The Path from Digital Business to AI EnterpriseOctober 10, 2016
For several years, companies have made large investments to become more digital. The connections and interactions with customers, suppliers, employees, partners and the increasing number of devices coming online has led to a data tsunami. Some companies struggle to survive in this data deluge while others thrive with an insatiable appetite for data. The more data they have, the more they learn and excel.
Leading industry analysts have identified this transition from digital enterprise to AI enterprise as the required path for companies to analyze and visualize large amounts of data, while unleashing the power of AI algorithms on their data. These companies think beyond the traditional business questions and more rapidly respond to the unpredictability of today’s market.
Companies need tremendous computing power to gain faster insight, access dynamic correlations and achieve superhuman knowledge. With NVIDIA GPUs, customers can process data 10-100x faster. That means applications simply run quicker. Fast compute also allows people to visualize 100x more data with a fraction of the hardware in sub-seconds. As they start to harness the advantages of GPUs, deep learning and AI, these companies develop tremendous competitive advantages over their competition.
NVIDIA offers a portfolio of products to help businesses get started. Our DGX-1 AI supercomputer in a box headlines the options. In the chart below, see the tremendous acceleration possibility with DGX-1. NVIDIA accelerated solutions are also available through OEMs and cloud providers to meet the many use cases and form factors customers need.
In the last few weeks, there have been many discussions about the tools and approaches these leading companies are using. Here is a quick summary of some of the great activities around the globe.
O’Reilly AI Conference and Strata Hadoop World
In Jim McHugh’s keynote at the O’Reilly AI conference last week, he shared how deep learning and AI are making our lives better. This will change many industries, but in his talk he focused on self-driving cars and healthcare AI assistants. These deep learning techniques perform the data work for you and unleash the power of AI to detect and diagnose diseases early, regenerate damaged tissues and greatly improve our quality of life. One great example was Zebra Medical’s use of AI to help the millions of people at risk of osteoporosis.
In the The Cube Monday Night Event, we saw the contributions of the growing accelerated analytics ecosystem. MapD, Kinetica, IBM, DataBricks, Bonsai, SQream Technologies, Skymind, Lumiata and NYU joined NVIDIA on stage to share how they are leveraging GPUs in the data center to deliver customer solutions.
- Keynote: Claudio Silva from NYU on AI-driven sports and GPU-accelerated SQL database and Lumiata CEO Ash Damle on AI-driven healthcare.
- Panel 1: NVIDIA GPU-accelerated Spark (DataBricks, IBM, NVIDIA, Skymind)
- Panel 2: NVIDIA GPU-accelerated database, analytics and visualization (MapD, Kinetica, SQream Technologies, Bonsai and NYU)
One notable point of the night was when NYU’s Silva shared that he experienced a 6,000x speedup using NVIDIA GPUs vs. CPU. Some of the other impressive results are:
- Kinetica helped a customer rendering millions of data points in sub-seconds instead a single million in 20-30 minutes.
- MapD was benchmarked at processing 40 billion rows in 200 milliseconds.
- SQream Technologies customers see a consistent 100 times speedup using their technologies and they were able to ingest 24 terabytes of event data and respond in half a second or less.
Kinetica, MapD and Accenture Labs (for Graphistry) also joined us on stage to do live demos at our Strata session. The speed at which they analyzed and dynamically visualized the data was truly impressive. Seeing is believing, thus I encourage you to get a demo of these great offerings.
AI Summit in San Francisco
Meanwhile in San Francisco, at the AI Summit, NVIDIA’s Kimberly Powell gave an impressive presentation on the deep learning AI revolution. She shared how deep learning is solving high-impact problems like healthcare and cars. She explained how we have all been working the last 5-10 years on using digital technology to become a smarter enterprise. We will see deep learning finding its way into business analytics very quickly over the next six months.
NVIDIA CEO Jen-Hsun Huang announced a partnership with SAP, demonstrating how the ecosystem continues to grow. SAP just received a DGX-1 and is super excited to start using it for deep learning technologies and acceleration of its solutions.
This is a new era of computing, AI and AI for analytics on the horizon.
Our analytics world tour is just a beginning. There are more great events coming. For more information, sign up for our upcoming DGX for Analytics webinar on Oct. 18 or come visit us at Gartner Symposium Conference.