Of Windows, Newton’s, iPad’s and 10GBASE-T

[Editor’s note: George Zimmerman’s post is copied here as a guest post to allow the guest posts on this topic to be gathered in one place. Check out George’s other posts at Up and Down the Network Stack.]

A recent article in EDN by Robert Cravotta pointed out how changes in the adoption of new technologies (so-called “inflection points”) are often related to integration of multiple technologies together. These are “easier to spot in hindsight”, and often comprise the right mix of integration while hiding the complexity to the end user. Cravotta cites TI’s first DSP, Microsoft’s Windows 3.0 and Apple’s iPhone as examples of inflection points, and, more particularly for the integration of multiple elements and hiding complexity than for being first in their segments. While Cravotta asserts “inflection points are not just about technical capability”, I’m sure he would agree that technical capability is often necessary to change the rate and range of adoption.  Those of us with a few gray hairs remember Windows 2.0 & Apple’s Newton – both like their later successful successors, both extremely limited by the technical capability of processors in their day. Windows 2.0 was a thin shell on top of DOS, and Apple’s Newton – revolutionary in concept as a pen-based device, was so far behind the capability curve that no one tried anything similar for a nearly a decade. Apple only today is introducing a new tablet (the iPad), after the iPhone has shown the world that touch-sensing, user interface, network connectivity and processing technology have caught up with the demands of the application.

Years ago, as Chief Scientist at PairGain Technologies, I witnessed a similar wave of enabling technology in consumer broadband. Telephone modems dominated the world, and anything more was thought to either not be possible or require too much power. Remember 33.6 and 56 kbits?, I distinctly recall giving a talk to a communications systems group at a major research institute and answering a question from the audience that can roughly be translated as “didn’t Shannon say that was impossible?”. Later it was, “won’t that take too much power?” Yet today, we see ADSL, VDSL and cable modems as routinely as telephones. That wave of inflection transformed the network, interpersonal communications, how we conduct commerce, and our social networks. And it follows Cravotta’s premise – because the connectivity is invisible.

Today I believe we are in the midst of a similar inflection. Boundaries between computing devices are disappearing. This is happening because of a combination of processing power, including multiple cores, virtualization, and network connectivity. All of the things that will make the iPad a non-Newton (whether it is successful may still depend on other factors); but they are changing our world in a much more substantial way. Personally, my part in this drama is making the network connectivity fast and simple enough that it becomes invisible. The 10GBASE-T transceivers that I’m involved with at Solarflare today, as well as the 10G controllers and Network Interface Cards, provide connectivity at power, cost, and performance levels that still aren’t fully realized – and they’ll continue to become less visible (lower power, higher performance) as we introduce each new generation. This revolution is just starting, but the explosion in LAN bandwidth that it brings will combine with virtualization, increased processing power, and yes, mobile interface/compute terminals to produce the next inflection point, breaking down the walls between computing boxes. I can’t wait for it to happen.

George Zimmerman, PhD, Chief Technology Officer for Solarflare Communications