Flickr

Microsoft (NSDQ:MSFT), supercomputer maker Cray (NSDQ:CRAY), and the Swiss National Supercomputing Centre (CSCS), a world-class scientific computing center, announced the results of a deep learning collaboration that expands the horizons of running Artificial Intelligence (AI), Machine Learning (ML), and in particular Deep Learning (an especially promising branch of ML) applications, using Microsoft software and Cray supercomputers.

Under the leadership of CEO Satya Nadella, Microsoft continues to be among the most forward-looking and pragmatic tech companies lately. With MSFT stock trading near all-time high levels, the above announcement holds immense strategic significance to Nadella's vision for Microsoft. Let's take a closer look.

The collaboration and its first results were unveiled at the 2016 Neural Information Processing Systems (NIPS) Conference in Barcelona. The three partners adapted and scaled the Microsoft Cognitive Toolkit - a free open source toolkit that trains deep learning algorithms to learn like the human brain - to run efficiently on more than 1,000 NVIDIA Tesla P100 GPU accelerators on the Cray XC50 supercomputer at CSCS.

According to the Cray news release, the computing power of the XC50 shortens the training time from weeks to hours and allows users "to solve a whole new class of problems, such as moving from image recognition to video recognition, and from simple speech recognition to natural language processing with context."

“Applying a supercomputing approach to optimize deep learning workloads represents a powerful breakthrough for training and evaluating deep learning algorithms at scale,” said Dr. Xuedong Huang, distinguished engineer, Microsoft AI and Research. “Our collaboration with Cray and CSCS has demonstrated how the Microsoft Cognitive Toolkit can be used to push the boundaries of deep learning.”

According to Microsoft, leading teams such as Skype, Cortana, Bing, Xbox, and industry-leading data scientists already use the Microsoft Cognitive Toolkit, built with sophisticated algorithms and tools for Big Data, to develop commercial-grade AI applications. The XC50's Aries network interconnect, with optimized node-to-node communication performance, contributes to speeding up the Microsoft Cognitive Toolkit's operations.

Microsoft seems to be pivoting to a new business model based on cloud computing, and high profile government and corporate clients are signing up. Next-generation Microsoft cloud services include AI applications. "[We] are building out Azure [Microsoft's cloud computing platform] as the first AI supercomputer," said Microsoft CEO Satya Nadella. It seems likely that the partnership with Cray could permit advancing this key strategy of the Redmond giant. In related news, Microsoft seems poised to become a leader in the emerging high-tech sector of quantum computing, always with an eye to the cloud.

AI, ML, and Deep Learning applications can efficiently leverage the power of massively parallel supercomputers. The XC50, unveiled in November at the 2016 Supercomputing Conference in Salt Lake City, is Cray's fastest supercomputer. According to the company, the machine can reach a peak performance of one petaflop - a thousand trillion floating point operations per second - in a single cabinet. Fusing high-performance computing with deep learning is part of Cray’s vision of the convergence of supercomputing and big data.

“Only Cray can bring the combination of supercomputing technologies, supercomputing best practices, and expertise in performance optimization to scale deep learning problems,” said Dr. Mark S. Staveley, Cray’s director of deep learning and machine learning. “We are working to unlock possibilities around new approaches and model sizes, turning the dreams and theories of scientists into something real that they can explore. Our collaboration with Microsoft and CSCS is a game changer for what can be accomplished using deep learning.”

“Cray’s proficiency in performance analysis and profiling, combined with the unique architecture of the XC systems, allowed us to bring deep learning problems to our [system] and scale them in a way that nobody else has,” added Prof. Dr. Thomas C. Schulthess, director of the Swiss National Supercomputing Centre (CSCS). “What is most exciting is that our researchers and scientists will now be able to use our existing Cray XC supercomputer to take on a new class of deep learning problems that were previously infeasible.”

Putting It All Together

This partnership between Microsoft and Cray is likely to permit enhancing the performance of Microsoft's strategic cloud computing arm, thus giving a good positive contribution to Microsoft's stock in the medium and long term. At the same time, the partnership with Microsoft is likely to boost Cray's stock much sooner, because Cray is a smaller company uniquely focused on the supercomputing sector.