On 28/06/19 6:43 AM, Chris Muller wrote:
> 
>     The foundation has strayed away from its original intent to design and
>     build affordable development boards to promote hands on computing
>     amongst students [1]. Its main appeal was its openness (global access),
>     low-cost (i.e. every student could have one or more) and low-power
>     (battery powered) and wireless connectivity (wifi/BT). With RPi 4, the
>     power hogs in GPU and Ethernet have become worse.
> 
> 
> According to the Bitscope rep I asked, the Pi 4 should budget about 
> 8-10W per node.  To me, that seems pretty amazing for a 4-core, 4GB RAM 
> computer.

That is a typical power draw. I was referring to the upwards creep in 
the power adapter -  5V/1A to 5V/2A to 5V/3A now.

But my disappointment is not about specs. It is about the divergence 
from the motivations that led to RPi - an open physical computing kit 
less than the price of a CS textbook. This was also reiterated in 
opening para of RPi Zero announcement [1] in 2015. RPi is not a 
"batteries included" kit. A complete study kit would be $50+ and it 
would increase now.

[1] https://www.raspberrypi.org/blog/raspberry-pi-zero/

Cortex boards are unarguably more power efficient than x86 boards and I 
do like many of the design choices in RPi. But when it comes to laptop 
computing needs, Cortex is yet to seriously challenge single thread 
performance of x86/x64 boards. In positioning itself towards desktop 
computing space, RPi 4 is falling between two stools.

Of course, RPi is free to choose to change their direction. I just 
expressed my disappointment about losing a excellent player in the 
student kit space.

Regards .. Subbu