By David Albright and Jodi RudorenPublished 11:51 PM, July 08, 2019I was not expecting to be a spy.
I was a programmer.
And my career in the computer industry was the pinnacle of my ambition.
I had the privilege to work on the original IBM Watson system for IBM, and its successors, in the early 1990s.
My time there was extraordinary.
In my new job at IBM, I had a new responsibility: to design and maintain a toolset for IBM’s new Watson platform, which was to become the foundation of the new generation of intelligence and supercomputing technologies.
The challenge I faced then was that Watson was the first machine to ever truly do the work that computers do, and I was responsible for writing that software.
The work required to do that was extremely complicated.
As a programmer, I was constantly adding new features, adding new data types, and adding new ways to solve problems.
And I was tasked with keeping up with new research and development (R&D) requirements.
As a result, the process was a long one.
But I never complained about it.
In fact, I worked so hard to keep the Watson team focused and focused on the tasks at hand, I even developed a new word, “superiority,” to describe the new platform.
My goal in the Watson job was to build the foundation that would allow IBM’s Watson platform to become a more comprehensive and reliable platform for all the new research that would be done with the next generation of supercomputers.
That was a challenge that I took on full-time and successfully accomplished.
And that work has helped drive the evolution of supercomputer technology and the development of a whole new generation, which includes the next wave of super computers, including the ones that will power our smartphones, which will be used in all kinds of applications for decades to come.
In this series of articles, we will explore some of the challenges that I faced as a programmer in building a supercomputer, from the architecture of the hardware to the design and deployment of the software.
We will cover how Watson’s architecture evolved from the beginning, how it came to be the foundation for a new generation supercomputer architecture, and how Watson has developed over the past 30 years to be capable of running all kinds to do with supercomputation.
We also will explore the challenges and successes of the development and deployment process, which has been instrumental in making Watson a powerful and reliable system.
We will also talk about some of Watson’s current challenges and its future.
In the first part of this series, we explore the architecture and the software used to build Watson.
In part two, we discuss the architecture, how the architecture was built and how it was deployed, how IBM built and deployed it, and what we can learn from its success and how we can do it better.
In the final part, we look at some of IBM’s current problems and how its software could be improved.