Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 49
4
Implications of Changes in the Global Advanced
Computing Landscape for U.S. National Security
T he viability, efficiency, and, ultimately, success of
the global economic ecosystem depend in part on
the flow of goods and services throughout the
global value chain, from design through fabrication to
consumption. In computing and information technology,
Chapter 1) and the rise of the post-personal computer
(PC) ecosystem of smart devices, coupled with cloud-
computing capabilities, further complicate the landscape.
This chapter discusses several emerging changes in the
global advanced computing landscape that have
the locus of innovation, influence, and early access can implications for U.S. national security, including
and has shifted throughout the history of the modern parallelism in hardware and software (Section 4.1), the
information technology (IT) era. These shifts can have integrity and reliability of the global supply chain
significant implications for U.S. competitiveness and (Section 4.2), the decline of custom production (Section
national security. 4.3), convergence of civilian and defense technological
In the committee's view, the United States currently capabilities (Section 4.4), the rise of a new post-PC
enjoys a technological advantage in advanced computing paradigm driven by mass information and
hardware and software capabilities but that technological communications technology (ICT) consumerization
gap is narrowing, not only due to the technical (Section 4.5), new market-driven innovation centers
challenges described in Chapter 1, but also because other (Section 4.6), the future educational and research
global economic competitors (e.g., China) are making a landscape on advanced computing (Section 4.7),
concerted effort to develop their own indigenous cybersecurity and software (Section 4.8), and possible
computing design and manufacturing capabilities. defense ICT outcomes (Section 4.9).
Moreover, the design and fabrication of such
technologies are increasingly globally distributed. 4.1 Parallelism in Hardware and Software
Market success, of course, is only partially correlated
with technological preeminence, as the ecosystem of In U.S. defense and national security, one element of
producers and consumers and market size, together with the U.S. advantage in defense ICT has accrued from
network effects, are also key determinants. rapid increases in application performance, which in turn
In the committee's view, national security concerns has depended on rapid increases in single-processor
for the United States related to anticipated long-term performance. As the latter ends, continued application
developments in advanced computing come not just from performance increases will likely only be possible if
potential threats to U.S. technological superiority, but there is a shift to the development of applications that
also from changes to the nature and structure of technical can take advantage of parallel hardware. The inability of
innovation and to the marketplace for computing and defense software to make this transition faster than
information technology. Intensifying competition will competitors in the global market or our potential
affect the global supply chain and reshape the numbers adversaries has significant implications for U.S.
and types of commercial players that survive in a rapidly competitiveness and national security.
evolving marketplace. The diminishing performance The slowdown in performance increases for single-
returns from traditional silicon advances that have helped core processors is a matter of physical (e.g., power
existing software systems run ever faster (described in density and dissipation and quantum barriers) and
49
OCR for page 50
50 THE GLOBAL ECOSYSTEM IN ADVANCED COMPUTING
technological (e.g., gate length, lithography, power U.S. national security has long relied on an
dissipation, wire scaling, and materials) limits. Thus, information technology advantage. Given the dramatic
even if resources were plentiful, it would not be possible shift to multicore chips and explicit parallelism, defense
to simply buy or make appropriately targeted ICT will need to transition to tools, techniques, and
investments that would result in continuing exponential processes that can meet defense needs through effective
speed-ups for single-core processors. An additional use of new parallel software models and emerging
matter of physics is the power constraints that are driving hardware approaches. Such a transition will be difficult,
the industry from homogeneous multicore chips to and even if this transition is made successfully--a
heterogeneous parallelism, for example, using graphics challenge not just for defense, but for even the most
processing units, accelerators, and reconfigurable advanced commercial interests as well--the growth rate
fabrics. Once again, national security processes and of computing performance is expected to continue to
deployments will need to adapt to use heterogeneous slow, making it easier for the rest of the world, including
parallelism to maintain advantage. adversaries, to catch up.
Exponentially increasing processor speed has
traditionally served as a proxy for higher-performing, 4.2 Integrity and Reliability of the Global Supply
more capable, and more innovative systems. Absent this Chain
traditional metric of continually increasing performance
(whether from sequential or parallel systems), focusing Maintaining the integrity of the global supply chain
on other metrics will likely come to the fore. In many is a serious challenge. The supply chain for integrated
cases, design and innovation efforts will focus on circuits (computer chips) is of particular interest given
combinations of improvements in diverse dimensions, that they are key components of all computing systems.
such as cost, energy, weight, robustness, and security. Some fabrication facilities are still present in the United
Traditional performance improvements would help to States. For instance, Intel is the primary operator of
achieve these, but if such improvements are not large-scale, state-of-the-art semiconductor fabrication
forthcoming, other means of achieving these facilities in the United States, though it also has such
improvements will be needed. facilities outside the United States. IBM and other
Developing, verifying, and deploying software to companies operate facilities in the United States that
complement advanced hardware is fraught with target more specialized markets and national security
challenges. Moves to homogeneous and then needs.4 However, the United States is increasingly
heterogeneous parallelism will amplify these dependent on foreign sources of microchip production
challenges.1 These challenges are especially prevalent in and on device assembly and testing capabilities that are
defense and national security systems. Moreover, concentrated in a handful of countries.
defense is notable for its relatively slow adoption of Developing secure sources of production is also
innovative hardware and software that now emerge from challenging. A global supply chain increases the
the commercial rather than the military sector. New and likelihood that compromised and counterfeit products
faster-moving threats with fewer legacy concerns may can be introduced in mission-critical infrastructure.5
make this status quo untenable.2,3
1
The 2009 National Research Council (NRC) report Critical trasted with U.S. companies' more near-term focus. Too much of a
Code: Software Producibility for Defense assesses the growing short-term focus cuts into long-term success. Moreover, once the
importance of software for national security and examines how the former made less money available, addressing the latter became
U.S. DOD can most effectively meet its future software needs. more difficult.
2 4
The 2009 NRC report, Achieving Effective Acquisition of See http://www.nsa.gov/business/programs/tapo.shtml. Last
Information Technology in the Department of Defense calls for the accessed on July 2, 2012.
5
DOD to acquire information technology systems using a funda- Indeed, an immediate challenge for U.S. access to the global
mentally different acquisition process based on iterative, incre- semiconductor value chain is that some U.S. defense contractors
mental development practices. have been deceived into using counterfeit electronics parts. At a
3
This is reminiscent of the 20th century U.S. automobile sector. November 2011 hearing, the Senate Armed Services Committee
The U.S. auto industry moved from being the best in the world to noted that such fake parts could have disastrous consequences for
being high cost and slow to adopt new processes and technologies. the performance of U.S. defense equipment such as helicopter
This decline was masked for years by the lack of credible compe- night-vision systems and aircraft video display units. See U.S.
tition. The arrival of Japanese and other foreign automakers Senate Committee on Armed Services, Hearing to receive testi-
changed the competitive landscape two ways. First, the Japanese mony on the Committee's investigation into counterfeit electronic
focus on manufacturing efficiency exposed U.S. companies' pro- parts in the Department of Defense supply chain, November 8,
cess problems and eroded near-term profits. Second, sustained 2011. See also DOD's TRUST in Integrated Circuits Program
long-term Japanese investments (e.g., on energy efficiency) con- (available at (http://www.darpa.mil/Our_Work/MTO/Programs/
OCR for page 51
IMPLICATIONS FOR U.S. NATIONAL SECURITY 51
Developing ways to ascertain and monitor the of dominant market players that make commoditized
provenance of semiconductor products will become ever products may significantly increase costs.
more important. Related issues of hardware and software
verification and validation will continue to be critical 4.4 Convergence of Civilian and Defense
issues, particularly as the complexity of systems Technological Capabilities
continues to rise.
In addition to challenges related to integrity and The convergence of civilian and defense
security, the global interdependence of design, technologies is accelerating, driven by rapid and cost-
component fabrication, and assembly means that risks of effective technological progress in a highly competitive
disruption due to natural disasters, political conflict, or commercial marketplace, especially as compared with
constrained access to raw materials become greater. A the often lengthy and rigid procurement processes in the
single event, such as the March 2011 earthquake-tsunami defense sector. Convergence is most evident in
in Japan or the more recent floods in Thailand, can electronics, where a growing proportion of U.S. defense
disrupt global product deliveries for months.6 Similarly, needs are being met by COTS technologies. At the same
restrictions on shipments of rare earths, key elements of time, the U.S. defense establishment's ability to
chip fabrication, can stall production lines. The influence the development of the global semiconductor
globalization of science and technology (S&T) and of industry, similarly to what happened with
the computing marketplace in combination with supercomputers (which create the chip components of
specialization (only a few suppliers of a particular COTS products) through sheer volume has been reduced.
component) and just-in-time inventory practices all add For instance, the U.S. military accounted for a large
to the risk as well. More generally, a disaster or a well- proportion of sales from the global semiconductor
targeted action from an adversary could constrain or industry in that industry's formative years, but that
interrupt global supplies, potentially placing the United proportion had fallen to just 1 percent of global
States in a defensive position due to competing demands microcircuit sales by the late 2000s.7
between U.S. defense needs, commercial production The convergence between civilian and defense
requirements, and the producing region's own needs. hardware capabilities and ease of access to openly
available technological products that may be just as good
4.3 Decline of Custom Production or even more advanced than equivalent defense
technologies has implications for U.S. defense.8 In
A decrease in the number of specialized companies particular, such convergence allows greater opportunity
able to make custom products for defense needs is also for adversaries to narrow the technological gap with the
relevant to national security. Although commercial off- United States. In such an environment, time to
the-shelf (COTS) products are widely used in defense integration and time to deployment will be the primary
materials, there are specialized components and products factors that determine technical superiority, rather than
that are not commodity products. who is the first to develop a particular technology.
This reduction is driven in part by the exponentially This suggests that deeper awareness of the differing
rising cost of state-of-the-art fabrication facilities, which processes and timescales for hardware and software
places a premium on volume production. In turn, this development must be part of the design and procurement
limits the economic incentive for any company to process. Semiconductor design and fabrication, as well
respond to the defense needs for specialized devices-- as subsequent integration of fabricated chips, have a
for example, the capacity to design and fabricate substantial lead time. Although it is possible to develop
radiation-resistant integrated circuits (ICs). Further, the portions of new software systems with simulators and
concentration of design and production to a small group emulators, integration and complete testing is dependent
on hardware availability. Thus, the overall time to
deployment of new hardware and software systems will
Trusted_Integrated_Circuits_%28TRUST%29.aspx, last accessed be especially critical when the software requirements for
on February 7, 2012) that seeks to "provide trust in the absence of
a `trusted foundry'."
6 7
As an example, the shortages of disk drives and flash memory Annual Industrial Capabilities Report to Congress, 2008
resulting from the Japan and Thailand natural disasters affected (Washington D.C.: Office of Under Secretary of Defense Acquisi-
many devices and vendors. See http://www.isuppli.com/ tion, Technology and Logistics Industrial Policy, February 2008).
8
Home-and-Consumer-Electronics/News/Pages/IHS-iSuppli-News- An ongoing NRC study, Ethical and Societal Implications of
Flash-Thailand-Flood-Spurs-Nearly-4-Million-Unit-Shortfall-in- Advances in Militarily Significant Technologies that are Rapidly
PC-Shipments-in-Q1-2012.aspx Last accessed on February 7, Changing and Increasingly Globally Accessible is exploring these
2012. issues.
OCR for page 52
52 THE GLOBAL ECOSYSTEM IN ADVANCED COMPUTING
defense missions must be developed based on predicted, design firms without fabrication capabilities, along with
rather than current hardware. This further emphasizes the the rise of system-on-a-chip (SoC) ecosystems are
importance of hardware-software co-design and rapid increasingly enabling new companies and enterprises to
testing and of drawing lessons from consumer device offer devices that compete with the traditional x86-
deployment. oriented PC ecosystem.
The expected length of the life cycle for consumer In both the x86 and ARM SoC ecosystems, some
devices continues to decrease; for instance, the elements of each SoC are likely to be common (for
replacement time for smartphones is now less than a example, general-purpose cores); others will be tailored
year. Comparatively slow and cumbersome Department to specific applications (for example, cryptography
of Defense (DOD) procurement and deployment cycles blocks, media encoders and decoders, digital signal
mean that units may lack access to current-generation processors, or network interfaces) and drawn from an
technology. Defense organizations must balance rapid array of internationally available and licensable silicon
adoption for commodity technologies against more design blocks. This mix-and-match model, now
measured and careful integration and deployment of prevalent in the mobile device space, challenges the
devices and technologies that are unique to defense traditional software development and maintenance
needs. Risks increase when applying the same process model, where legacy software could execute unchanged
and evaluation to both without distinguishing the risks (often without recompilation) as described in Chapter 1.
and benefits.9 At the same time, proven technology-- A DOD shift to application-tailored classes of chips will
even if it is not the most current--may provide better require software refactoring and optimization for each
results with cost-effective performance. Managing these new class of chips, each with different functionality,
tensions suggests that requirements and designs should adding complexity to the software design and
be based not just on current technology but on maintenance life cycle. Unless the software design
projections of technology available two or even three process and toolset for distinctive defense software is
generations ahead. adapted to this shift, the useful lifetime of the chips will
be determined by software availability, not hardware.
4.5 Rise of a New Post-PC Paradigm Driven by In addition to the rise of a new and complementary
Mass ICT Consumerization COTS ecosystem, the consumerization of ICT has
profound implications for how organizations manage
One area in which COTS has become the principal their own ICT. The proliferation and popularity of new
technological driver is in the ongoing consumerization of device functionality challenges traditional approaches to
ICT and the emergence of what might be called a post- organizational technology uptake. Consumers drive
PC technological paradigm. Smartphones, tablets, cloud- adoption of technology in large organizations by forcing
computing capabilities, and other related commercial central ICT organizations to respond to consumer
technologies are the hallmarks for this new era. Industry acquisition outside the organization. This socially
projections10 suggest there could be as many as 50 activated disruption changes the planning and
billion devices connected to the Internet within a decade. deployment of software and services. The DOD is not
Global sales of mobile phones now exceed those of PCs, immune to this effect. As the perceived and actual
and the Chinese phone market alone exceeds that of the differences between commodity technology availability
United States or Europe.11 For much of the world's and centrally mandated deployments rises, individuals
population, a phone is the primary computing device. and groups may circumvent best policies and practices in
More generally, low-power designs, based on system security and information flow in order to access
licensable components and created by semiconductor improvements in functionality.12 In addition, the
proliferation of mobile devices with personally
9
identifiable data and institutional data brings information
A 2009 NRC report, Achieving Effective Acquisition of Infor- leakage risks due to the possibilities of device loss and
mation Technology in the Department of Defense calls for the
DOD to acquire information technology systems using a funda-
theft.
mentally different acquisition process based on iterative, incre-
mental development practices.
10
CISCO White Paper, 2011, The Internet of Things: How the
12
Next Evolution of the Internet is Changing Everything. Available A recent NRC report, Toward Better Usability, Security, and
at http://www.cisco.com/web/about/ac79/docs/innov/IoT_IBSG_ Privacy of Information Technology, examines some of the com-
0411FINAL.pdf. Last accessed on February 7, 2012. peting motivations for users of technology and identifies research
11
See http://www.strategyanalytics.com/default.aspx?mod= opportunities and ways to embed usability considerations in de-
reportabstractviewer&a0=6871. Last accessed on February 7, sign and development related to security and privacy, and vice
2012. versa.
OCR for page 53
IMPLICATIONS FOR U.S. NATIONAL SECURITY 53
Similarly, the rise of big data and rich data analytics, curiosity-driven, smaller-scale research.13 A diverse
made possible by the proliferation of these inexpensive portfolio of research was supported by multiple agencies
networked devices and by massive cloud data centers, in the Networking and Information Technology Research
are challenging traditional notions of computing. The and Development (NITRD) Program over many years.
balance of value is shifting from isolated devices and This era of diverse funding has undergone changes in
software to capabilities embodied by an integrated recent years. There is now an increasing monoculture of
system of devices, data, and services. A data-rich, research funding for computing research, centered on the
consumer-driven world where data is ubiquitous and funding model of the National Science Foundation
often accessible has profound implications for U.S. DOD (NSF). Because NSF emphasizes single-investigator and
notions of information superiority, privacy, and security. small-group research, it has not historically supported
The increasing diversity and independence of global long-term, large-scale infrastructure for either chip and
supply chains for new generations of COTS devices will system fabrication or compiler and tool infrastructure.
challenge existing approaches to system security. In contrast, China, for example, has been increasing
Demand for software verification of diverse components its R&D investment in advanced computing over the past
with multiparty provenance will increasingly be the decade and appears willing to invest in research aimed at
norm, but thus far verification of even existing systems both incremental and higher-end computer innovation.
remains a challenging research problem. When coupled At the same time, China is investing heavily in the
with device heterogeneity and specialization for training of advanced scientists and engineers at the
performance, verifying functionality and the absence of undergraduate and postgraduate level both at home and
implicit or explicit security backdoors will require new abroad. If these trends continue, the still-wide gulf in the
organizational and software security approaches. educational and R&D capabilities between China and the
United States will narrow.
4.6 New Market-Driven Innovation Centers
4.8 Cybersecurity and Software
The emergence of foreign markets that are larger,
are potentially more lucrative, and have better long-term The growing R&D competitiveness of other
growth potential than in the United States and other countries has potentially far-reaching ramifications for
developed countries also has significant implications for the United States in cybersecurity. The DOD and the
the ability of the United States to shape technological U.S. government cybersecurity strategy depends upon
directions. A shift in the global commercial center of the U.S. commercial information technology sector
gravity (either as the result of a new development or of remaining as the world leader.14 Software development
decreased public or private research investments) may is an increasingly central driver of computing
lead to a shift in the global research and development innovation, whether it is parallel tools and applications
(R&D) center of gravity. For example, this could occur if for new devices or advanced software services and data
international firms are required to locate in these markets analytics running atop massive, highly parallel cloud
to remain competitive, to meet the requirements of data centers.
government regulations in the target markets, and to The interconnected nature of globally designed and
better understand those markets. The availability of manufactured consumer devices contributes to increased
trained and talented researchers and developers, risk of data and software security breaches and makes
particularly in parallel computing, will also affect these clean separation of functions--a traditional tenet of good
placements, as today's devices are dependent on parallel security--ever more difficult. The globalization of this
applications and system software to meet performance software development, as well as state-sponsored cyber-
and functional targets. espionage, raises important software and cybersecurity
questions.15 Cybersecurity may well become a pivotal
4.7 The Future Educational and Research long-term area of competition between the United States
Landscape in Advanced Computing
13
NRC, 1999, Funding a Revolution: Government Support for
In the committee's view, the United States became Computing Research, Washington, D.C.: The National Academies
the leader in advanced computing because of its Press (available online at http://books.nap.edu/catalog.php?record
_id=6323).
significant and sustained investment in long-term basic 14
Office of the Deputy Assistant Secretary of Defense for
research, especially its combination of risky, big bets, Industrial Policy, 2010, DoD Cyberstrategy: Leveraging the In-
some of which had significant financial returns, and dustrial Base, December.
15
The NRC has a deep portfolio of work on cybersecurity:
http://www.nas.edu/cybersecurity.
OCR for page 54
54 THE GLOBAL ECOSYSTEM IN ADVANCED COMPUTING
and foreign competitors with fast-growing software and heterogeneous multicore SoCs on all computer
industries, most notably India and China.16 systems. This trend and the proliferation of device types
present daunting challenges, especially given the
4.9 Possible Defense IT Outcomes historical hegemony of the United States in mainstream
computing. Barring concerted action involving major
The slowdown in the growth of single-processor technology breakthroughs and a major shift in U.S.
computing performance described in Chapter 1 brought industrial competitive policy, this accelerating
an end to the virtuous cycle of ever-faster sequential innovation shift may open the door to a latecomer
processors coupled with increasingly feature-rich innovation advantage (discussed in Chapter 3).
software built atop a sequential model. Explicit The challenges and the opportunities for the United
parallelism in both hardware and software is now States are in capitalizing on its historical strengths in
required to realize greater performance and desired systems design, engineering, and integration. Defense
functionality. The consequences of this shift are deep systems and their information technology components
and profound for computing and for the sectors of the are often large and complex, with interconnected and
economy that depend on and assume, implicitly or often redundant components. Advanced computing is a
explicitly, ever-increasing performance. From a critical element of such systems, but only one element. If
technology standpoint, this has lead to heterogeneous the United States focuses on nimble and rapid system
multicore chips and a shift to new innovation axes that integration, with designs that emphasize reliability and
include but are not limited to chip performance. In turn, verification, it can continue to build effective defense
these technical shifts are reshaping the computing systems.
industry, with global consequences. Otherwise, the DOD could find itself with deployed
Today, global equilibration, access to standard computing technology that is no better than, or even
hardware Internet protocol (IP) blocks, and open inferior, to its adversaries.17 Such technical parity (or
foundries have lowered the barrier to entry for even inferiority) could occur due to either a loss of U.S.
international competitors, particularly in Asia. As a technological capabilities or the inability to deploy the
result, it is possible that the locus of innovation may shift appropriate new technologies sufficiently rapidly to
further from the United States. Technology limitations maintain a competitive advantage.
are forcing a new ecosystem of mix-and-match IP blocks
16
N. Gregory, S. Nollen, and S. Tenev, 2009, New Industries
from New Places: The Emergence of the Software and Hardware
17
Industries in China and India, Stanford University Press and Further, computing technologies could also potentially be
World Bank, Washington, D.C. manufactured by adversaries.
