Эльбрус/тесты/результаты: различия между версиями

Материал из ALT Linux Wiki
(Updated JS results)
(Added "Тесты Crystal Mark 2004")
Строка 70: Строка 70:
| 256M || 105.46ns || 137.1 || 120.48ns || 186.7 ||  12.17ns || 31.6 ||  37.77ns || 43.5
| 256M || 105.46ns || 137.1 || 120.48ns || 186.7 ||  12.17ns || 31.6 ||  37.77ns || 43.5
|}
|}
== Тесты Crystal Mark 2004 R3/R7 ==
{| class="wikitable"
! CPU || Threads || Frequency || ALU || FPU || MEM R (Mb/s) || MEM W (Mb/s) || Anounced
|-
| 486 DX4 || 1 || 75 || 119 || 77 || 9 || 11 || 1993
|-
| P1 (P54C) || 1 || 200 || 484 || 420 || 80 || 65 || 1994
|-
| P1 MMX (P55C) || 1 || 233 || 675 || 686 || 112 || 75 || 1997
|-
| P2  || 1 || 400 || 1219 || 1260 || 222 || 150 || 1998
|-
| Transmeta Crusoe TM5800 || 1 || 1000 || 2347 || 1689 || 405 || 223 || 2000
|-
| P3 (Coopermine) || 1 || 1000 || 3440 || 3730 || 355 || 170 || 2000
|-
| P4 (Willamete) || 1 || 1600 || 3496 || 4110 || 1385 || 662 || 2001
|-
| Celeron (Willamete) || 1 || 1800 || 3934 || 4594 || 1457 || 657 || 2001
|-
| Athlon XP (Palomino) || 1 || 1400 || 4450 || 6220 || 430 || 520 || 2001
|-
| P4 (Northwood) || 1 || 2400 || 5661 || 6747 || 1765 || 754 || 2002
|-
| P4 (Prescott) || 1 || 2800 || 5908 || 6929 || 3744 || 851 || 2004
|-
| Athlon 64 (Venice) || 1 || 1800 || 6699 || 7446 || 1778 || 906 || 2005
|-
| Celeron 530 (Conroe-L) || 1 || 1733 || 7806 || 9117 || 3075 || 1226 || 2006
|-
| P4 (Prescott) || 2 || 3000 || 9719 || 10233 || 3373 || 1578 || 2004
|-
| Atom D525 || 4 || 1800 || 10505 || 7605 || 3407 || 1300 || 2010
|-
| Athlon 64 X2 (Brisbane) || 2 || 2300 || 16713 || 19066 || 3973 || 2728 || 2007
|-
| Core i3-6100 || 2 || 3700 || 17232 || 10484 || 5553 || 9594 || 2015
|-
| Pentium T3200 (Merom) || 2 || 2000 || 20702 || 18063 || 4150 || 1598 || 2008
|-
| Atom x5-Z8350 || 4 || 1440 || 21894 || 18018 || 4799 || 2048 || 2016
|-
| Core i3-M330 || 4 || 2133 || 25595 || 26627 || 6807 || 4257 || 2010
|-
| Core 2 Duo || 2 || 3160 || 28105 || 18196 || 6850 || 2845 || 2008
|-
| Atom Z3795 || 4 || 1600 || 40231 || 34963 || 12060 || 5797 || 2016
|-
| AMD A6-3650 || 4 || 2600 || 46978 || 35315 || 9711 || 3870 || 2011
|-
| Core 2 Quad || 4 || 2833 || 47974 || 31391 || 9710 || 5493 || 2008
|-
| Core i3-4130 || 4 || 3400 || 54296 || 39163 || 19450 || 9269 || 2013
|-
| AMD Phenom II X4 965 (Agena) || 4 || 3400 || 59098 || 56272 || 11162 || 5973 || 2009
|-
| Core i7-2600 || 8 || 3400 || 95369 || 71648 || 19547 || 9600 || 2011
|-
| Core i7-9900K || 16 || 3600 || 270445 || 238256 || 44618 || 17900 || 2018
|-
| Elbrus-8C RTC-x86 || 8 || 1300 || 65817 || 29977 || 49800 || 7945 || 2016
|-
| Elbrus-8CB RTC-x86 || 8 || 1500 || 77481 || 37972 || 62100 || 13940 || 2018
|-
| Elbrus-1C+ RTC-x86 || 1 || 1000 || 6862 || 2735 || 6230 || 1800 || 2015
|}


== Тесты .NET (C#) [EntityFX Bench] ==
== Тесты .NET (C#) [EntityFX Bench] ==

Версия от 14:46, 13 октября 2020

Нативные тесты

Исходный код: Anybench

Для архитектуры E2K собраны с LCС, для X86 и ARM с GCC или CLang.

Native classic benchmarks
Test Elbrus E8CB Elbrus E8C Elbrus 4C Elbrus 1C+ Сore i7-2600 Atom Z8350 Orange Pi PC2
Dhrystone 8 974.78 7779.40 3 548.80 4 302.53 22 076.97 4 677.30 2 949.12
Whetstone 2 037.62 1 748.37 970.80 1 277.55 5 729.52 2 085.24 980.26
Whetstone MP 16 194.00 13 818.00 2 455.00 1 312.00 31319.00 6 636.00 3 798.00
Coremark 5 510.19 4 907.57 2 364.24 2 901.49 22 692.89 6 893.09 3 869.72
Coremark MP 39 941.90 35 395.62 9 078.68 2 848.32 119 670.91 23 814.68 14 901.28
Linpack 1 269.79 1 075.27 674.68 814.76 4 302.89 1 021.44 163.44
Scimark 2 (Composite) 472.24 511.43 - 379.23 1 800.00 509.44 191.59
MP MFLOPS (32 ops/w) 378 976.00 139 265.00 35 782.00 15 676.00 84 536.00 10 665.00 6 033.00

Тест латентности кеша

Исходный код: Test TLB

Elbrus 8C (1.3 GHz) Elbrus 8CB (1.5 GHz) Amd A6-3650 (2.6 GHz) Alwinner A64 (1152 MHz)
Size Latency Cycles Latency Cycles Latency Cycles Latency Cycles
4k 5.39ns 7 4.52ns 7 1.16ns 3 3.48ns 4
8k 5.39ns 7 4.52ns 7 1.16ns 3 3.48ns 4
16k 5.39ns 7 4.52ns 7 1.16ns 3 3.48ns 4
32k 5.39ns 7 4.52ns 7 1.16ns 3 3.58ns 4.1
64k 5.39ns 7 4.52ns 7 1.16ns 3 6.83ns 7.9
128k 8.48ns 11 7.10ns 11 4.00ns 10.4 7.28ns 8.4
256k 8.48ns 11 7.10ns 11 4.00ns 10.4 7.69ns 8.9
512k 8.56ns 11.1 7.13ns 11.1 4.00ns 10.4 8.04ns 9.3
1M 25.43ns 33.1 21.31ns 33 4.03ns 10.5 34.36ns 39.6
2M 25.43ns 33.1 21.31ns 33 12.14ns 31.6 37.05ns 42.7
4M 25.43ns 33.1 21.31ns 33 12.06ns 31.3 37.36ns 43
6M 25.44ns 33.1 21.31ns 33 12.21ns 31.7 37.59ns 43.3
8M 25.46ns 33.1 21.31ns 33 12.01ns 31.2 37.37ns 43
16M 46.09ns 59.9 27.99ns 43.4 12.02ns 31.3 37.57ns 43.3
32M 105.44ns 137.1 118.83ns 184.2 11.98ns 31.2 37.09ns 42.7
64M 105.44ns 137.1 120.01ns 186 12.02ns 31.3 37.07ns 42.7
128M 105.47ns 137.1 120.00ns 186 12.21ns 31.7 37.57ns 43.3
256M 105.46ns 137.1 120.48ns 186.7 12.17ns 31.6 37.77ns 43.5

Тесты Crystal Mark 2004 R3/R7

CPU Threads Frequency ALU FPU MEM R (Mb/s) MEM W (Mb/s) Anounced
486 DX4 1 75 119 77 9 11 1993
P1 (P54C) 1 200 484 420 80 65 1994
P1 MMX (P55C) 1 233 675 686 112 75 1997
P2 1 400 1219 1260 222 150 1998
Transmeta Crusoe TM5800 1 1000 2347 1689 405 223 2000
P3 (Coopermine) 1 1000 3440 3730 355 170 2000
P4 (Willamete) 1 1600 3496 4110 1385 662 2001
Celeron (Willamete) 1 1800 3934 4594 1457 657 2001
Athlon XP (Palomino) 1 1400 4450 6220 430 520 2001
P4 (Northwood) 1 2400 5661 6747 1765 754 2002
P4 (Prescott) 1 2800 5908 6929 3744 851 2004
Athlon 64 (Venice) 1 1800 6699 7446 1778 906 2005
Celeron 530 (Conroe-L) 1 1733 7806 9117 3075 1226 2006
P4 (Prescott) 2 3000 9719 10233 3373 1578 2004
Atom D525 4 1800 10505 7605 3407 1300 2010
Athlon 64 X2 (Brisbane) 2 2300 16713 19066 3973 2728 2007
Core i3-6100 2 3700 17232 10484 5553 9594 2015
Pentium T3200 (Merom) 2 2000 20702 18063 4150 1598 2008
Atom x5-Z8350 4 1440 21894 18018 4799 2048 2016
Core i3-M330 4 2133 25595 26627 6807 4257 2010
Core 2 Duo 2 3160 28105 18196 6850 2845 2008
Atom Z3795 4 1600 40231 34963 12060 5797 2016
AMD A6-3650 4 2600 46978 35315 9711 3870 2011
Core 2 Quad 4 2833 47974 31391 9710 5493 2008
Core i3-4130 4 3400 54296 39163 19450 9269 2013
AMD Phenom II X4 965 (Agena) 4 3400 59098 56272 11162 5973 2009
Core i7-2600 8 3400 95369 71648 19547 9600 2011
Core i7-9900K 16 3600 270445 238256 44618 17900 2018
Elbrus-8C RTC-x86 8 1300 65817 29977 49800 7945 2016
Elbrus-8CB RTC-x86 8 1500 77481 37972 62100 13940 2018
Elbrus-1C+ RTC-x86 1 1000 6862 2735 6230 1800 2015


Тесты .NET (C#) [EntityFX Bench]

Исходный код: EntityFX Bench

Бинарники Net Framework/Mono, DotNetCore: [1]

.Net benchmarks
CPU Elbrus 1C+ Elbrus 4C Elbrus 8C Elbrus 8CB Pentium 4 2800 Atom X5 Z8350 Pentium T8400 Core i3 M330 AMD A6 3650 Core i5 4460 Core i7 2600 Allwinner A64 Amlogic S922X
OS Linux 4.9 Linux 4.9 Linux 4.19 Linux 5.4 Linux 5.4 Windows 10 Linux 5.4 Windows 10 Linux 5.4 Windows 7 SP1 Windows 10 Linux 4.14 Linux 4.9
Runtime Mono 5.16 Mono 5.16 Mono 5.16 NetCore 3.1.1 (RTC) NetCore 3.1.1 (RTC) Mono Net 4.7.1 NetCore 3.1.8 Mono Net 4.7.1 NetCore 3.1.3 Net 4.7.1 Net 4.7.1 Net 4.7.1 NetCore 3.1.1 Mono NetCore 3.1.1 Mono
Threads 1 4 8 8 8 1 4 4 2 4 4 4 4 8 8 4 4 6
MemoryBenchmark 760.14 622.24 685.80 2422.27 2891.87 1898.20 2722.99 2352.93 4087.59 11077.00 5200.87 3859.76 8186.80 12330.27 9802.08 447.93 701.25 1257.51
ParallelMemoryBenchmark 761.32 2414.69 4603.87 16957.66 20658.81 2053.45 8506.48 7609.98 7051.09 17913.30 8142.90 14326.32 28198.26 41026.21 33404.25 1564.49 2092.21 6621.61
RandomMemoryBenchmark 435.22 338.56 588.17 2099.81 2620.34 1366.04 1755.61 1569.19 2682.57 3967.03 2183.32 2287.64 7822.70 9929.41 5325.75 261.84 351.96 1357.76
ParallelRandomMemoryBenchmark 432.84 2017.68 6468.92 23310.51 27765.76 1436.56 11133.39 10272.32 7793.85 12790.70 8603.97 15015.87 60364.21 64813.43 44307.60 2217.48 2983.75 9839.12
Scimark2Benchmark 567.95 389.98 743.62 1602.46 1691.94 1253.99 1435.89 1508.57 2507.98 2778.18 2505.88 2142.08 5525.90 5847.06 5757.59 330.27 462.51 1253.99
ParallelScimark2Benchmark 567.95 1673.70 5814.07 12390.16 12876.85 1253.99 5743.56 5743.56 5015.95 5743.56 5306.66 8131.40 21821.16 26921.01 25957.53 1321.10 1883.59 5691.26
DhrystoneBenchmark 188.96 143.43 218.55 5755.26 5058.62 735.34 2784.29 3123.51 1306.77 2604.44 1964.71 1019.92 15392.15 5101.88 5459.31 207.17 1803.07 785.43
ParallelDhrystoneBenchmark 188.96 573.71 1723.39 38069.44 48628.34 730.79 10693.23 11742.74 2397.27 3911.21 4068.30 3776.89 57757.54 20874.22 20749.00 764.94 6697.78 4136.60
WhetstoneBenchmark 469.62 357.30 833.15 376.47 407.61 412.62 940.33 1105.60 2068.04 1769.97 1652.29 1547.00 2719.55 3043.93 3681.81 455.98 388.74 500.83
ParallelWhetstoneBenchmark 388.83 1206.25 5127.83 8108.30 9430.39 548.21 3955.52 4656.08 3923.04 4171.13 4817.22 5990.89 9328.49 18751.93 22840.84 1342.66 2376.01 4909.97
ArithemticsBenchmark 111.34 84.81 142.69 3452.37 4036.59 45.47 999.72 843.85 272.51 1812.49 1190.11 178.11 3256.83 4289.95 4349.14 150.75 296.73 316.23
ParallelArithemticsBenchmark 281.14 855.99 2909.45 29867.62 35469.08 211.19 3765.26 4222.82 3546.56 3108.32 3408.16 4792.36 13089.84 18593.81 17479.06 546.81 1416.41 2868.45
MathBenchmark 486.82 373.61 637.97 727.87 855.71 629.40 911.21 815.32 2334.08 2293.70 1772.34 1707.83 4002.56 4323.03 4759.68 489.01 556.92 1314.58
ParallelMathBenchmark 489.24 1493.84 5011.26 5350.82 7110.93 622.19 3674.23 3138.42 4315.21 4860.99 3889.19 6699.91 14202.83 23102.72 21920.96 1804.51 2124.36 7018.97
CallBenchmark 393.50 299.63 463.67 12801.83 15289.98 2314.05 2762.38 2780.03 2403.37 2138.95 1822.57 3190.05 3870.43 4147.75 3949.32 601.68 711.39 1637.24
IfElseBenchmark 578.74 440.61 763.63 3691.38 4402.63 759.79 1220.58 1220.11 2403.80 1780.26 1600.94 2095.29 3281.63 3238.56 3354.13 356.41 473.46 1036.86
StringManipulation 323.07 248.85 303.84 1765.31 1856.23 860.94 1229.26 1950.88 1276.23 2754.18 3680.63 1095.26 4784.59 4643.01 6411.47 437.34 846.89 1021.18
ParallelStringManipulation 322.18 943.09 2154.68 2020.58 2311.31 863.34 1847.18 2273.31 2390.88 2853.96 3440.47 4121.52 3820.96 4890.19 5773.23 1520.99 1309.55 4818.13
HashBenchmark 373.05 290.95 456.06 0 0 614.49 549.50 620.60 1588.19 1367.62 1876.64 1406.19 2674.67 2464.86 3231.08 251.70 564.13 695.25
ParallelHashBenchmark 372.57 992.32 2392.49 0 0 612.40 1151.26 1369.97 2679.98 1586.14 2291.17 3209.11 2324.80 3027.87 6174.11 876.97 1363.53 2737.16
Total Time 00:37:15 00:43:36 00:31:11 00:17:59 00:16:23 00:22:55 00:15:12 00:14:52 00:10:25 00:11:26 00:11:09 00:12:27 00:08:06 00:08:42 00:07:52 00:37:32 00:26:08 00:18:32

Тесты JavaScript (Browser JS) [EntityFX Bench]

Исходный код: EntityFX Bench

CPU Elbrus 8C Elbrus 1C+ AMD A6 3650 Core i7 2600 Atom X5-Z8350 Core i3 M330 Qualcomm 625 Allwinner A64 Mediatek MT6589
OS Linux 4.19 Linux 4.19 Linux 5.4 Windows 10 Windows 10 Windows 10 Android 7 Linux 4.14 Android 4.4
Runtime Firefox 52 Firefox 52 Firefox 81 Firefox 81 Chrome 85 Chrome 85 Firefox 81 Firefox 81 Chrome 80
MemoryBenchmark 655.06 453.69 4476.48 6350.50 1501.43 1733.91 1739.09 996.74 404.80
RandomMemoryBenchmark 20.97 16.17 132.43 187.39 193.08 344.14 42.17 19.61 83.04
Scimark2Benchmark 1387.60 777.67 4271.13 6335.21 2412.89 3378.66 1227.36 699.49 525.10
DhrystoneBenchmark 822.33 542.33 2311.75 3619.98 1501.82 2569.53 931.43 473.48 561.53
WhetstoneBenchmark 368.97 49.72 2091.21 2784.54 297.19 588.93 1006.00 447.75 143.29
ArithemticsBenchmark 310.88 209.90 1059.45 1140.97 348.73 723.30 321.36 183.63 200.89
MathBenchmark 644.43 466.20 1618.67 3331.11 358.42 612.93 345.75 518.28 172.10
CallBenchmark 215.94 149.24 1618.52 2262.96 3114.29 3134.30 427.03 244.66 919.67
IfElseBenchmark 719.99 346.18 3743.22 3409.48 758.09 1835.70 676.84 391.25 514.56
StringManipulation 290.16 208.01 1877.51 4473.87 1257.04 2054.65 763.02 669.67 525.24
HashBenchmark 58.50 36.66 225.58 407.07 128.87 272.45 105.00 58.79 62.02
Total Points 5494.83 3255.76 23425.94 34303.09 11871.85 17248.52 7585.06 4703.36 4112.24
Total Time (sec) 3016 4105 599 412 864 506 1781 3159 1827


Javascript. Популярные бенчмарки

Octane 2

Octane 2

Cpu Result (ms)
Intel Pentium 4 2800 3387
AMD A6-3650 11741
Elbrus 8C (rtc x86 32bit) 2815
Elbrus 8C 2102
Elbrus 1C+ 739

Kraken Benchmark

Kraken Benchmark

Cpu Result (ms)
Elbrus 8C 10493.4
Elbrus 8CB RTX x86 9567.5
Elbrus 8CB 8714.2
Intel Pentium 4 2800 9486.6
AMD A6-3650 (4 core 2.6 GHz) 3052.5

Sunspider

Sunspider

Cpu Result (ms)
Elbrus 8C 3059.8
Elbrus 8CB 2394.6
Intel Pentium 4 2800 1295.5
AMD A6-3650 485.6

PHP benchmark

Test Elbrus 8C Elbrus 8CB Pentium 4 2800 AMD A6-3650 Allwinner A64
Version 7.0.33 7.0.33 7.2.24 7.4.3 7.0.33
01_math 58.15 kOp/s 69.72 kOp/s 104.19 kOp/s 295.97 kOp/s 44.33 kOp/s
02_string_concat 3.56 MOp/s 3.92 MOp/s 4.00 MOp/s 13.15 MOp/s 3.07 MOp/s
03_1_string_number_concat 418.29 kOp/s 472.77 kOp/s 631.10 kOp/s 1.51 MOp/s 332.99 kOp/s
03_2_string_number_format 506.39 kOp/s 573.89 kOp/s 724.44 kOp/s 1.69 MOp/s 432.88 kOp/s
04_string_simple_functions 77.06 kOp/s 91.50 kOp/s 198.03 kOp/s 332.67 kOp/s 59.48 kOp/s
05_string_multibyte 2.48 kOp/s 2.90 kOp/s -.--Op/s 57.53 kOp/s 2.50 kOp/s
06_string_manipulation 22.10 kOp/s 26.91 kOp/s 78.96 kOp/s 127.08 kOp/s 35.73 kOp/s
07_regex 48.24 kOp/s 54.60 kOp/s 128.41 kOp/s 233.76 kOp/s 47.64 kOp/s
08_1_hashing 113.58 kOp/s 132.62 kOp/s 180.46 kOp/s 306.24 kOp/s 71.44 kOp/s
08_2_crypt 361.21 Op/s 403.62 Op/s 571.99 Op/s 813.60 Op/s 238.00 Op/s
09_json_encode -.--Op/s -.--Op/s 88.33 kOp/s 233.62 kOp/s 48.67 kOp/s
10_json_decode -.--Op/s -.--Op/s 68.02 kOp/s 143.01 kOp/s 33.57 kOp/s
11_serialize 73.67 kOp/s 81.57 kOp/s 130.16 kOp/s 307.52 kOp/s 62.20 kOp/s
12_unserialize 63.89 kOp/s 69.02 kOp/s 79.33 kOp/s 301.98 kOp/s 46.21 kOp/s
13_array_fill 2.08 MOp/s 2.50 MOp/s 5.30 MOp/s 9.69 MOp/s 1.97 MOp/s
14_array_range 50.36 kOp/s 57.54 kOp/s 31.68 kOp/s 61.01 kOp/s 25.25 kOp/s
14_array_unset 2.08 MOp/s 2.48 MOp/s 7.17 MOp/s 14.05 MOp/s 2.16 MOp/s
15_loops 13.57 MOp/s 16.21 MOp/s 38.75 MOp/s 150.46 MOp/s 12.64 MOp/s
16_loop_ifelse 4.74 MOp/s 5.64 MOp/s 13.41 MOp/s 28.34 MOp/s 4.48 MOp/s
17_loop_ternary 3.18 MOp/s 3.79 MOp/s 7.29 MOp/s 12.10 MOp/s 2.90 MOp/s
18_1_loop_defined_access 3.28 MOp/s 3.90 MOp/s 9.03 MOp/s 18.90 MOp/s 3.18 MOp/s
18_2_loop_undefined_access 604.06 kOp/s 665.54 kOp/s 1.13 MOp/s 2.60 MOp/s 485.23 kOp/s
19_type_functions 250.57 kOp/s 293.21 kOp/s 806.37 kOp/s 1.56 MOp/s 193.89 kOp/s
20_type_conversion 382.32 kOp/s 458.44 kOp/s 812.72 kOp/s 1.57 MOp/s 298.61 kOp/s
21_0_loop_exception_none 7.45 MOp/s 8.91 MOp/s 19.67 MOp/s 56.57 MOp/s 6.97 MOp/s
21_1_loop_exception_try 6.48 MOp/s 7.74 MOp/s 19.11 MOp/s 52.18 MOp/s 6.39 MOp/s
21_2_loop_exception_catch 184.22 kOp/s 216.00 kOp/s 573.09 kOp/s 1.38 MOp/s 147.28 kOp/s
22_loop_null_op 3.25 MOp/s 3.74 MOp/s 8.39 MOp/s 16.03 MOp/s 3.08 MOp/s
23_loop_spaceship_op 4.30 MOp/s 5.12 MOp/s 8.50 MOp/s 17.98 MOp/s 3.96 MOp/s
24_xmlrpc_encode -.--Op/s -.--Op/s -.--Op/s -.--Op/s -.--Op/s
25_xmlrpc_decode -.--Op/s -.--Op/s -.--Op/s -.--Op/s -.--Op/s
26_1_class_public_properties 3.32 MOp/s 4.08 MOp/s 10.51 MOp/s 26.70 MOp/s 3.22 MOp/s
26_2_class_getter_setter 1.31 MOp/s 1.51 MOp/s 4.66 MOp/s 9.41 MOp/s 969.36 kOp/s
26_3_class_magic_methods 525.45 kOp/s 592.98 kOp/s 1.35 MOp/s 3.77 MOp/s 409.73 kOp/s
Total 1.23 MOp/s 1.43 MOp/s 2.60 MOp/s 5.33 MOp/s 987.79 kOp/s
Time 488.324 sec 419.895 sec 231.485 sec 113.087 sec 609.787 sec