AI at VMware: A History, Present and What’s Next
A few months ago, Michael Gandy and Joel Leichnetz from the Xplorer Group presented a session at VMworld about AI at VMware. We thought it would be interesting to share with the larger community as Artificial Intelligence (AI) has become such a buzzword (and often over used).
A brief history of AI
AI is not a new concept. It is actually pretty old and can be traced back to the Greek Antiquity: Talos (Greek: Τάλως) was a giant automaton made of bronze to protect Europa in Crete from pirates and invaders. Fast forward to 20th century, in 1956, John McCarthy coined the term AI for the first time and the same year, Arthur Samuel (IBM) wrote the first AI engine to play checkers. In 1980, a lot of Venture Capital (VC) money went into AI startup but the technology was not mature enough and a lot of them either failed or pivoted quickly to something more profitable. The Industry entered in what was called the “AI Winter” for almost 20 years, where very few VCs were investing in AI. It didn’t stop researchers such as Yann LeCun to work on new techniques on Machine Learning (ML) and Deep Learning (DL). This was really the breakthrough with the availability of massively distributed compute power that helped the AI come back, and now we hear about it everywhere. A good example is a graph compiled by CNBC showing the increase in the past years of CEOs mentioning AI at their earning calls.
AI Timeline from Antiquity to today
AI at VMware
VMware is an innovation leader and we have been investing in AI and advanced analytics in general for a long time now. As a company, we are looking at solving the hardest problems for our customers today and in the future. To do so, we are leveraging the best techniques available to create the best solutions. In 2011, we started investigating how we could use ML in our product by engaging in a project with MIT. Shortly after, we introduced dynamic threshold and capacity planning in vRealize Operations in 2012 and automatic data consolidation and schema extraction in vRealize Log Insight in 2014. Today, the pace of innovation in AI and ML are accelerating and VMware’s commitment is clear with the acquisition of Apteligent, along with the recent Wavefront and Arkin/vRealize® Network Insight™ acquisitions. VMware continues to provide our customers with solutions that offer deep insights and granular analytics to improve the performance, availability and experience of their digital services.
AI Timeline at VMware
AI is not an end but a powerful tool to solve hard problems. At VMware, we are looking at the best (i.e. powerful yet efficient) techniques to solve our customers problems and depending on the use cases, we will choose the most appropriate Analytic Solution as described in the diagram below.
Data Analytics pattern at VMware
VMware’s Office of the CTO (OCTO) is looking over the horizon: our Research and Academic group is working with top universities with 8 engagements with multiple universities on ML/AI across the globe. The Research group is looking at the “big data for the 99% of enterprises” as David Tennenhouse, our Chief Research Officer, would say. A lot of ML techniques developed today solve very specific problems that are not the problems of most enterprises and hence specific techniques needs to be developed. Read the full article here. Last but not least, our internal innovation programs such as Borathon, hackathons happening all around the world and xLabs, an innovation incubator, are seeing a lot of interest and development in that space.
To stay up to date with our latest news, please follow us on Twitter @vmwOCTO.