ORB Visualization
(soon)
21st Century Computing and Networking
The core programming languages and operating systems underpinning United States computational and networking infrastructures are based largely on computer science dating from the late 1960s. Consequently, the US economy, US National Security, and American scientific leadership has come to depend on antiquated foundational technology. Because dominant operating systems and programming languages have set the framework for software production, companies must conform or risk possible rejection by the market, and therefore, companies seeking to meet quarterly profit targets find it very difficult to provide anything more than incremental change.
The stock of basic computer research infrastructure built up by Cold War investments has largely been depleted during the 1990s. Universities have been intellectually constrained by legacy software frameworks accompanying low-cost computers. Government has too often downgraded requirements in security and functionality in order to exploit low-cost COTS technology (commodity-off-the-shelf). This vicious circle has locked the US into an evolutionary path that is increasingly narrow, lacking in diversity, low on fundamental innovation, and unresponsive to the ever-growing reliance of society on computers.
American software jobs are being lost to low-wage foreign producers because US software productivity is not high enough to offset wage differentials. Major military weapons systems can become prohibitively expensive due to low software productivity. For similar reasons, flight control systems and other critical software can require up to 10 years to reach certification. Major IT infrastructures can be similarly expensive and difficult to modernize or adapt to changing circumstances. Cumbersome user interfaces make it slow and difficult to adapt computers to changing purposes.
As the US has come to depend ever more critically on computers and networking, our economy and national security have become increasingly vulnerable to attack and disruption because industry-standard hardware architectures, operating systems, and programming languages added security only as an after-thought. Computer viruses and break-ins cost the economy in lost GNP and pose serious risks to national security.
Historically, US government investments have stimulated research into and catalyzed commercialization of revolutionary computer technologies, including magnetic core memory and missile early warning (SAGE) in the 1950s, time-shared computing and displays in the 1960s, packet-based networking and speech recognition in the 1970s. Indeed, the first computers in the 1940s came out of wartime cryptography, and subsequent military R&D yielded radar, sonar, signal processing, satellite imagery and control, avionics and air traffic control, not to mention GPS.
Today, the Federal government spends about $2 billion per year on networking and information technology, which is coordinated across 12 agencies and departments by the National Coordination Office for Information Technology Research and Development (NITRD). Although Federal IT R&D covers computer architectures, software, networking, and human computer interaction, funds are primarily directed at agency-specific applications and generally build upon industry-standard computing bases. Current Federal funding of basic computing research is too limited and too diffuse to create working systems that can transform the foundational infrastructure of computing and networking in America.
Government leadership is now urgently required to fund development of innovative foundational computing technologies beyond the current response range of industry and to revitalize basic research in the fundamental computing & networking sciences.
Policy Recommendations
… Fund a new 10-year $5 billion Strategic Computing Program to develop 21st-century computer architectures, operating systems, programming languages and intelligent user interfaces that enable ultra-high productivity software development, agile software adaptation, and personalized end-user services.
… Fund a new 5-year $2.5 billion Safe Computing Program to develop provably-secure computer architectures, operating systems, and networking infrastructures for national security applications, offering timely, fine-grain control of security.
… Follow Safe Computing with a new 5-year $2.5 billion Computing Research Modernization Program to seed universities and research institutions with new ultra-high productivity computer systems.