale=C,Utf16=off,HugeFiles=on,64 bits,6 CPUs LE)
LE
CPU Freq: - - - - - - 512000000 - -
RAM size: 3785 MB, # CPU hardware threads: 6
RAM usage: 1323 MB, # Benchmark threads: 6
Compressing | Decompressing
Dict Speed Usage R/U Rating | Speed Usage R/U Rating
KiB/s % MIPS MIPS | KiB/s % MIPS MIPS
22: 7534 561 1308 7330 | 140002 507 2355 11940
23: 7241 556 1326 7378 | 136968 507 2335 11852
24: 7221 565 1373 7764 | 133806 508 2313 11744
25: 7141 568 1437 8154 | 131281 509 2297 11683
---------------------------------- | ------------------------------
Avr: 562 1361 7657 | 508 2325 11805
Tot: 535 1843 9731
7-Zip (a) [64] 16.02 : Copyright (c) 1999-2016 Igor Pavlov : 2016-05-21
p7zip Version 16.02 (locale=C,Utf16=off,HugeFiles=on,64 bits,6 CPUs LE)
LE
CPU Freq: 64000000 - - - - 256000000 - - -
RAM size: 3785 MB, # CPU hardware threads: 6
RAM usage: 1323 MB, # Benchmark threads: 6
Compressing | Decompressing
Dict Speed Usage R/U Rating | Speed Usage R/U Rating
KiB/s % MIPS MIPS | KiB/s % MIPS MIPS
22: 7584 562 1313 7378 | 140042 507 2357 11943
23: 7323 562 1328 7462 | 136997 507 2337 11854
24: 7269 569 1374 7816 | 133600 507 2314 11726
25: 7130 567 1435 8141 | 131397 509 2299 11694
---------------------------------- | ------------------------------
Avr: 565 1363 7699 | 507 2327 11804
Tot: 536 1845 9752
Compression: 7616,7657,7699
Decompression: 11685,11805,11804
Total: 9651,9731,9752
##########################################################################
Testing maximum cpufreq again, still under full load. System health now:
Time big.LITTLE load %cpu %sys %usr %nice %io %irq Temp
09:02:10: 2400/2016MHz 5.70 98% 1% 96% 0% 0% 0% 46.9°C
Checking cpufreq OPP for cpu0-cpu1 (Cortex-A53):
Cpufreq OPP: 2016 Measured: 2013 (2013.606/2013.557/2013.507)
Checking cpufreq OPP for cpu2-cpu5 (Cortex-A73):
Cpufreq OPP: 2400 Measured: 2398 (2398.769/2398.769/2398.658)
##########################################################################
Hardware sensors:
gpio_fan-isa-0000
fan1: 1600 RPM (min = 0 RPM, max = 1600 RPM)
cpu_thermal-virtual-0
temp1: +34.2 C (crit = +110.0 C)
ddr_thermal-virtual-0
temp1: +36.3 C (crit = +110.0 C)
##########################################################################
Thermal source: /sys/class/hwmon/hwmon0/ (cpu_thermal)
System health while running tinymembench:
Time big.LITTLE load %cpu %sys %usr %nice %io %irq Temp
08:50:48: 2400/2016MHz 1.07 1% 0% 0% 0% 0% 0% 31.0°C
08:51:08: 2400/2016MHz 1.05 16% 0% 16% 0% 0% 0% 29.1°C
08:51:28: 2400/2016MHz 1.04 16% 0% 16% 0% 0% 0% 29.6°C
08:51:48: 2400/2016MHz 1.02 16% 0% 16% 0% 0% 0% 29.4°C
08:52:08: 2400/2016MHz 1.02 16% 0% 16% 0% 0% 0% 29.1°C
08:52:28: 2400/2016MHz 1.01 16% 0% 16% 0% 0% 0% 31.4°C
08:52:48: 2400/2016MHz 1.01 16% 0% 16% 0% 0% 0% 32.4°C
08:53:08: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 30.8°C
System health while running ramlat:
Time big.LITTLE load %cpu %sys %usr %nice %io %irq Temp
08:53:28: 2400/2016MHz 1.00 2% 0% 1% 0% 0% 0% 32.1°C
08:53:34: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 29.6°C
08:53:40: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 29.7°C
08:53:46: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 29.7°C
08:53:52: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 29.6°C
08:53:58: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 29.6°C
08:54:04: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 30.5°C
08:54:10: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 30.8°C
08:54:16: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 30.6°C
08:54:22: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 30.3°C
08:54:28: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 30.6°C
System health while running OpenSSL benchmark:
Time big.LITTLE load %cpu %sys %usr %nice %io %irq Temp
08:54:34: 2400/2016MHz 1.08 2% 0% 1% 0% 0% 0% 32.2°C
08:54:51: 2400/2016MHz 1.06 16% 0% 16% 0% 0% 0% 29.7°C
08:55:07: 2400/2016MHz 1.12 16% 0% 16% 0% 0% 0% 31.7°C
08:55:23: 2400/2016MHz 1.09 16% 0% 16% 0% 0% 0% 29.8°C
08:55:39: 2400/2016MHz 1.07 16% 0% 16% 0% 0% 0% 32.1°C
08:55:55: 2400/2016MHz 1.05 16% 0% 16% 0% 0% 0% 29.9°C
08:56:11: 2400/2016MHz 1.04 16% 0% 16% 0% 0% 0% 32.3°C
System health while running 7-zip single core benchmark:
Time big.LITTLE load %cpu %sys %usr %nice %io %irq Temp
08:56:23: 2400/2016MHz 1.03 2% 0% 2% 0% 0% 0% 33.5°C
08:56:30: 2400/2016MHz 1.03 16% 0% 16% 0% 0% 0% 30.0°C
08:56:37: 2400/2016MHz 1.02 16% 0% 16% 0% 0% 0% 29.9°C
08:56:44: 2400/2016MHz 1.02 16% 0% 16% 0% 0% 0% 30.0°C
08:56:51: 2400/2016MHz 1.02 16% 0% 16% 0% 0% 0% 29.9°C
08:56:58: 2400/2016MHz 1.02 16% 0% 16% 0% 0% 0% 29.8°C
08:57:05: 2400/2016MHz 1.01 16% 0% 16% 0% 0% 0% 30.0°C
08:57:12: 2400/2016MHz 1.01 16% 0% 16% 0% 0% 0% 30.0°C
08:57:19: 2400/2016MHz 1.01 16% 0% 16% 0% 0% 0% 29.7°C
08:57:26: 2400/2016MHz 1.01 16% 0% 16% 0% 0% 0% 29.8°C
08:57:33: 2400/2016MHz 1.01 16% 0% 16% 0% 0% 0% 30.0°C
08:57:40: 2400/2016MHz 1.01 16% 0% 16% 0% 0% 0% 29.9°C
08:57:47: 2400/2016MHz 1.01 16% 0% 16% 0% 0% 0% 29.9°C
08:57:54: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 29.9°C
08:58:01: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 31.1°C
08:58:08: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 31.5°C
08:58:15: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 31.6°C
08:58:22: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 31.7°C
08:58:29: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 31.7°C
08:58:36: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 31.7°C
08:58:43: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 31.7°C
08:58:50: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 31.7°C
08:58:57: 2400/2016MHz 1.00 16% 0% 16% 0% 0% 0% 31.7°C
System health while running 7-zip multi core benchmark:
Time big.LITTLE load %cpu %sys %usr %nice %io %irq Temp
08:59:02: 2400/2016MHz 1.00 3% 0% 2% 0% 0% 0% 38.4°C
08:59:13: 2400/2016MHz 1.84 97% 0% 96% 0% 0% 0% 44.7°C
08:59:25: 2400/2016MHz 2.11 85% 0% 84% 0% 0% 0% 44.3°C
08:59:35: 2400/2016MHz 2.97 80% 1% 79% 0% 0% 0% 44.1°C
08:59:45: 2400/2016MHz 3.36 74% 0% 73% 0% 0% 0% 40.6°C
08:59:56: 2400/2016MHz 4.07 96% 1% 94% 0% 0% 0% 40.9°C
09:00:06: 2400/2016MHz 4.13 85% 0% 83% 0% 0% 0% 40.6°C
09:00:17: 2400/2016MHz 4.49 97% 0% 96% 0% 0% 0% 46.1°C
09:00:29: 2400/2016MHz 4.65 85% 0% 84% 0% 0% 0% 46.0°C
09:00:42: 2400/2016MHz 4.72 83% 1% 82% 0% 0% 0% 45.7°C
09:00:53: 2400/2016MHz 4.70 81% 1% 80% 0% 0% 0% 41.7°C
09:01:07: 2400/2016MHz 5.05 98% 1% 96% 0% 0% 0% 46.1°C
09:01:20: 2400/2016MHz 5.55 88% 0% 87% 0% 0% 0% 46.7°C
09:01:33: 2400/2016MHz 5.79 87% 0% 86% 0% 0% 0% 46.8°C
09:01:46: 2400/2016MHz 5.61 84% 1% 82% 0% 0% 0% 46.6°C
09:01:57: 2400/2016MHz 5.47 81% 1% 80% 0% 0% 0% 43.2°C
09:02:10: 2400/2016MHz 5.70 98% 1% 96% 0% 0% 0% 46.9°C
##########################################################################
Linux 6.1.11-meson64 (odroidn2) 03/27/23 _aarch64_ (6 CPU)
avg-cpu: %user %nice %system %iowait %steal %idle
6.41 0.00 0.45 0.02 0.00 93.11
Device tps kB_read/s kB_wrtn/s kB_dscd/s kB_read kB_wrtn kB_dscd
mmcblk1 5.04 117.72 134.88 0.00 518916 594573 0
zram0 0.14 0.54 0.00 0.00 2384 4 0
zram1 0.12 0.12 0.56 0.00 516 2452 0
total used free shared buff/cache available
Mem: 3.7Gi 181Mi 3.4Gi 5.0Mi 133Mi 3.4Gi
Swap: 1.8Gi 0B 1.8Gi
Filename Type Size Used Priority
/dev/zram0 partition 1938044 0 5
CPU sysfs topology (clusters, cpufreq members, clockspeeds)
cpufreq min max
CPU cluster policy speed speed core type
0 0 0 1000 2016 Cortex-A53 / r0p4
1 0 0 1000 2016 Cortex-A53 / r0p4
2 0 2 1000 2400 Cortex-A73 / r0p2
3 0 2 1000 2400 Cortex-A73 / r0p2
4 0 2 1000 2400 Cortex-A73 / r0p2
5 0 2 1000 2400 Cortex-A73 / r0p2
Architecture: aarch64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
CPU(s): 6
On-line CPU(s) list: 0-5
Thread(s) per core: 1
Core(s) per socket: 6
Socket(s): 1
NUMA node(s): 1
Vendor ID: ARM
Model: 4
Model name: Cortex-A53
Stepping: r0p4
CPU max MHz: 2400.0000
CPU min MHz: 1000.0000
BogoMIPS: 48.00
NUMA node0 CPU(s): 0-5
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec store bypass: Vulnerable
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Vulnerable
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 cpuid
SoC guess: Amlogic Meson G12B (S922X) Revision 29:c (40:2)
DT compat: hardkernel,odroid-n2-plus
amlogic,s922x
amlogic,g12b
Compiler: /usr/bin/gcc (Debian 10.2.1-6) 10.2.1 20210110 / aarch64-linux-gnu
Userland: arm64
Kernel: 6.1.11-meson64/aarch64
CONFIG_HZ=250
CONFIG_HZ_250=y
CONFIG_PREEMPTION=y
CONFIG_PREEMPT=y
CONFIG_PREEMPT_BUILD=y
CONFIG_PREEMPT_COUNT=y
CONFIG_PREEMPT_NOTIFIERS=y
CONFIG_PREEMPT_RCU=y
##########################################################################
Kernel 6.1.11 is not latest 6.1.21 LTS that was released on 2023-03-22.
See https://endoflife.date/linux for details. Perhaps some kernel bugs have
been fixed in the meantime and maybe vulnerabilities as well.
##########################################################################
opp-table-0:
1000 MHz 761.0 mV
1200 MHz 781.0 mV
1398 MHz 811.0 mV
1512 MHz 861.0 mV
1608 MHz 901.0 mV
1704 MHz 951.0 mV
1800 MHz 1001.0 mV
1908 MHz 1030.0 mV
2016 MHz 1040.0 mV
opp-table-1:
1000 MHz 731.0 mV
1200 MHz 751.0 mV
1398 MHz 771.0 mV
1512 MHz 771.0 mV
1608 MHz 781.0 mV
1704 MHz 791.0 mV
1800 MHz 831.0 mV
1908 MHz 861.0 mV
2016 MHz 911.0 mV
2108 MHz 951.0 mV
2208 MHz 1011.0 mV
2304 MHz 1030.0 mV
2400 MHz 1040.0 mV
opp-table-gpu:
125 MHz 800.0 mV
250 MHz 800.0 mV
286 MHz 800.0 mV
400 MHz 800.0 mV
500 MHz 800.0 mV
667 MHz 800.0 mV
800 MHz 800.0 mV
##########################################################################
Results validation:
* No mismatch between advertised and measured max CPU clockspeed
* No swapping
* Background activity (%system) OK
* No throttling
Status of performance related governors found below /sys (w/o cpufreq):
* ffe40000.gpu: simple_ondemand / 125 MHz (powersave performance simple_ondemand / 125 250 286 400 500 667 800)
| Hardkernel ODROID-N2Plus | 2400/2016 MHz | 6.1 | Debian GNU/Linux 11 (bullseye) arm64 | 9710 | 2373 | 1366180 | 4220 | 7720 | - |