I'm anxious to hear more details about Silk but from what I understand so far, the protocol between the device and the backend EC2 servers is SPDY (http://www.c

I'm anxious to hear more details about Silk but from what I understand so far, the protocol between the device and the backend EC2 servers is SPDY (http://www.chromium.org/spdy/spdy-protocol http://www.chromium.org/spdy/spdy-protocol) or something pretty similar. SPDY is a transport protocol, basically an improvement over HTTP to avoid the latencies associated with opening and closing many TCP sessions in order to render a single web page. This approach also allows Amazon to maintain a huge local web cache including predictive pre-fetching. As many people have pointed out this stuff isn't 'new' other than perhaps the scale at which Amazon operates compared to existing systems. I don't think any of this is going to change the computational requirements at the device for handling client-side processing (i.e. Javascript), which is what the author seems to suggest.