Big Data's Rocket Fuel: GPUs Help Revolutionize Customer Behavior Analysis

By Gidon Ben-Zvi

6.21.2017 twitter linkedin facebook

Digital transformation has become imperative for all businesses, small, medium and large. Applying cutting edge analytics and the most up-to-date data is the brick and mortar businesses’ answer to dealing with evolving customer behavior. Indeed, while many enterprises, such as Amazon, are built from the ground up as digital and data-driven organizations, real-world retailers are also keeping on the cutting edge of technological advancement, hoping to retain their market share against technologically-driven upstarts.
 

Fast, Powerful and Cheap: GPUs Power a Revolution

The main ingredient for succeffully combining the benefits of existing networks of stores and their distribution infrastructures with the possibilities brought about by new technologies is Big Data. And if Big Data is the vehicle driving this revolution in customer behavior analysis, GPUs are its rocket fuel. A GPU-based system can accelerate processing by over 100X on many different types of workloads, including analytics and visualization. GPUs and CPUs have effectively combined to provide a new blueprint for the data center.
However, GPUs can handle thousands of threads simultaneously, which is an order of magnitude larger than what CPUs can handle. Moreover, GPUs are more cost efficient and power efficient. Finally, CPUs processing power has essentially stalled, due to heat, power and transistor size.
 

Not Keeping Up: While data is growing at 40 percent annually, CPUs are only increasing by 20 percent a year, as measured in raw FLOPS
(Source:
https://www.wirelessweek.com/article/2017/03/what-gpu-and-why-should-telcos-care).

 

Big Data Analytics: Helping Customize the Customer Experience

As today’s customers come to expect a personalized experience when interacting with a business, customer analytics is expected to become the backbone of the customer journey, creating touch points at every level of the funnel and at every moment of interaction. Today, technology experts are predicting that analytics will do more than ever to drive customer satisfaction. As such, GPU-powered Big Data has arrived, in a big way.
According to a recent Zebra 2017 Retail Vision Study, 73% of retailers rate managing Big Data as important or business-critical to their operations.
 

Get Smart or Die: By 2021, at least 75% of retailers anticipate investing in predictive and software analytics for loss prevention and price optimization along with cameras and video analytics for operational purposes and improving the overall customer experience
(Source: https://www.forbes.com/sites/louiscolumbus/2017/03/19/internet-of-things-will-revolutionize-retail/#a1460795e582).

 
Specifically, market-basket analysis, customer segmentation and centralized customer data and intelligence are the top tech initiatives retailers are prioritizing. By 2021, nearly 79% of retailers will be able to customize the store visit for customers as a majority of them will know when a specific customer is in the store. In addition, GPU-fueled Big Data insights can help enterprises understand customer activities to predict potential churn or attrition. For example, some of the indicators pointing to attrition could be cancellation of pre-authorized payments, customer complaints and social media sentiment.
 

Speed, Flexibility and SQream

Powered by GPUs, Big Data vendors like SQream Technologies are making it possible for enterprises to harness next-generation database technology to understand and use multi-terabytes of customer behavior data. The SQream database is powered by NVIDIA GPUs. SQream’s patented software delivers ground breaking flexibility in an interactive business environment, where rapid insights from large datasets are time critical for data-driven business decision making. With minimum cost and infrastructure changes required, SQream DB™ enables businesses to interact dynamically with their digital assets, by easily ingesting, storing and analyzing hundreds of terabytes and beyond in near real-time. SQream DB is designed with simplicity in mind, to fit in almost any IT environment, and usable by anyone with even basic SQL knowledge.
 

Conclusion: Survival of the Most Agile

Enterprises that leverage the power of real-time streaming analytics will become more sensitive, agile and gain a better understanding of their customers’ needs and habits to provide an overall better experience. As a result, Big Data platforms, powered by GPUs, will transform the ability of a data-centric business to identify new revenue streams, save costs and improve their customer intimacy.