Discussion

Benchmarks between Banana Pro and Raspberry Pi 2

16 3907
The user blindpet has made a benchmarks between Banana Pro and Raspberry Pi 2, more details can refer to the link below:
http://www.htpcguides.com/raspbe ... -pi-pro-benchmarks/

test1.jpg
test2.jpg
test3.jpg
test4.jpg
test5.jpg
test6.jpg
test7.jpg
Wow! Great job on Banana Pro!

I also have other Banana Pro benchmarks in my Raspberry Pi 2 Review:

http://www.mikronauts.com/raspbe ... 2-model-b-review/6/

And the BPro does quite well - check out the application launch times, and more.

The BPro was also the only SBC whose power consumption went to zero on a software shutdown.

It proves also the Rasberry choice of multiplexing USB (so IO) and Ethernet is really a pity

Elfen  
This is interesting.

cxy  
It is very interesting

Elfen  
I'm not saying that there cheating involved here but there are certain notes to follow.

The R-Pi Ethernet is 10/100, and the B-Pi Ethernet is 10/100/1000, so the B-Pi will always win this hands down. This is a given.

Were these Pi's software over-clocked? I don't think so. So note - the R-Pi is 700MHz, and the R-Pi2 is 900MHz while the B-Pi is 1GHz. That there says that the B-Pi will win as is it the faster CPU. "But the R-Pi2 is a Quad Core and the B-Pi is Dual Core?" So? That does not mean a thing unless you are using software that takes advantage of that, if not most of the major processing it done by 1 CPU while the others sit and wait. In that, because the B-Pi is running at 1GHz and the R-Pi2 is running at 900Mhz, the B-Pi will always win. What would happen if you software over clock the R-Pi2 and even the R-Pi to 1GHz? Hmmm... I bet it will be a much closer race.

SATA Drive test? Of course the B-Pi will win because the R-Pi does not have that.

I'm not arguing that the R-Pi/R-Pi2 is better. I prefer the B-Pi over the R-Pi for most applications. All I'm saying is "Lets level the playing field a bit and lets see what happens." The B-Pi is better for most things than the R-Pi, but not everything. The R-Pi does have its place.

Elfen  
I read the article and I noticed one BIG and MAJOR Issue here - the OS is different on both systems! That skews everything!

Lets make the test fair - boot Raspbian on all 4 units and see how they compare without the over-clocking the R-Pis and then with over-clocking the R-Pi to 1 GHz and see how they compare! Until then, this is not a true comparison!

Edited by destroyedlolo at Thu Mar 5, 2015 21:41
Elfen replied at Thu Mar 5, 2015 20:07
The R-Pi Ethernet is 10/100, and the B-Pi Ethernet is 10/100/1000, so the B-Pi will always win this hands down. This is a given.


Even with both connected to a 100 Mb hub, the rPI will be worst, by design ... due to the fact both USB and Ethernet are sharing the same channel. The A20 has different channels for that, so far better performances.

So note - the R-Pi is 700MHz, and the R-Pi2 is 900MHz while the B-Pi is 1GHz.


By default, bPI don't go above 900 Mhz : faster clock are deactivated in LeMaker's fex file.

That does not mean a thing unless you are using software that takes advantage of that, if not most of the major processing it done by 1 CPU while the others sit and wait


It's true on non multitasking environment like Arduino or PICs : on bPI and such, you have a plain OS so tasks are dispatched on available cores. In addition, you have some of the tests above that are multi-treads (clearly indicated in the report) and in such situations, rPI is obviously faster than bPI.

SATA Drive test? Of course the B-Pi will win because the R-Pi does not have that.


rPI2 has 4 cores, bPI only 2 / rPI doesn't have SATA, bPI has. The choice is mostly here.

The R-Pi does have its place.


Clear Now, I take also in account the Raspberry foundation and clearly, I don't like a company that claiming everywhere they have rights on others' work without never provided any proof (and it's annoying some 3rd parties like sourceforge follow them w/o arguing, w/o thinking) ... But it's only personal reasons

Elfen  
Edited by Elfen at Sat Mar 7, 2015 11:06
Clear  Now, I take also in account the Raspberry foundation and clearly, I don't like a company that claiming everywhere they have rights on others' work without never provided any proof (and it's annoying some 3rd parties like sourceforge follow them w/o arguing, w/o thinking) ... But it's only personal reasons


This is so true on so many levels.

I need to sit down and set up the same tests you did here since I have all three systems (R-Pi, R-Pi2, and B-Pi, not B-Pro). Thing is I need to set up Raspian for B-Pi for all the machines  because B-Pi's version of Raspian does run on a R-Pi. But the biggest problem I have is RaspberryPi.org's statement that the R-Pi2 is "100% compatible with the older R-Pi's." This is not true, as the R-Pi2 will not run on Raspian 2014 or Raspian 2013, but it will run on the newly released Raspian 2015. Just like the R-Pi model B+ will not run on Raspian 2013 but does Raspian 2014. Only the Model B v1.1 can run Raspian 2013 to 2015.

In texting on LinkedIn with one of the R-Pi Engineers, he stated that to run Raspian 2013 on  Model B+ requires a firmware downgrade. I laughed at this. And from this, I can deduce  that the R-Pi2 would also need a firmware downgrade to run the older Raspians. Would I would guess the downgrade would be needed for a R-Pi2 to run the Banana Pi Raspian Version like the Models B and B+ can run. Problem is such a firmware downgrade for the R-Pi2 does not exist.

Thanks for the information. It is helpful.

You have to log in before you can reply Login | Sign Up

Points Rules