By Ryan Martin | 4Q 2018 | IN-5237

If the twentieth century was all about the mechanization of physical work, the twenty-first century is all about the mechanization of mental work. We can think of “mental work” as the process of thinking that unfolds when making difficult decisions. These decisions are ultimately what establish a state of understanding or belief. In computing terms, this state of understanding or belief is manifested in cached memory. The challenge of working with cached memory in the world of IoT is heterogeneity—of endpoints, applications, data, and suppliers—and the resulting need to shift from physical-time to logical-time order computing, which prioritizes the relative order in which data is generated, transmitted, and/or processed (think blockchain).

Registered users can unlock up to five pieces of premium content each month.

Log in or register to unlock this Insight.

 

From Physical- to Logical-Time Order Computing

NEWS


If the twentieth century was all about the mechanization of physical work, the twenty-first century is all about the mechanization of mental work. We can think of “mental work” as the process of thinking that unfolds when making difficult decisions. These decisions are ultimately what establish a state of understanding or belief. In computing terms, this state of understanding or belief is manifested in cached memory. The challenge of working with cached memory in the world of IoT is heterogeneity—of endpoints, applications, data, and suppliers—and the resulting need to shift from physical-time to logical-time order computing, which prioritizes the relative order in which data is generated, transmitted, and/or processed (think blockchain).

The Instrumentation of Search

IMPACT


When a person performs a search on Google, they aren’t searching the web; they are searching an index of the web. These search programs typically start by fetching a few webpages. Then they fetch the pages that link to those pages, and the pages that link to those pages, etc., until a sufficiently large portion of the web is indexed. When a search is performed, the software sorts through all possible results and then ranks the list of potentials by relevance based on a growing number of factors.

The need to optimize search engines materialized due to the explosion of information trafficking the Internet. While personal computers from the 1980s and 1990s were serial processors, modern PCs and smartphones perform multiple operations at the same time. For problems that are extremely complex, such as modeling the earth’s climate or simulating molecular behavior, processing on general purpose hardware takes way too long. Supercomputers solve these problems by batching processes in parallel but have yet to meet the demands of searching the physical web at IoT scale. This is where technologies like Artificial Intelligence (AI) and machine learning come in—to find patterns in vast amounts of existing data–and where extra computing horsepower can be a big help. It’s also where the attention is starting to shift from traditional supercomputing to quantum computing: To find solutions to problems where data does not exist.

The Mechanization of Complex Decision Making

RECOMMENDATIONS


Quantum mechanics employ a fundamentally different set of rules than we are used to (compared to electronics, for example, which deals in volts and currents, rather than polarization)—and this means there is the potential to solve a fundamentally different set of problems. It could be to help researchers create new medicines and materials through a better understanding of molecular interactions; improve supply chain logistics by delivering products with the least amount of fuel, using the most cost-effective mix of providers; optimize the generative design process; or manage risk in constantly fluctuating financial markets. There are a lot of benefits that come with greater access to high-horsepower computing, though the most tangible near-term opportunity is to improve the speed and quality of AI application innovation.

About a dozen companies are already collaborating to create a business case for the current class of quantum computers. The near-term challenge (other than development of the systems themselves) is to offer a compelling enough Minimum Viable Product (MVP) to engage the right centers of influence (e.g., developers, product management, and information technology/operational technology)—the beneficiaries of which include everythingfrom farms (using satellite data to increase crop yields) and ports (using blockchain to improve supply chain management transparency in shipping and logistics) to researchers (working with IBM-Q to find new treatments in medicine), encryption specialists (using high-performance computing to prevent cybercrime), and manufacturing (to speed material engineering as well as the general prototyping and design process).

For more information on the key players in this market and their strategy assessments, see ABI Research’s latest report,Quantum Computing: Core Technologies, Development and Use Cases (AN-4950), as well as ABI Research’s Smart Manufacturing service.

Services

Companies Mentioned