Expect a Sea Change in Application and Information Architectures

Intersecting waves of technology are converging: Cloud computing, massive parallel processing, distributed file systems like Hadoop, dynamic provisioning to the cloud, embedded computing, open mobile platforms and APIs, and semantic publishing, to cite just a few. Patty’s Pioneers (technology architects) have been talking recently about how these phenomena will profoundly change the way we do business, use computers and address problems. We already have the ability to send huge volumes of information to the compute cloud to perform long-running, compute-intensive pattern detection. Google works that way. The genome project and many other scientific endeavors are enabled by this capability. You can solve BIG problems. You can offload bursty, compute-intensive applications. You can enable new capabilities like high-frequency trading.

Information, data types, and file formats are changing too. Scientific and technical data isn’t always organized in relational databases or spreadsheets. There are huge flat files, analog files, arrays, sequences, 3D models, and a myriad of data types and patterns that most business IT was never designed to address. We are sensing and detecting real-time changes in everything from weather to ocean pH. We perform chemical analyses, create molecular models, and find interrelationships among everything. We have massive amounts of genetic sequence data, with genome browsers, model organism databases, molecule- or process-specific databases.

Information that we want to share and have customers find and use is increasingly being tagged to accelerate semantic processing. Many of Patty’s Visionaries (customer-centric e-business leaders) have already embarked on large semantic tagging projects. They are convinced that, in their fields of endeavor, important concepts need to pop out of the sea of words and data so that these concepts can be linked to the appropriate behaviors and services. A company name came be linked to news, financial data, personnel information, products, patents, geography, partners, and customers. A chemical formula or a gene sequence can be linked, viewed, and acted upon with a staggering array of tools.

This next generation of distributed computing and pattern matching will unleash a new economy – with new opportunities for value creation.

As I contemplate the tremendous and revolutionary changes that will hit all of our businesses as a result of the new ways in which we can now detect patterns and process information, I am grateful for the magic of the human brain. Our intuition and emotions combine with our sense-making and pattern matching to make magical leaps across disciplines.

This next generation of computing architectures is upon us. It’s time to take the plunge. Say goodbye to Microsoft Word, say hello to N-dimensional arrays and patterns. Say goodbye to PDAs and laptops. Say hello to holograms you can gesture through.

Comments

[Scott Jordan]

I'd like to address rjnerd's comment.

While there are certainly similarities and (no pun intended) parallels, there are many differences between old-timey timesharing and this newfangled cloudy stuff.

o I can already, as a novice user and in the space of a few minutes, create a fully-functioning virtual machine in the cloud that looks and acts just like a server. Using CohesiveFT.com's service, I don't even have to install and configure my software stack. A software venture I might've ginned up with hundreds of thousands of dollars' worth of industrial-strength, high-uptime rack hardware in 1999 can now be implemented on Amazon EC2 for a pittance.

o I can already, as a novice user and in the space of a few minutes, establish a huge server installation in the cloud, performing massively parallel Hadoop database and processing operations. This was, in fact, done by an enterprising New York Times technician to convert 4 TB of archival resources into web-ready .pdfs into 11 million .pdfs overnight... for $240.

And so on. The difference is that timesharing the old-timey way was inflexible and costly and proprietary and localized. Creating applications and infrastructure that customers could access and utilize was Just Not Done. Cloud computing is as flexible as a PC, as powerful as a supercomputer, almost incredibly cheap, instantly available, and accessible to folks around the world, wherever there are bright ideas, and wherever there are customers. It's a crucible for innovation in the way that traditional timeshared computing hasn't been since its dawning days.

[Munish K Gupta]

I fully agree, exciting time are ahead for us, with devices becoming smarter and always on, data accesible from anywhere and anytime, collaboration with fellow user's - will make us all more connected. Also, it will lead/create a divide between have and have not's. Access to information will become one big differentiator in the new digital world.

[daviding.wordpress.com]

Thanks for opening up my vocabulary to "semantic publishing". Unfortunately, your link into the entry in Wikipedia tends to suggest that we're really early in that technology. (I've left comments on the discussion page, and presume that Wikipedia moderators will figure out whether those links should be corrected or deleted).

There isn't an entry in Wikipedia for "semantic tagging", so I tried Google, without getting much enlightenment. I understand the basics of the semantic web from a technical perspective; I'm just trying to figure out if I should change my behaviour in some way to prepare for this future trend.

[rjnerd]

Why do we need a snazzy new term "cloud computing" for something that has for decades (until now) has been called "timesharing".

The display is snazzier than the VT-52 that used to grace my sunroom, the modem is many orders of magnitude faster than my old Vadic, and the editor is less functional (but prettier) than the emacs I used to use, but the architecture, and security issues haven't changed.