why ras. pi have ethernet combined with usb unlike banana li

[Bhavesh]

I think it's one of biggest drawback. It has Ethernet shared with usb. so Ethernet can't get 1000 scale.

Else, It would be perfect for using it as NAS. Is it going to be change in near future?

[W. H. Heydt]

The Banana Pi (which should be considered a clone of the Cubieboard 2, *not* a clone of the Raspberry Pi) uses an Allwiner A20 SoC. The SoC used for the Pis, either a BCM2835 or BCM2836, does not have a built in Ethernet interface. As a result, the Pi has to use an external chip, the LAN9514, to have more than one USB port and an Ethernet port.

The only way around this would be to use a completely different SoC, which would require designing a whole new PCB, redoing the bill of materials, and retooling the production lines. In other words...not going to happen any time prior the existence of a "Pi3", and possibly not even then.

[riklaunim]

They could have changed it for RPi2 but they didn't... But there is a lot of ARM single board computers with native Ethernet. Note that not every one has true/full Gigabit Ethernet.

[rpdom]

They could have changed it for RPi2 but they didn't... But there is a lot of ARM single board computers with native Ethernet. Note that not every one has true/full Gigabit Ethernet.

Not easily, they couldn't.

It's a slightly different matter to tack on a few extra ARM cores than to add a complete new block of Ethernet hardware.

[eriktheitalian]

If big ones not sabotage raspberry then raspberry evolution can be done. What is this mean ? There is only classical desktop usage alternative is x86 + linux or windows. Raspberry selling numbers very good. Its very good for broadcom SOC source coding. They can release real desktop computer alternative SOC.

Android is not designed for coding works. Not server related. There is no alternative for x86+windows. But users hungry for this.

[jamesh]

Changing the SoC is a BIG THING. Some examples as to why...

Consider the hardware implications.

• GPIO will be completely different, meaning all code written so far that uses them will break
  Probably no camera (Very few SoC have camera interfaces)
  Probably no chance for a dedicated display
  Completely different GPU.

And what about software?

• New distro required, so all the work done on Raspbian is lost.
  A lot of software written will need to be modified if is uses SoC specific features, like GPIO.
  All the optimisation work done for VC4 will be wasted.

It's not impossible, just a heck of a lot of work

[fruitoftheloom]

Changing the SoC is a BIG THING. Some examples as to why...

Consider the hardware implications.

• GPIO will be completely different, meaning all code written so far that uses them will break
  Probably no camera (Very few SoC have camera interfaces)
  Probably no chance for a dedicated display
  Completely different GPU.

And what about software?

• New distro required, so all the work done on Raspbian is lost.
  A lot of software written will need to be modified if is uses SoC specific features, like GPIO.
  All the optimisation work done for VC4 will be wasted.

It's not impossible, just a heck of a lot of work

Yes just another re-hash of previous posts on this subject

[texy]

Changing the SoC is a BIG THING. Some examples as to why...

Consider the hardware implications.

• GPIO will be completely different, meaning all code written so far that uses them will break
  Probably no camera (Very few SoC have camera interfaces)
  Probably no chance for a dedicated display
  Completely different GPU.

And what about software?

• New distro required, so all the work done on Raspbian is lost.
  A lot of software written will need to be modified if is uses SoC specific features, like GPIO.
  All the optimisation work done for VC4 will be wasted.

It's not impossible, just a heck of a lot of work

Yes just another re-hash of previous posts on this subject

...and locking because of it.....