Table of contents for IBM Big Data Analytics Analyst Insights 2013
- Driving Transformation with Big Data Analytics #BDA13
- Business Value of Big Data and Analytics #BDA13
- Big Data and Analytics: Fueling Competitive Advantage #BDA13
- Using Analytics for Competitive Advantage #BDA13
- Big Data and Analytics Use Cases #BDA13
- IBM Watson Engagement Advisor #BDA13
- An IBM Innovation Panel #BDA13
- Enhancing the Client Experience #BDA13
- Addressing the Analytics Skills Gap #BDA13
- Big Data Inspires Analytics Innovations at IBM #BDA13
- Speed of Analytics: Why Infrastructure and Platforms Matter #BDA13
- IBM Solutions for Insight Driven Marketing #BDA13
- IBM Counter-Fraud Point of View #BDA13
- Closing thoughts on IBM Big Data Analytics #BDA13
Now time for a panel of IBMers focused on innovation moderated by John Hagerty. Some topics from the panel:
- IBM Research
A big part of innovation at IBM is their focus on research. IBM research creates strategic initiatives driven by a Global Technology Outlook. This same technology outlook drives some “big bets” on things like cloud or smarter cities. The famous grand challenges, like Deep Blue and Watson, as well as a focus on “first of a kind” projects like the streaming data in neo-natal care example from earlier round out the drivers. All of this results in projects in both the software and services parts of IBM. Examples of topics from the Global Technology Outlook include data and analytics back in 2005 and more recently a focus on petascale (big data) analytics.
The drop in price of memory and increase in size of memory are real game changes for systems because a memory-based approach offers dramatic improvements in performance. But it requires careful architecture to drive the right data and the right processing into memory and off disk.
- Scaling analytics
Another area is a focus on larger numbers of decision-makers as analytics get driven down to the front line. This results in increase demand for scale. IBM is looking into various things like FPGAs (Field Programmable Gate Arrays) that offer another layer of parallelism while BLU also uses Single Instruction Multiple Data chips where a single instruction can be executed against multiple rows.
There’s a lot of research into the consumability of visualizations, how to visualize data in ways that can be easily consumed by users. There’s also an interest in the consumability of analytical services, especially given the challenges in hiring enough analytic expertise.
One of my favorite things about IBM is its continued focus on and investment in research. They do a nice mix of longer-term projects with a long time horizon (5-10 years) and a commensurate failure rate with low-risk, practical product-focused research targeted on the next couple of years. As always with research and innovation at IBM there is almost too much to discuss…