A History of Silicon Valley

Table of Contents | Timeline of Silicon Valley | A photographic tour
History pages | Editor | Correspondence
Purchase the book

These are excerpts from Piero Scaruffi's book
"A History of Silicon Valley"


(Copyright © 2016 Piero Scaruffi)

The Selfies (2011-16)

click here for the other sections of this chapter

Beyond the Cloud

The biggest outside threat to Google and Apple was probably Amazon. On the surface the Seattle-based rival was simply a huge retailer (in 2014 carrying more than 200 million items sold by over two million third-party vendors), and mainly famous (or infamous) for destroying the business of bookstores, but deeper down Amazon was an extremely sophisticated technology company that pioneered online customer reviews and ratings, that in 2006 introduced pay-as-you-go cloud computing (later used by the likes of Instagram, Pinterest and Spotify), that in 2007 turned ebook readers into commodities, that in the 2010s turned its colossal PHX6 fulfilment center at Phoenix (Arizona) into an automated, robot-intensive facility (while making the whole supply chain available to other merchants as well), and that was rapidly moving online business towards same-day delivery and mobile shopping.

New methods for software development made it easier for startups to create distributed applications.

Introduced in 2008 and based on open-source Xen technology, Amazon's Elastic Compute Cloud (EC2) provided a self-service portal for developers to rent their own virtual private servers on which to run their own applications.

Containers were a legacy of the Unix operating system (SUN's Solaris edition of 2004, an evolution of Bill Joy's "chroot" command incorporated into Berkeley Unix in 1982). An application run within a container is isolated from the hardware platform. If it sounds like virtualization, that's because they are closely related. Containers are a faster way to implementat virtual machines and a better way to optimize hardware resources. The net result for the software developer is the same: application portability across platforms. In 2008 Solomon Hykes and Sebastien Pahl founded DotCloud (later Decker) in San Francisco to provide an open-source platform based on containers that quickly built an entire community of developers. Google then legitimized the field when it introduced Kubernetes to manage clusters of containers.

The Nebula project, started in 2008 by Ray O'Brien at NASA Ames in Mountain View, evolved in 2010 into an open-source cloud platform, OpenStack, technically a joint venture with Texas-based cloud service Rackspace (a 1998 offshoot of San Antonio's Trinity University). Rackspace donated the software behind OpenStack's Swift storage service, and in 2011 acquired Anso Labs, the San Francisco-based company that maintained Nasa's Nebula cloud (and that developed the Nova computer). In 2012 some of Anso Labs' original team (including former NASA Ames scientists Chris Kemp and Devin Carlen) left RackSpace and founded Nebula in Mountain View. By 2015 more than 500 companies had joined OpenStack, basically an open-source alternative to Amazon's and VMware's cloud services.

The new generation of middleware was represented by platforms like Anypoint, introduced in 2013 by MuleSoft, a company started in 2006 in San Francisco by Ross Mason and Dave Rosenberg. It connected and integrated software as a service (SaaS) on the cloud with legacy databases and applications running on personal devices: "any application, any data source, any device, any API".

In 2015 Cisco acquired cloud-security provider OpenDNS, founded in 2006 by David Ulevitch in San Francisco (who in 2001 in Saint Louis had launched the free DNS management service EveryDNS).

Jyoti Bansal founded AppDynamics (in 2008 in San Francisco) to improve performance of cloud-based applications, which were typically heavily distributed (the startup would become a unicorn in 2015).

Notice that peer-to-peer, revitalized by the blockchain mechanism, was competing with cloud computing to introduce collectivization in cyberspace. The two models were wildly different, but both were removing the data from the individual's home computer and relocating the data somewhere else: either on a distributed network of participants or on a distributed network of corporate computers. Peer-to-peer computing had generated a lot of philosophical, sociological and political discussion by the independents who had created it and were using it. Cloud computing was less naturally analyzed and discussed, because it mostly belonged to governments and corporations, its "architects" mostly unknown.


click here for the other sections of the chapter "The Selfies (2011-16)"
(Copyright © 2016 Piero Scaruffi)

Table of Contents | Timeline of Silicon Valley | A photographic tour | History pages | Editor | Correspondence