PDA

Bekijk de volledige versie : Сравнение производительности процессоров роутеров



ryzhov_al
28-08-2012, 21:53
Коллеги, предлагаю поучаствовать в древней забаве - померять у кого длин...
.. больше производительность CPU на разных моделях роутеров.
Просто для того, чтобы представлять, какими вычислительными ресурсами мы располагаем.

Будем использовать cpubench (https://bitbucket.org/lokkju/openpli-oe/src/4c4185cfe184/recipes/cpubench/files), издавна созданный OpenWRT'шниками для этих целей.
В приложении найдёте статический бинарник, работа которого не зависит от текущей прошивки.
Просто напишите модель роутера и вывод теста, образец в следующем посте.

Тест выполняйте при остановленных торрент-качалках и других внешних сервисах.
Если ваш результат практически не отличается от уже имеющегося, то публиковать его не стоит.

Подразумевается, что у вас не разогнанный/модифицированный роутер.
Если это не так, то напишите тактовую частоту CPU/суть мода.
В табличку сведены лучшие показатели, в порядке убывания производительности:



Vendor, Model, Version
Freq. CPU
Memory bench
[secs]
Computation
of pi [secs]
Computation
of e [secs]
Float bench
[secs]
Total time
[secs]


Asus RT-AC68U
@1600MHz* (http://wl500g.info/showthread.php?30444&p=266271#post266271)
1.19
3.29
2.83
0.03
7.3


Asus RT-N56U B1
@900MHz* (http://wl500g.info/showthread.php?30444&p=269143#post269143)
2.14
3.69
3.69
1.47
11.0


Asus RT-AC56U
@1000MHz* (http://wl500g.info/showthread.php?30444&p=261151#post261151)
1.51
5.32
4.19
0.04
11.1


Asus RT-AC68U
@1000MHz* (https://wl500g.info/showthread.php?30444&p=273779#post273779)
1.70
4.90
4.42
0.04
11.1


Asus RT-N66U
@662MHz* (http://wl500g.info/showthread.php?30444&p=266269#post266269)
1.84
4.77
3.85
1.88
12.3


Asus RT-AC66U
@662MHz* (http://pastebin.com/c10piJ4t)
1.81
4.75
3.84
1.92
12.3


Asus RT-AC56U
@800MHz (default) (http://wl500g.info/showthread.php?30444&p=261151#post261151)
1.62
6.64
5.22
0.05
13.5


Asus RT-N18U
@800MHz (default) (http://wl500g.info/showthread.php?t=30444&p=269884#post269884)
1.59
6.70
5.29
0.05
13.6


Asus RT-N66U B1
@600MHz (default) (http://wl500g.info/showthread.php?t=30444&p=254322#post254322)
1.98
5.23
4.23
2.11
13.6


Asus RT-AC66U
@600MHz (default) (http://pastebin.com/F2Ek1Taj)
1.99
5.28
4.27
2.14
13.7


Asus RT-N16
@533MHz* (http://wl500g.info/showthread.php?p=136268#post136268)
2.25
5.63
5.13
2.38
15.4


Asus RT-N56U
@500MHz (default) (http://forum.ixbt.com/topic.cgi?id=14:56360-63#2475)
2.87
6.01
4.89
2.45
16.2


Asus RT-N65U
@500MHz (default) (http://wl500g.info/showthread.php?p=256026#post256026)
2.55
6.07
5.00
2.57
16.2


Asus RT-N14U
@580MHz (default) (http://wl500g.info/showthread.php?p=264021#post264021)
2.41
5.98
5.98
2.28
16.6


Asus RT-N15U
@500MHz (default) (http://wl500g.info/showthread.php?t=30444&p=254324#post254324)
2.82
6.36
5.13
2.48
16.8


Asus RT-AC58U
@717MHz (default) (http://wl500g.info/showthread.php?t=30444&p=273349#post273349)
1.99
7.76
7.19
0.05
17.0


Asus RT-N10U
@480MHz* (http://wl500g.info/showthread.php?t=30444&p=254939#post254939)
3.01
6.63
5.37
2.58
17.6


Asus RT-N16
@480MHz (default) (http://wl500g.info/showthread.php?t=30444&p=254324#post254324)
2.47
6.75
5.80
2.64
17.7


Asus RT-AC51U
@580MHz (default) (http://wl500g.info/showthread.php?t=30444&p=269884#post269884)
2.58
6.56
6.40
2.36
17.9


Netgear WNR3500L V2
@480MHz (default) (http://wl500g.info/showthread.php?t=30444&p=254326#post254326)
3.07
6.67
5.75
2.54
18.0


Netgear WNR3500L V1
@453MHz (default) (http://wl500g.info/showthread.php?p=254323#post254323)
2.72
7.12
6.27
2.79
18.9


TP-Link TL-MR3020
@400MHz (default) (http://wl500g.info/showthread.php?p=255608#post255608)
3.41
8.28
8.28
0.43
20.4


Raspberry Pi model B
@700MHz (default) (http://wl500g.info/showthread.php?p=255098#post255098)
2.67
8.83
8.86
0.64
21.0


ZyXEL Keenetic
@384MHz (default) (http://wl500g.info/showthread.php?t=30444&p=254532#post254532)
3.29
9.16
9.34
3.66
25.4


Asus RT-N10U
@300MHz (default) (http://wl500g.info/showthread.php?t=30444&p=254553#post254553)
3.91
10.62
8.63
4.17
27.3


Asus WL-500gP V1
@300MHz* (http://wl500g.info/showthread.php?t=30444&p=254321#post254321)
4.79
9.17
9.77
4.70
28.4


Asus WL-500gP V1
@264MHz (default) (http://wl500g.info/showthread.php?t=30444&p=254321#post254321)
5.31
10.42
11.10
5.36
32.2


D-Link DIR-320 A1
@240MHz (default) (http://wl500g.info/showthread.php?p=255642#post255642)
6.45
11.86
12.97
6.54
37.8


Asus WL-500gP V2
@240MHz (default) (http://wl500g.info/showthread.php?p=255103#post255103)
6.50
12.23
13.19
6.66
38.6


* - overclocked CPU

ryzhov_al
28-08-2012, 22:03
Asus RT-N66u

$ ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 31us
Time to run memory bench: 1.99[secs]
Time to run computation of pi (2400 digits, 10 times): 5.28[secs]
Time to run computation of e (9009 digits): 4.27[secs]
Time to run float bench: 2.08[secs]
Total time: 13.6s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2012-08-28 || ''Author'' || 2.0s || 5.3s || 4.3s || 2.1s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

ryzhov_al
28-08-2012, 22:04
D-Link DIR-320 ver.A1

$ ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 68us
Time to run memory bench: 6.32[secs]
Time to run computation of pi (2400 digits, 10 times): 11.88[secs]
Time to run computation of e (9009 digits): 14.62[secs]
Time to run float bench: 6.55[secs]
Total time: 39.4s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2012-08-28 || ''Author'' || 6.3s || 11.9s || 14.6s || 6.5s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

MercuryV
28-08-2012, 22:43
ASUS WL-500g Premium V1

$ ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 61us
Time to run memory bench: 5.31[secs]
Time to run computation of pi (2400 digits, 10 times): 10.42[secs]
Time to run computation of e (9009 digits): 11.10[secs]
Time to run float bench: 5.36[secs]
Total time: 32.2s

ASUS WL-500g Premium V1 @ 300 MHz

$ ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 53us
Time to run memory bench: 4.79[secs]
Time to run computation of pi (2400 digits, 10 times): 9.17[secs]
Time to run computation of e (9009 digits): 9.77[secs]
Time to run float bench: 4.70[secs]
Total time: 28.4s

skuwakin
28-08-2012, 23:03
root@RT-N66U:/tmp/mnt/sda3/ftp_pub/test# ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 24us
Time to run memory bench: 1.98[secs]
Time to run computation of pi (2400 digits, 10 times): 5.23[secs]
Time to run computation of e (9009 digits): 4.23[secs]
Time to run float bench: 2.11[secs]
Total time: 13.6s


странный какой-то тест... в памяти было полно процессов (снял только торрент), а начальные цифры чуток поменьше. чем во втором посте, а последняя цифирь больше....

MercuryV
28-08-2012, 23:11
Netgear WNR3500L V1

$ ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 33us
Time to run memory bench: 2.72[secs]
Time to run computation of pi (2400 digits, 10 times): 7.12[secs]
Time to run computation of e (9009 digits): 6.27[secs]
Time to run float bench: 2.79[secs]
Total time: 18.9s

Netgear WNR3500L V1 @ 480 Mhz

$ ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 32us
Time to run memory bench: 2.56[secs]
Time to run computation of pi (2400 digits, 10 times): 6.72[secs]
Time to run computation of e (9009 digits): 5.83[secs]
Time to run float bench: 2.62[secs]
Total time: 17.7s

Netgear WNR3500L V1 @ 500 Mhz

$ ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 31us
Time to run memory bench: 2.43[secs]
Time to run computation of pi (2400 digits, 10 times): 6.55[secs]
Time to run computation of e (9009 digits): 5.58[secs]
Time to run float bench: 2.52[secs]
Total time: 17.1s

theMIROn
28-08-2012, 23:38
RT-N16 @ 480 Mhz (default)

# ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 31us
Time to run memory bench: 2.47[secs]
Time to run computation of pi (2400 digits, 10 times): 6.75[secs]
Time to run computation of e (9009 digits): 5.80[secs]
Time to run float bench: 2.64[secs]
Total time: 17.7s

RT-N15U @ 500 Mhz (default)

# ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 29us
Time to run memory bench: 2.82[secs]
Time to run computation of pi (2400 digits, 10 times): 6.36[secs]
Time to run computation of e (9009 digits): 5.13[secs]
Time to run float bench: 2.48[secs]
Total time: 16.8s

Vampik
28-08-2012, 23:51
Netgear WNR3500L V1 @ 480 MHz


$ ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 31us
Time to run memory bench: 2.58[secs]
Time to run computation of pi (2400 digits, 10 times): 6.71[secs]
Time to run computation of e (9009 digits): 5.81[secs]
Time to run float bench: 2.61[secs]
Total time: 17.7s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2012-08-29 || ''Author'' || 2.6s || 6.7s || 5.8s || 2.6s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

Netgear WNR3500L V2 @ 480 MHz (default)


$ ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 31us
Time to run memory bench: 3.07[secs]
Time to run computation of pi (2400 digits, 10 times): 6.67[secs]
Time to run computation of e (9009 digits): 5.75[secs]
Time to run float bench: 2.54[secs]
Total time: 18.0s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2012-08-29 || ''Author'' || 3.1s || 6.7s || 5.8s || 2.5s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

AlexeyS
29-08-2012, 08:13
Asus WL500gp v1 @ 288 MHz



$ cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 52us
Time to run memory bench: 4.86[secs]
Time to run computation of pi (2400 digits, 10 times): 9.59[secs]
Time to run computation of e (9009 digits): 10.22[secs]
Time to run float bench: 4.75[secs]
Total time: 29.4s

QuickeneR2
29-08-2012, 08:55
У OpenWRT есть еще такая табличка замеров скорости шифрования
http://wiki.openwrt.org/inbox/benchmark.openssl
Это, конечно, еще не Мб/с через OpenVPN, но все равно любопытно

ryzhov_al
29-08-2012, 09:25
У OpenWRT есть еще такая табличка замеров скорости шифрования
http://wiki.openwrt.org/inbox/benchmark.openssl
Это, конечно, еще не Мб/с через OpenVPN, но все равно любопытноГодная ссылка, спасибо!
Померял у себя:) Получается, что RT-N66u в скорости (полностью софтового) шифрования уступает только Atom 330, хотя по bogomips'ам находится примерно в середине таблицы. Другими словами, bogomips'ам веры нет.

Zyxmon
04-09-2012, 11:12
ZyXEL keenetic (чип Ralink RT3052) mipsel, 384MHz


./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 42us
Time to run memory bench: 3.29[secs]
Time to run computation of pi (2400 digits, 10 times): 9.16[secs]
Time to run computation of e (9009 digits): 9.34[secs]
Time to run float bench: 3.66[secs]
Total time: 25.4s

theMIROn
05-09-2012, 15:18
RT-N10U @ 300 Mhz (default)

# ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 47us
Time to run memory bench: 3.91[secs]
Time to run computation of pi (2400 digits, 10 times): 10.62[secs]
Time to run computation of e (9009 digits): 8.63[secs]
Time to run float bench: 4.17[secs]
Total time: 27.3s

AkuNin
06-09-2012, 09:42
Netgear WNR3500L V1 @ 532 MHz

# ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 29us
Time to run memory bench: 2.27[secs]
Time to run computation of pi (2400 digits, 10 times): 6.09[secs]
Time to run computation of e (9009 digits): 5.24[secs]
Time to run float bench: 2.40[secs]
Total time: 16.0s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2012-09-06 || ''Author'' || 2.3s || 6.1s || 5.2s || 2.4s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

Omega
12-09-2012, 16:20
Можно также добавить и эти данные от Олега в таблицу :D


Производительность процессора:

RT-N16@533MHz

# /tmp/media/openwrt_cpu_bench_v06.bin
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 25us
Time to run memory bench: 2.25[secs]
Time to run computation of pi (2400 digits, 10 times): 5.63[secs]
Time to run computation of e (9009 digits): 5.13[secs]
Time to run float bench: 2.38[secs]
Total time: 15.4s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 1970-01-01 || ''Author'' || 2.3s || 5.6s || 5.1s || 2.4s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

WL500gp@300Mhz

This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 53us
Time to run memory bench: 4.71[secs]
Time to run computation of pi (2400 digits, 10 times): 9.18[secs]
Time to run computation of e (9009 digits): 10.08[secs]
Time to run float bench: 5.15[secs]
Total time: 29.1s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2009-03-10 || ''Author'' || 4.7s || 9.2s || 10.1s || 5.1s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

Т.е. перемалывает данные он примерно в 2 раза быстрее

А вот здесь у Aven'a тоже куча данных: :) http://digiland.tw/viewtopic.php?id=409&p=1


Router: RT-N16 (532Mhz) BogoMIPS: 264.19 Firmware: 1.0.0.6

Overhead for getting time: 26us
Time to run memory bench: 2.28[secs]
Time to run computation of pi (2400 digits, 10 times): 5.75[secs]
Time to run computation of e (9009 digits): 5.23[secs]
Time to run float bench: 2.41[secs]
Total time: 15.7s
Router: RT-N16 (480Mhz) BogoMIPS: 238.59 Firmware: 9.74.0.1

Overhead for getting time: 28us
Time to run memory bench: 2.58[secs]
Time to run computation of pi (2400 digits, 10 times): 6.46[secs]
Time to run computation of e (9009 digits): 5.78[secs]
Time to run float bench: 2.69[secs]
Total time: 17.5s
Router: RT-N16 (453Mhz) BogoMIPS: 225.28 Firmware: 1.0.0.6

Overhead for getting time: 30us
Time to run memory bench: 2.79[secs]
Time to run computation of pi (2400 digits, 10 times): 7.03[secs]
Time to run computation of e (9009 digits): 6.41[secs]
Time to run float bench: 2.96[secs]
Total time: 19.2s
Router: RT-N56U CPU: Ralink RT3662 500Mhz (MIPS 74K) Flash: 8M RAM: 128MB BogoMIPS: 249.34 Firmware: 9.0.6.4

Overhead for getting time: 27us
Time to run memory bench: 5.52[secs]
Time to run computation of pi (2400 digits, 10 times): 5.92[secs]
Time to run computation of e (9009 digits): 5.40[secs]
Time to run float bench: 2.58[secs]
Total time: 19.4s
Router: RT-N13U CPU: Ralink RT3052F (MIPS 24K) RAM: 32M BogoMIPS: 255.48 Firmware: 1.0.0.9

Overhead for getting time: 33us
Time to run memory bench: 3.72[secs]
Time to run computation of pi (2400 digits, 10 times): 9.43[secs]
Time to run computation of e (9009 digits): 9.99[secs]
Time to run float bench: 3.71[secs]
Total time: 26.9s
Router: RT-N10 CPU: Broadcom BCM5356 300MHz Flash: 4MB RAM: 16MB

Overhead for getting time: 96us
Time to run memory bench: 4.53[secs]
Time to run computation of pi (2400 digits, 10 times): 9.93[secs]
Time to run computation of e (9009 digits): 8.65[secs]
Time to run float bench: 4.08[secs]
Total time: 27.2s
Router: RT-N10+ CPU: Ralink RT3050F 320MHz Flash: 4MB RAM: 32MB

Overhead for getting time: 44us
Time to run memory bench: 5.01[secs]
Time to run computation of pi (2400 digits, 10 times): 11.22[secs]
Time to run computation of e (9009 digits): 11.92[secs]
Time to run float bench: 4.25[secs]
Total time: 32.4s
Router: WL-500W

Overhead for getting time: 60us
Time to run memory bench: 5.59[secs]
Time to run computation of pi (2400 digits, 10 times): 10.78[secs]
Time to run computation of e (9009 digits): 11.87[secs]
Time to run float bench: 5.99[secs]
Total time: 34.2s
Router: WL-500gP v2 Firmware: oleg 1.9.2.7-10

Overhead for getting time: 84us
Time to run memory bench: 6.79[secs]
Time to run computation of pi (2400 digits, 10 times): 11.91[secs]
Time to run computation of e (9009 digits): 13.30[secs]
Time to run float bench: 7.51[secs]
Total time: 39.5s
Router: WL-520gU Firmware: oleg 1.9.2.7-10

Overhead for getting time: 84us
Time to run memory bench: 7.38[secs]
Time to run computation of pi (2400 digits, 10 times): 12.05[secs]
Time to run computation of e (9009 digits): 13.54[secs]
Time to run float bench: 7.54[secs]
Total time: 40.5s
Router: WL500gx

Overhead for getting time: 108us
Time to run memory bench: 8.55[secs]
Time to run computation of pi (2400 digits, 10 times): 14.55[secs]
Time to run computation of e (9009 digits): 16.32[secs]
Time to run float bench: 9.13[secs]
Total time: 48.6s
Router: WL-HDD Firmware: oleg 1.9.2.7-9

Overhead for getting time: 139us
Time to run memory bench: 10.13[secs]
Time to run computation of pi (2400 digits, 10 times): 24.17[secs]
Time to run computation of e (9009 digits): 29.57[secs]
Time to run float bench: 13.31[secs]
Total time: 77.2s
Router: WL-500b Firmware: oleg 1.9.2.7-9

Overhead for getting time: 139us
Time to run memory bench: 10.12[secs]
Time to run computation of pi (2400 digits, 10 times): 24.19[secs]
Time to run computation of e (9009 digits): 29.62[secs]
Time to run float bench: 13.31[secs]
Total time: 77.2s
Router: WL-500g Firmware: oleg 1.9.2.7-8

Overhead for getting time: 151us
Time to run memory bench: 10.32[secs]
Time to run computation of pi (2400 digits, 10 times): 25.05[secs]
Time to run computation of e (9009 digits): 30.30[secs]
Time to run float bench: 15.81[secs]
Total time: 81.5s
Тестируют ещё и так : http://wpte.kicks-ass.net/index.php/2010/08/07/benchmarking/ ;)

З.Ы. http://code.google.com/p/wl500g-repo/source/browse/feeds/rtndev/cpubench/Makefile?spec=svn894&r=894 :)

theMIROn
19-09-2012, 10:01
RT-N10U @ 480 Mhz

# ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 31us
Time to run memory bench: 3.01[secs]
Time to run computation of pi (2400 digits, 10 times): 6.63[secs]
Time to run computation of e (9009 digits): 5.37[secs]
Time to run float bench: 2.58[secs]
Total time: 17.6s

don-pedro
25-09-2012, 07:43
Raspberry Pi model B @ 700MHz :-)

pi@raspberrypi ~ $ ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 1us
Time to run memory bench: 2.67[secs]
Time to run computation of pi (2400 digits, 10 times): 8.83[secs]
Time to run computation of e (9009 digits): 8.86[secs]
Time to run float bench: 0.64[secs]
Total time: 21.0s

ryzhov_al
25-09-2012, 07:57
Можно также добавить и эти данные от Олега в таблицу :D

А вот здесь у Aven'a тоже куча данных: :) http://neture.homeip.net/digiland/viewtopic.php?id=409&p=1Не можно. В таблице только те данные, что мы сами собрали на одном бинарнике бенчмарка из первого поста. Для этого и привожу ссылки на источник.

Zyxel Keenetic, ASUS RT-N56U и Raspberri Pi приведены исключительно для справки.


Raspberry Pi model B @ 700MHz :-)Спасибо! Обновил табличку.

uip88
25-09-2012, 10:42
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 68us
Time to run memory bench: 6.50[secs]
Time to run computation of pi (2400 digits, 10 times): 12.23[secs]
Time to run computation of e (9009 digits): 13.19[secs]
Time to run float bench: 6.66[secs]
Total time: 38.6s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2012-09-25 || ''Author'' || 6.5s || 12.2s || 13.2s || 6.7s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

ryzhov_al
01-10-2012, 11:40
# ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time...
(typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 46us
Time to run memory bench: 3.82[secs]
Time to run computation of pi (2400 digits, 10 times): 10.65[secs]
Time to run computation of e (9009 digits): 8.76[secs]
Time to run float bench: 4.19[secs]
Total time: 27.4s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2012-10-01 || ''Author'' || 3.8s || 10.6s || 8.8s || 4.2s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

ryzhov_al
18-10-2012, 09:32
root@OpenWrt:/tmp# ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 11us
Time to run memory bench: 3.41[secs]
Time to run computation of pi (2400 digits, 10 times): 8.28[secs]
Time to run computation of e (9009 digits): 8.28[secs]
Time to run float bench: 0.43[secs]
Total time: 20.4s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2012-10-18 || ''Author'' || 3.4s || 8.3s || 8.3s || 0.4s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||
Чуть подробнее (http://wl500g.info/entry.php?2929-TP-Links-TL-MR3020):

oot@OpenWrt:/tmp# uname -a
Linux OpenWrt 3.3.8 #1 Thu Oct 11 11:12:22 MSK 2012 mips GNU/Linux

Virtuals
19-10-2012, 11:01
dir320 A1G
Linux WL-00265A88D06F 2.6.22.19 #2 Sun May 6 05:36:11 MSK 2012 mips GNU/Linux


./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 71us
Time to run memory bench: 6.45[secs]
Time to run computation of pi (2400 digits, 10 times): 11.86[secs]
Time to run computation of e (9009 digits): 12.97[secs]
Time to run float bench: 6.54[secs]
Total time: 37.8s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2012-10-19 || ''Author'' || 6.4s || 11.9s || 13.0s || 6.5s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

harsh132
01-11-2012, 10:40
admin@RT-N65U:/tmp/opt# cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 28us
Time to run memory bench: 2.55[secs]
Time to run computation of pi (2400 digits, 10 times): 6.07[secs]
Time to run computation of e (9009 digits): 5.00[secs]
Time to run float bench: 2.57[secs]
Total time: 16.2s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2012-10-31 || ''Author'' || 2.6s || 6.1s || 5.0s || 2.6s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

Большое спасибо ryzhov_al за потраченное время и помощь)))

ryzhov_al
28-01-2013, 13:23
NAS WD My Book Live, камень PowerPC @800MHz, тоже RISC:

MyBookLive:/DataVolume/opt/tmp# ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 1us
Time to run memory bench: 1.93[secs]
Time to run computation of pi (2400 digits, 10 times): 4.28[secs]
Time to run computation of e (9009 digits): 6.03[secs]
Time to run float bench: 0.02[secs]
Total time: 12.3s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2013-01-28 || ''Author'' || 1.9s || 4.3s || 6.0s || 0.0s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

Оставлю здесь для истории, в таблицу, наверное, вносить не буду.

staticroute
03-02-2013, 01:06
Может пригодится Netgear WNDR3800 on OpenWRT 12.09RC1

CPU 680Mhz, 128Mb RAM


system type : Atheros AR7161 rev 2
machine : NETGEAR WNDR3700/WNDR3800/WNDRMAC
processor : 0
cpu model : MIPS 24Kc V7.4
BogoMIPS : 452.19
wait instruction : yes
microsecond timers : yes
tlb_entries : 16
extra interrupt vector : yes
hardware watchpoint : yes, count: 4, address/irw mask: [0x0000, 0x0ff8, 0x0ff8, 0x0ff8]
ASEs implemented : mips16
shadow register sets : 1
kscratch registers : 0
core : 0
VCED exceptions : not available
VCEI exceptions : not available



OpenWrt:tmp $ ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 7us
Time to run memory bench: 1.85[secs]
Time to run computation of pi (2400 digits, 10 times): 5.08[secs]
Time to run computation of e (9009 digits): 5.09[secs]
Time to run float bench: 0.25[secs]
Total time: 12.3s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2013-02-03 || ''Author'' || 1.9s || 5.1s || 5.1s || 0.3s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||


OpenWrt:tmp $ uname -a
Linux OpenWrt 3.3.8 #1 Sun Nov 18 04:31:55 UTC 2012 mips GNU/Linux

cpubench пришлось собрать в openwrt, так как ваш байнари не запустился.

Еще можно погнать до 800Мгц и потестить, может попозже выложу резалт.

Megagloom
08-02-2013, 11:18
TPLINK TL-WDR4300 v1
CPU Atheros AR9344 560 MHz, 128Mb RAM

root@Router:/tmp# ./cpubench.atheros
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 9us
Time to run memory bench: 2.09[secs]
Time to run computation of pi (2400 digits, 10 times): 5.35[secs]
Time to run computation of e (9009 digits): 4.62[secs]
Time to run float bench: 0.32[secs]
Total time: 12.4s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2013-02-08 || ''Author'' || 2.1s || 5.3s || 4.6s || 0.3s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||
бинарник брал здесь (http://neture.homeip.net/digiland/viewtopic.php?id=409&p=1)

ryzhov_al
19-04-2013, 17:22
Добавил в таблицу результаты (http://forum.ixbt.com/topic.cgi?id=14:59030-26#945) бенчмарка нового SoC'а RT6856 (Zyxel Keenetic II Giga), который из-за наличия кучи конвееров видится (http://forum.ixbt.com/topic.cgi?id=14:59030-26#928) в системе как четырёхядерный:)

Спасибо McMCC.

slaiver
20-04-2013, 09:49
[admin@router root]$ cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 28us
Time to run memory bench: 2.75[secs]
Time to run computation of pi (2400 digits, 10 times): 7.84[secs]
Time to run computation of e (9009 digits): 7.99[secs]
Time to run float bench: 2.28[secs]
Total time: 20.9s

QuickeneR2
14-05-2013, 14:17
Подтверждаю результат Megagloom выше, бинарник брал по его ссылке


# ./cpubench.atheros
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 9us
Time to run memory bench: 2.10[secs]
Time to run computation of pi (2400 digits, 10 times): 5.32[secs]
Time to run computation of e (9009 digits): 4.40[secs]
Time to run float bench: 0.32[secs]
Total time: 12.1s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2011-09-08 || ''Author'' || 2.1s || 5.3s || 4.4s || 0.3s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||
Результаты последовательных замеров 13.7 12.1 16.1 15.0 12.1 12.1 13.1 12.1 13.9 12.1 12.1 (видимо, была какая-то фоновая активность), так что взял 12.1 как минимально стабильный

QuickeneR2
29-05-2013, 14:33
Приехал Ubiquiti Edge Router Lite - могу померять скорость, если научите правильно компилировать. Первая попытка (apt-get install gcc; gcc -O0 openwrt_cpu_bench_v06.c) провалилась - выдала 32 секунды, на уровне 500gP, хотя органолептически он куда быстрее.

ryzhov_al
20-06-2013, 16:37
Asus RT-AC56U @800MHz (default):

./cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 1us
Time to run memory bench: 1.62[secs]
Time to run computation of pi (2400 digits, 10 times): 6.64[secs]
Time to run computation of e (9009 digits): 5.22[secs]
Time to run float bench: 0.05[secs]
Total time: 13.5s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2013-06-02 || ''Author'' || 1.6s || 6.6s || 5.2s || 0.0s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

Asus RT-AC56U @1000MHz (overclocked):

This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 1us
Time to run memory bench: 1.51[secs]
Time to run computation of pi (2400 digits, 10 times): 5.32[secs]
Time to run computation of e (9009 digits): 4.19[secs]
Time to run float bench: 0.04[secs]
Total time: 11.1s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2013-06-02 || ''Author'' || 1.5s || 5.3s || 4.2s || 0.0s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||
О чём это говорит? Скорее всего, ни о чём. Вряд ли тест использует многоядерные возможности.

Процессор:

# cat /proc/cpuinfo
Processor : ARMv7 Processor rev 0 (v7l)
processor : 0
BogoMIPS : 1599.07

processor : 1
BogoMIPS : 1595.80

Features : swp half thumb fastmult edsp
CPU implementer : 0x41
CPU architecture: 7
CPU variant : 0x3
CPU part : 0xc09
CPU revision : 0

Hardware : Northstar Prototype
Revision : 0000
Serial : 0000000000000000

Тулчейн:

arm-brcm-linux-uclibcgnueabi-cc -v
Using built-in specs.
COLLECT_GCC=arm-brcm-linux-uclibcgnueabi-cc
COLLECT_LTO_WRAPPER=/root/asuswrt-arm/release/src-rt-6.x.4708/toolchains/hndtools-arm-linux-2.6.36-uclibc-4.5.3/bin/../libexec/gcc/arm-brcm-linux-uclibcgnueabi/4.5.3/lto-wrapper
Target: arm-brcm-linux-uclibcgnueabi
Configured with: /projects/toolchains_build/buildroot-2012.02-brcm/output/toolchain/gcc-4.5.3/configure --prefix=/projects/hnd/tools/linux/hndtools-arm-linux-2.6.36-uclibc-4.5.3 --build=i686-pc-linux-gnu --host=i686-pc-linux-gnu --target=arm-brcm-linux-uclibcgnueabi --enable-languages=c,c++ --with-sysroot=/projects/hnd/tools/linux/hndtools-arm-linux-2.6.36-uclibc-4.5.3/arm-brcm-linux-uclibcgnueabi/sysroot --with-build-time-tools=/projects/hnd/tools/linux/hndtools-arm-linux-2.6.36-uclibc-4.5.3/arm-brcm-linux-uclibcgnueabi/bin --disable-__cxa_atexit --enable-target-optspace --disable-libgomp --with-gnu-ld --disable-libssp --disable-multilib --disable-tls --enable-shared --with-gmp=/projects/hnd/tools/linux/hndtools-arm-linux-2.6.36-uclibc-4.5.3 --with-mpfr=/projects/hnd/tools/linux/hndtools-arm-linux-2.6.36-uclibc-4.5.3 --with-mpc=/projects/hnd/tools/linux/hndtools-arm-linux-2.6.36-uclibc-4.5.3 --enable-threads --disable-decimal-float --with-float=soft --with-abi=aapcs-linux --with-arch=armv7-a --with-tune=cortex-a9 --with-pkgversion='Buildroot 2012.02' --with-bugurl=http://bugs.buildroot.net/
Thread model: posix
gcc version 4.5.3 (Buildroot 2012.02)

Omega
02-09-2013, 23:45
Asus RT-AC56U @800MHz (default):

What can you expect out of an ARM A9 dual-core CPU?

OpenVPN throughput benchmark:
iperf -c 10.16.0.1 -M 1400 -N -l 64K -t 30

The iperf and OpenVPN servers were on the router, and clients were a Win7 laptop connected to the WAN port.


RT-AC66U (FW 3.0.0.4.270.25):
AES-128-CBC [152] 0.0-30.0 sec 79.5 MBytes 22.2 Mbits/sec

RT-AC56U
AES_128-CBC [156] 0.0-30.0 sec 217 MBytes 60.7 Mbits/sec
For OpenVPN, this router is a beast :)

Here's the raw OpenSSL performance. openssl speed test:


RT-AC66U:
type 16 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
aes-128 cbc 9398.56k 9941.17k 10138.71k 10195.09k 10234.54k

RT-AC56U (with ARM ASM optimizations backported from OpenSSL 1.0.1):
type 16 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
aes-128 cbc 25596.63k 28151.82k 29251.64k 29363.88k 29587.09k
That's right. OpenSSL raw performance is nearly 3x faster on the RT-AC56U.
http://forums.smallnetbuilder.com/showthread.php?t=11380 :D

Asus RT-AC56U @1200MHz (overclocked):


nvram set clkfreq=1200,800
nvram commit && reboot

# CPU

CPU Model ARMv7 Processor rev 0 (v7l) - (Cores: 2)
CPU Frequency 1200 MHz
CPU Load Average (1, 5, 15 mins) 0.09, 0.07, 0.05

# cat /proc/cpuinfo

Processor : ARMv7 Processor rev 0 (v7l)
processor : 0
BogoMIPS : 2398.61

processor : 1
BogoMIPS : 2398.61

Features : swp half thumb fastmult edsp
CPU implementer : 0x41
CPU architecture: 7
CPU variant : 0x3
CPU part : 0xc09
CPU revision : 0

Hardware : Northstar Prototype
Revision : 0000
Serial : 0000000000000000
4. openssl speed:

# openssl speed aes-256-cbc

Doing aes-256 cbc for 3s on 16 size blocks: 5298660 aes-256 cbc's in 2.98s
Doing aes-256 cbc for 3s on 64 size blocks: 1473282 aes-256 cbc's in 3.01s
Doing aes-256 cbc for 3s on 256 size blocks: 380765 aes-256 cbc's in 3.00s
Doing aes-256 cbc for 3s on 1024 size blocks: 95997 aes-256 cbc's in 3.00s
Doing aes-256 cbc for 3s on 8192 size blocks: 12042 aes-256 cbc's in 3.00s

OpenSSL 1.0.0j 10 May 2012 built on: Sat Aug 24 22:13:38 EDT 2013
options:bn(64,32) rc4(ptr,char) des(idx,cisc,16,long) aes(partial) idea(int) blowfish(ptr)
compiler: arm-brcm-linux-uclibcgnueabi-gcc -fPIC -DOPENSSL_PIC -DOPENSSL_THREADS -D_REENTRANT -DDSO_DLFCN -DHAVE_DLFCN_H -ffunction
-sections -fdata-sections -DTERMIO -O3 -Wall -DOPENSSL_BN_ASM_MONT -DSHA1_ASM -DSHA256_ASM -DSHA512_ASM -DAES_ASM -DGHASH_ASM

The 'numbers' are in 1000s of bytes per second processed.
type 16 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
aes-256 cbc 28449.18k 31325.60k 32491.95k 32766.98k 32882.69k
http://wl500g.info/showthread.php?30722-Asus-RT-AC56U-amp-RT-AC68U-802-11ac-Gigabit-DualBand-Wireless-Router&p=263212#post263212 :cool:

MercuryV
21-11-2013, 00:39
Overhead for getting time: 26us
Time to run memory bench: 2.41[secs]
Time to run computation of pi (2400 digits, 10 times): 5.98[secs]
Time to run computation of e (9009 digits): 5.98[secs]
Time to run float bench: 2.28[secs]
Total time: 16.6s

Omega
26-03-2014, 00:51
CPU: AMCC PowerPC APM82181 Rev.E at 1000 MHz (PLB=166, OPB=83, EBC=83 MHz)
Security support Bootstrap Option D - Boot
ROM Location NAND wo/ECC 2k page (8 bits),
booting from NAND 32 kB I-Cache 32 kB D-Cache
DNI HW ID: 29763875 NAND flash 128MB RAM 256MB U-Boot 2009.08 dni1 V1.4
Board: AMCC APM82181 Evaluation Board, PCIE0/SATA1, 1*USB OTG
I2C: ready
DRAM: 256 MB (ECC not enabled, 500 Mb/s, 32-bit, CL3)
NAND: 128 MiB
http://www.wikidevi.com/wiki/Netgear_WNDR4700

Test and Performance

Note: All benchmarks and test were carried out on the original firmware.
The results would be much better in the current branch of OpenWrt.

Benchmark: OpenSSL v0.9.8p Note: OpenSSL v0.9.8p from SDK

OpenSSL MD5 SHA-1 SHA-256 SHA-512 DES 3DES AES-128 AES-192 AES-256 RSA Sign RSA Verify DSA Sign DSA Verify
Version

0.9.8p 87764520 23929930 18292650 10642240 10366780 3586110 18779740 16537780 14734770 10.9 408.0 41.2 33.9

Benchmark: Cpubench v0.6 Note: Source code here.
https://code.google.com/p/wl500g-repo/source/browse/feeds/rtndev/cpubench/?r=1146#cpubench

[WNDR4700]# ./cpubench
This is CPU and memory benchmark for OpenWRT v0.6.
This will then take some time... (typically 30-60 seconds on a 200MHz computer)

Overhead for getting time: 1us
Time to run memory bench: 1.65[secs]
Time to run computation of pi (2400 digits, 10 times): 4.16[secs]
Time to run computation of e (9009 digits): 5.10[secs]
Time to run float bench: 0.02[secs]
Total time: 10.9s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2013-08-27 || ''Author'' || 1.6s || 4.2s || 5.1s || 0.0s || v0.6 || ''OS'' |
| ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||
Note: Tests were performed on non-optimized software and built by hand, using only the sole and conclusive construction environment what is available
at the moment. Manufacturer's SDK using the native libraries-definitely not reflect actual achievements and possibilities of this equipment.

http://wiki.openwrt.org/toh/netgear/wndr4700

Rozov
08-05-2014, 11:10
Друзья, подскажите, может кто натыкался на сравнение RT-N16 и RT-N66U.

Интересует сравнение производительности процессоров. Планирую прикупить RT-N66U т.к. RT-N16 не справляется с мультикаст потоком HD каналов в режиме IPTV сервера.

Посоветуйте стоит ли игра свеч?

260db
19-05-2014, 10:41
D-Link DIR-620 rev C1
Прошивка "от vampik" версия "3.0.6-r5066M".

@ CPU Clock = 530MHz (default); Memory Clock = 167Mhz (default)


$ cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 27us
Time to run memory bench: 2.72[secs]
Time to run computation of pi (2400 digits, 10 times): 6.02[secs]
Time to run computation of e (9009 digits): 4.91[secs]
Time to run float bench: 2.28[secs]
Total time: 15.9s

P.S. Такой-же роутер указанный в таблице почему-то работает на 5сек дольше, видимо зависит от прошивки...

@ CPU Clock = 530MHz (default); Memory Clock = 212Mhz (overclock)


$ cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 27us
Time to run memory bench: 2.49[secs]
Time to run computation of pi (2400 digits, 10 times): 5.97[secs]
Time to run computation of e (9009 digits): 4.90[secs]
Time to run float bench: 2.28[secs]
Total time: 15.6s


Если кто подскажет, как разогнать процессор больше 530MHz, буду очень благодарен :)

ryzhov_al
19-05-2014, 11:36
Спасибо, добавил в таблицу.

Признаться, два года назад эта табличка создавалась для того, чтобы потролить определённых товарищей на хоботе и в этом качестве она отработала на 200%.

Дальше пополнять таблицу в первом посте не стану, если она будет редактироваться, то только силами модератора(ов).

ryzhov_al
14-06-2014, 18:49
До кучи, разогнанный RT-N66U

admin@RT-N66U:/tmp/home/root# cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 22us
Time to run memory bench: 1.84[secs]
Time to run computation of pi (2400 digits, 10 times): 4.77[secs]
Time to run computation of e (9009 digits): 3.85[secs]
Time to run float bench: 1.88[secs]
Total time: 12.3s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt.org/HardwarePerformance
|| 2014-06-14 || ''Author'' || 1.8s || 4.8s || 3.8s || 1.9s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||
admin@RT-N66U:/tmp/home/root# nvram get clkfreq
662,331,165

hggomes
15-06-2014, 04:28
RT-AC68U@1600MHZ

admin@proxy:/tmp/home/root# nvram get clkfreq
1600,800

This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 1us
Time to run memory bench: 1.19[secs]
Time to run computation of pi (2400 digits, 10 times): 3.29[secs]
Time to run computation of e (9009 digits): 2.83[secs]
Time to run float bench: 0.03[secs]
Total time: 7.3s

ryzhov_al
15-06-2014, 09:24
RT-AC68U@1600MHZVery impressive.
Added to first post.

Markov
23-06-2014, 10:21
Подскажите как запустить этот бенчмарк, есть на руках Archer C7, хочу выложить его тест.

don-pedro
24-06-2014, 14:06
Подскажите как запустить этот бенчмарк, есть на руках Archer C7, хочу выложить его тест.
Например, так:
1) Установить на роутер OpenWrt
2) скомпилировать под ваш роутер cpubench
3) запустить cpubench
:)

Стоит ли всё это делать ради того, чтоб померяться бенчмарками, решать вам. Всё равно потом родную прошивку возвращать - OpenWrt ещё не совсем дозрело да вашей железки.

vleonv
29-10-2015, 21:53
/opt/home/admin # cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 17us
Time to run memory bench: 2.14[secs]
Time to run computation of pi (2400 digits, 10 times): 3.69[secs]
Time to run computation of e (9009 digits): 3.69[secs]
Time to run float bench: 1.47[secs]
Total time: 11.0s

Omega
03-01-2016, 06:36
ASUS RT-N18U (https://wikidevi.com/wiki/ASUS_RT-N18U) (Forum thread (http://4pda.ru/forum/index.php?showtopic=611325))
Broadcom BCM47081A0 ARM (800MHz)
128MB Flash, 256MB DDR3 RAM, USB3.0

Overhead for getting time: 2us
Time to run memory bench: 1.66[secs]
Time to run computation of pi (2400 digits, 10 times): 6.91[secs]
Time to run computation of e (9009 digits): 5.31[secs]
Time to run float bench: 0.05[secs]
Total time: 13.9s

D-Link DIR-868L (https://wikidevi.com/wiki/D-Link_DIR-868L_rev_A1)
Broadcom BCM47081A0 ARM (800MHz)
128MB Flash, 128MB DDR3 RAM, USB3.0

Overhead for getting time: 2us
Time to run memory bench: 1.71[secs]
Time to run computation of pi (2400 digits, 10 times): 6.62[secs]
Time to run computation of e (9009 digits): 5.37[secs]
Time to run float bench: 0.05[secs]
Total time: 13.7s

Xiaomi MiWiFi
Broadcom BCM4709 ARM (1GHz)
16MB Flash, 256MB DDR3 RAM, 1TB HDD

Overhead for getting time: 1us
Time to run memory bench: 1.17[secs]
Time to run computation of pi (2400 digits, 10 times): 5.31[secs]
Time to run computation of e (9009 digits): 4.31[secs]
Time to run float bench: 0.04[secs]
Total time: 10.8s

http://digiland.tw/viewtopic.php?id=409&p=5 :rolleyes:

Omega
03-10-2016, 00:08
• Asus RT-N18U (FW 380.57.7 HGG-FINAL)

. . http://digiland.tw/uploads/9152_aea.png

• ASUS RT-AC51U

Overhead for getting time: 23us
Time to run memory bench: '2.58'[secs]
Time to run computation of pi (2400 digits, 10 times): '6.56'[secs]
Time to run computation of e (9009 digits): '6.40'[secs]
Time to run float bench: '2.36'[secs]
Total time: '17.9's

Source: http://digiland.tw/viewtopic.php?id=409&p=5 ;)

Zyxmon
05-10-2016, 08:37
Можно использовать Entware для оценки производительности.
Вычисление числа pi не использует ядра/потоки, если их несколько.
Встроенный бенчмарк p7zip - использует.

Методика и результаты - http://forums.zyxmon.org/viewtopic.php?t=5382
Использовались роутеры, НАС'ы, мультимедиа плееры.

Вычисления в 1 поток:


QNAP 231+ (ARMv7 1.4GHz): ______________________ 66.61 user 0.00 system 1:07.00 elapsed 99%CPU
Tronsmart CX919 (RK3188T Cortex A9 1.4GHz): ____ 80.44 user 0.02 system 1:20.61 elapsed 99%CPU
QNAP 121 (Feroceon 88F6282 (armv5l) 2GHz): _____ 84.61 user 0.01 system 1:26.13 elapsed 98%CPU
QNAP 269 (Intel Atom D2701 2.13GHz): ___________ 87.18 user 0.08 system 1:27.28 elapsed 99%CPU
Keenetic Ultra II (MT7621AT MIPS 880MHz): ______ 208.64 user 0.34 system 3:29.05 elapsed 99%CPU
IconBit (Realtek RTD1186DD MIPS 700MHz): _______ 213.47 user 0.10 system 4:54.75 elapsed 72%CPU
Keenetic Ultra (RT6856 MIPS 700MHz): ___________ 262.76 user 1.80 system 4:26.13 elapsed 99%CPU
Keenetic II (RT6856 MIPS 700MHz): ______________ 265.55 user 0.05 system 4:26.10 elapsed 99%CPU
Keenetic Giga II (RT6856 MIPS 34K 700MHz): _____ 268.83 user 1.38 system 4:33.27 elapsed 98%CPU
Keenetic DSL (RT63368 MIPS 34K 700MHz): ________ 269.16 user 0.71 system 4:30.03 elapsed 99%CPU
Keenetic Omni II (MT7620N MIPS 24KEc 580MHz): __ 315.42 user 0.16 system 5:21.63 elapsed 98%CPU
Keenetic Extra (MT7620A MIPS24Kc 580MHz): ______ 317.09 user 0.38 system 5:25.82 elapsed 97%CPU

"Временной" победитель - 4х ядерный ARMv7 с частотой 1.4GHz на НАС'е QNAP.
Он в 1.5 раза обошел даже 2х ядерный Intel Atom 2.13GHz.

Вне конкурса участвовал Intel® Core™ i5-3470S CPU 2.90GHz

• real 0m19.387s
• user 0m19.388s
• sys 0m0.000s
Вычисление с использованием всех потоков, ядер:

Первая цифра - средняя загрузка CPU (значение 400 означает, что 4 ядра были загружены на 100%,
вторая цифра - значение на 1 ядро и последняя - на все ядра (суммарная производительность).
Сравниваем по последней цифре:


________________________________________________ CPU% /CPU TOTL
QNAP 269 (Intel Atom D2701 2.13GHz): ____________ 343 875 3043
QNAP 231+ (ARMv7 1.4GHz): _______________________ 194 1251 2430
Tronsmart CX919 (RK3188T Cortex A9 1.4GHz): _____ 336 651 2270
Keenetic Ultra II (MT7621AT MIPS 880MHz) : ______ 334 283 972
QNAP 121 (Feroceon 88F6282 (armv5l) 2GHz): ______ 98 750 734
Keenetic DSL (RT63368 MIPS 34K 700MHz): _________ 324 123 407
Keenetic Ultra (RT6856 MIPS 700MHz): ____________ 100 332 332
Keenetic Giga II (RT6856 MIPS 34K 700MHz): ______ 100 324 323
Keenetic Omni II (MT7620N MIPS 24KEc 580MHz): ___ 98 261 257
Keenetic Extra (MT7620A MIPS24Kc 580MHz): _______ 98 259 253

Omega
27-10-2016, 06:09
ryzhov_al
Можно будет составить другую таблицу ;)

• OpenVPN throughput benchmark:

# iperf -c 10.1.0.1 -M 1400 -N -l 64K -t 30
• OpenSSL performance speed test:

# openssl speed aes-128-cbc
# openssl speed aes-256-cbc

• MT7621A x4 threads
# openssl speed aes-128-cbc -elapsed -multi 4
# openssl speed aes-256-cbc -elapsed -multi 4

Benchmark: OpenSSL (OpenWRT) (https://wiki.openwrt.org/doc/howto/benchmark.openssl)

__________________________________________________ __________________________________________________ __________________________

Device | SoC | CPU | BogoMIPS| Version| MD5 | SHA~1 | SHA~256 | SHA~512
__________________________________________________ __________________________________________________ __________________________

Asus WL~500gPv1 | Broadcom BCM4704 @264MHz (BCM3302 V0.6) | 262.96 | 0.9.8e | 10649260 | 4264960 | 2501970 | 1203810
Asus WL~500gPv2 | Broadcom BCM5354 @240MHz (BCM3302 V2.9) | 239.10 | 0.9.8t | 10680320 | 3528210 | 2177780 | 1442780
Asus RT~N16 | Broadcom BCM4716 @480MHz (MIPS 74Kc V4.0) | 239.61 | 1.0.2e | 30531640 | 23920980 | 12333740 | 3836930
Asus RT~N56U A1 | Ralink RT3662F @500MHz (MIPS 74Kc V4.12) | 249.44 | 1.0.1i | 31172950 | 17290870 | 8386560 | 3905540
Asus RT~N56U B1 | MediaTek MT7621S @900MHz (MIPS 1004Kc V2.15) | 598.01 | 1.0.2d | 45868030 | 30101820 | 13991810 | 5380620
Asus RT~AC68U S | Broadcom BCM4708 @800MHz (ARMv7 rev 0 (v7l)) | 1594.16 | 1.0.2d | 77319820 | 20030390 | 16236840 | 6731270

WiTi Board | MediaTek MT7621A @880MHz (MIPS 1004Kc V2.15) | 586.13 | 1.0.2d | 45192190 | 39567150 | 20150160 | 6286730
WiTi x4 thr | MediaTek MT7621A @880MHz (MIPS 1004Kc V2.15) | 586.13 | 1.0.2d |108650150 | 80841390 | 40218280 |12650840
MikroTik RB750Gr3 | MediaTek MT7621A 1.2GHz (MIPS 1004Kc V2.15) | 796.67 | 1.0.2j | 63220740 | 44255230 | 21024090 | 6559060
MikroTik RB750Gr3 | MediaTek MT7621A x4 thr (MIPS 1004Kc V2.15) | 796.67 | 1.0.2j |130489340 | 90270720 | 42432850 |13103100
__________________________________________________ __________________________________________________ __________________________

Type 16 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
__________________________________________________ ______________________________

•'RT-N56U B1 MT7621S' @900MHz (overclocked)
aes-128 cbc 10891.69k 11434.13k 11575.65k 11624.11k 11643.56k
aes-256 cbc 8375.04k 8714.23k 8784.65k 8793.09k 8809.13k

md5 3218.11k 10591.36k 27489.26k 45838.34k 57035.43k
sha1 3132.05k 9257.71k 20830.09k 30095.70k 34820.11k
des cbc 6338.92k 6587.80k 6659.94k 6679.21k 6684.67k
des ede3 2317.51k 2352.15k 2371.04k 2366.16k 2372.67k
aes-128 cbc 10891.69k 11434.13k 11575.65k 11624.11k 11643.56k
aes-192 cbc 9443.80k 9895.38k 9990.57k 10022.19k 10018.82k
aes-256 cbc 8375.04k 8714.23k 8784.65k 8793.09k 8809.13k
sha256 2512.47k 6054.49k 11072.77k 13984.09k 15106.05k
sha512 660.55k 2642.84k 3897.28k 5355.97k 6060.44k
sign verify sign/s verify/s
rsa 2048 bits 0.094906s 0.002631s 10.5 380.1
dsa 2048 bits 0.026946s 0.032468s 37.1 30.8

| MIPS 1004Kc V2.15 | 598.01 | 1.0.2d | 45868030 | 30101820 | 13991810 | 5380620 |
| 6652930 | 2374720 | 11537980 | 10034520 | 8806400 | 10.5 | 379.2 | 36.9 | 30.3 |

•'MT7621A' @900MHz x4 threads (overclocked)
aes-128 cbc 21916.12k 22852.87k 23262.38k 23263.85k 23328.09k
aes-256 cbc 16773.94k 17312.23k 17518.51k 17574.23k 17521.63k

•'MT7621A' @1.0GHz x4 threads (overclocked)
aes-128 cbc 25749.26k 27047.57k 27310.01k 27478.02k 27488.63k
aes-256 cbc 19704.94k 20471.32k 20624.86k 20656.38k 20689.44k

•'MT7628' @575MHz - AES-Engine
aes-256 cbc (8192 bytes) - 98249.57k

•'RT-AC66U' @600MHz
aes-128 cbc 9398.56k 9941.17k 10138.71k 10195.09k 10234.54k

•'RT-AC56U' @1200MHz (overclocked)
aes-128 cbc 25596.63k 28151.82k 29251.64k 29363.88k 29587.09k
aes-256 cbc 28449.18k 31325.60k 32491.95k 32766.98k 32882.69k
__________________________________________________ ______________________________

З.Ы. Обсуждение роутера ASUS RT-N56U и ASUS RT-N56U B1 (часть 11), #102 (http://forum.ixbt.com/topic.cgi?id=14:62910:102#102) :D

ryzhov_al
27-10-2016, 07:40
ryzhov_al
Можно будет составить другую таблицу ;)

• OpenVPN throughput benchmark:
Или возьмите встроенный бенчмарк 7zip, он сам умеет (http://forums.zyxmon.org/viewtopic.php?p=15415#p15415) определять кол-во потоков.

Omega
17-08-2017, 20:14
CPU: Qualcomm IPQ4018 (@717MHz, 4 cores), Flash: 128MB, RAM: 128MB

WikiDevi: https://wikidevi.com/wiki/ASUS_RT-AC58U

ASUS RT-AC58U (AC1300G+)
FW: ASUS 3.0.0.4.380_7485


Overhead for getting time: 1us
Time to run memory bench: 1.99[secs]
Time to run computation of pi (2400 digits, 10 times): 7.76[secs]
Time to run computation of e (9009 digits): 7.19[secs]
Time to run float bench: 0.05[secs]
Total time: 17.0s

http://digiland.tw/viewtopic.php?pid=20350#p20350

ASUS RT-AC58U
FW: LEDE https://github.com/LEDE-RT-AC58U/LEDE_RT-AC58U


Test: https://wiki.openwrt.org/doc/howto/benchmark.openssl


The 'numbers' are in 1000s of bytes per second processed.

type 16 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
md5 3600.85k 13021.81k 35280.80k 61596.15k 78787.44k
sha1 2698.65k 7029.33k 13770.84k 18108.13k 19935.68k
des cbc 7975.87k 8278.05k 8366.61k 8390.68k 8396.12k
des ede3 2897.51k 2942.81k 2953.36k 2956.33k 2958.37k
aes-128 cbc 11883.04k 12394.74k 12556.84k 12638.21k 12609.15k
aes-192 cbc 10359.64k 10698.48k 10896.13k 10900.33k 10913.59k
aes-256 cbc 9135.53k 9409.40k 9483.23k 9500.75k 9506.53k
sha256 2691.05k 5805.14k 9816.06k 11828.39k 12592.82k
sha512 652.60k 2604.10k 3560.61k 4789.33k 5326.16k
sign verify sign/s verify/s
rsa 2048 bits 0.073285s 0.001854s 13.6 539.5
sign verify sign/s verify/s
dsa 2048 bits 0.021121s 0.023357s 47.3 42.8

| r4533 | ARMv7 Processor rev 5 (v7l) | 103.35 | ARMv7 Processor rev 5 (v7l) | 103.35
| ARMv7 Processor rev 5 (v7l) | 103.35 | ARMv7 Processor rev 5 (v7l) | 103.35
| Generi DT based system | 1.0.2k | 61596150 | 18108130 | 11828390 | 4789330
| 8390680 | 2956330 | 12638210 | 10900330 | 9500750 | 13.6 | 539.5 47.3 | 42.8 |

openssl speed aes-256-cbc


root@LEDE:~# /usr/bin/openssl speed aes-256-cbc

Doing aes-256 cbc for 3s on 16 size blocks: 1829834 aes-256 cbc's in 2.96s
Doing aes-256 cbc for 3s on 64 size blocks: 522608 aes-256 cbc's in 3.01s
Doing aes-256 cbc for 3s on 256 size blocks: 134013 aes-256 cbc's in 3.01s
Doing aes-256 cbc for 3s on 1024 size blocks: 33722 aes-256 cbc's in 3.01s
Doing aes-256 cbc for 3s on 8192 size blocks: 4223 aes-256 cbc's in 3.01s
OpenSSL 1.0.2l' 25 May 2017
built on: reproducible build, date unspecified
options:bn(64,32) rc4(ptr,char) des(idx,cisc,2,long) aes(partial) blowfish(ptr)
compiler: ccache_cc -I. -I.. -I../include -fPIC -DOPENSSL_PIC -DZLIB_SHARED -DZLIB -DOPENSSL_THREADS -D_REENTRANT -DDSO_DLFCN
-DHAVE_DLFCN_H -I/build/lede-snapshots/arm_cortex-a15_neon-vfpv4/build/sdk/staging_dir/target-arm_cortex-a15+neon-vfpv4_musl_eabi/usr/include -I/build/lede-snapshots/arm_cortex-a15_neon-vfpv4/build/sdk/staging_dir/target-arm_cortex-a15+neon-vfpv4_musl_eabi/include -I/build/lede-snapshots/arm_cortex-a15_neon-vfpv4/build/sdk/staging_dir/toolchain-arm_cortex-a15+neon-vfpv4_gcc-5.4.0_musl_eabi/usr/include -I/build/lede-snapshots/arm_cortex-a15_neon-vfpv4/build/sdk/staging_dir/toolchain-arm_cortex-a15+neon-vfpv4_gcc-5.4.0_musl_eabi/include/fortify -I/build/lede-snapshots/arm_cortex-a15_neon-vfpv4/build/sdk/staging_dir/toolchain-arm_cortex-a15+neon-vfpv4_gcc-5.4.0_musl_eabi/include -znow -zrelro -DOPENSSL_SMALL_FOOTPRINT -DHAVE_CRYPTODEV -DUSE_CRYPTODEV_DIGESTS -DOPENSSL_NO_ERR -DTERMIOS -Os -pipe -mcpu=cortex-a15 -mfpu=neon-vfpv4 -fno-caller-saves -fno-plt -fhonour-copts -Wno-error=unused-but-set-variable -Wno-error=unused-result -mfloat-abi=hard -iremap/build/lede-snapshots/arm_cortex-a15_neon-vfpv4/build/sdk/build_dir/target-arm_cortex-a15+neon-vfpv4_musl_eabi/openssl-1.0.2l:openssl-1.0.2l -Wformat -Werror=format-security -fstack-protector -D_FORTIFY_SOURCE=1 -Wl,-z,now -Wl,-z,relro -fpic -I/build/lede-snapshots/arm_cortex-a15_neon-vfpv4/build/sdk/feeds/base/package/libs/openssl/include -ffunction-sections -fdata-sections -fomit-frame-pointer -Wall -DOPENSSL_BN_ASM_MONT -DOPENSSL_BN_ASM_GF2m -DSHA1_ASM -DSHA256_ASM -DSHA512_ASM -DAES_ASM -DBSAES_ASM -DGHASH_ASM

The 'numbers' are in 1000s of bytes per second processed.
type 16 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
aes-256 cbc 9890.99k 11111.93k 11397.78k 11472.20k 11493.29k

Чем цифры больше (в последней строке), тем cpu быстрее. Этот тест задействует операции с плавающей запятой.
Цифры относятся к пакету lede/openwrt. Вариант Entware от Voxel быстрее на 30%.
http://forum.ixbt.com/topic.cgi?id=14:63557:4848#4848 ;)

bumbon4ik
28-01-2019, 21:38
RT AC68U

bumbon4ik@RT-AC68U-99E8:/tmp/home/root# nvram get clkfreq
1000,666

bumbon4ik@RT-AC68U-99E8:/tmp/home/root# cpubench
This is CPU and memory benchmark for OpenWRT v0.6. This will then take some time ... (typically 30-60 seconds on a 200MHz computer)
Overhead for getting time: 8us
Time to run memory bench: 1.70[secs]
Time to run computation of pi (2400 digits, 10 times): 4.90[secs]
Time to run computation of e (9009 digits): 4.42[secs]
Time to run float bench: 0.04[secs]
Total time: 11.1s

You can copy/paste the following line in the wiki table at: http://wiki.openwrt. org/HardwarePerformance
|| 2019-01-28 || ''Author'' || 1.7s || 4.9s || 4.4s || 0.0s || v0.6 || ''OS'' || ''DeviceModel'' || ''CPU model'' || ''CPU Frequency'' || ''LinkToHwPage'' ||

на стоке не нашёл)