Big data refers to the flood of data generated not only by the Web but also by embedded sensors, video and social networking. While the construct differs from cloud services, both are being driven
the ubiquity of data delivered by mobile devices and sensors linked to
the emerging cloud ecosystem.
The authors of the Berkeley study
stressed that the phrase “cloud computing” is overused and has been
drained of meaning by marketers. The researchers define it as a platform
for delivering “computing services – data storage, computation and
networking – to users at the time, to the location and in the quantity
they wish to consume, with costs based only on the resources used.”
that is needed to tap into cloud computing is an Internet connection.
For companies, cloud services “transform computing from a capital
expense to an operating expense,” the study found.
By contrast, the
numerous data centers created by cloud services providers like Amazon,
Google and Microsoft cost more than $500 million each, the authors
noted. Hence, these providers are increasingly likened to utilities
operating a critical new national infrastructure.
power of the cloud infrastructure rests with “the dynamic allocation of
resources and the ‘illusion’ of infinite scale,” the study found.
Indeed, it is the value delivered by cloud services that has prompted
researchers to tout it as the next big engine of economic growth. “Cloud
providers are competing on value-based differentiation on attributes
such as service level and functionality,” the Berkeley study concludes.
what does all this mean for reviving U.S. manufacturing and economic
In the foundry era, what can be designed and
what can be manufactured have emerged as a critical question for chip
makers. “The answer is not constant,” Zysman said. “It changes with who
the competitors are, what the tools are that can be made available.”
tech companies are being forced to decide whether they are producing a
“strategic asset or a vulnerable commodity,” Zysman said. For example,
Ericsson brought its semiconductor design for smartphones in-house
because the design itself became the product.
In a forthcoming
book with Georgia Tech researcher Dan Breznitz, Zysman will argue that
“we’re watching competition increasingly by phases of production rather
by sectors,” as he put it in the interview. For instance, product design often remains in Silicon
Valley, then Apple scales production at places like its iPhone and iPad
manufacturing lines at China’s Foxconn. “So the old notion of clusters
takes on a different meaning,” Zysman said. “It’s not clusters around
sectors [but] clusters around phases of production.”
longstanding concerns over the decline of western manufacturing, Zysman
stressed that not all is lost. When determining where the real value,
or intellectual property, resides in products like the iPhone, “the
value is in the West,” he stressed. Retaining IP doesn’t translate into
more value-added manufacturing jobs, however. “That’s an [income
distribution] issue because that [value] doesn’t go to the workforce.
That goes to the stock value of Apple,” Zysman noted. “It doesn’t change
the income distribution issues for workers.”
related "Rebuilding America" stories:
manufacturing, ecosystems seen as keys to U.S. innovation
by design: New skills needed to compete
to play hard ball on tech manufacturing