Microsoft Research

Poker has been a challenge problem in game theory, operations research, and artificial intelligence for decades.

We discuss the history and design tradeoffs for large-scale topologies in high-performance computing.

We introduce a fully online model of maximum cardinality matching in which all vertices arrive online. On the arrival of a vertex, its incident edges to previously-arrived vertices are revealed.

Episode 41 | September 12, 2018 Dr. Nancy Baym is a communication scholar, a Principal Researcher in MSR’s Cambridge, Massachusetts, lab, and something of a cyberculture maven.

The New York City subway is the largest rapid transit system in the world, serving approximately 5.5 million riders each day.

This project proposes a combination of new secure hardware for acceleration of machine learning (including custom silicon and GPUs), and cryptographic techniques to limit or eliminate information

Data-driven techniques help boost agricultural productivity by increasing yields, reducing losses and cutting down input costs.

We propose a platform to validate networks with high-fidelity emulations.

A reliable, performant, and low-cost network is a critical part of the cloud infrastructure. Our research aims to re-invent the next-generation network via emerging optical technologies.

We demonstrate Platform for Situated Intelligence, an open-source, extensible framework that enables the development, fielding, and study of situated, integrative-AI systems.

Microsoft has established itself as a pioneer in trustworthy computing. Features such as Always Encrypted in SQL Server show how we continue to keep trust while making world-class products.

Wolong: A Backend Optimizer for Deep Learning Computation and Open PAI Chair: Jilong Xue This talk presents Wolong, a backend optimization system for accelerating deep learning computation on

Ad-hoc audio clips, such as those from smart speakers, social media apps, security cameras and podcasts, are being recorded and shared online on a daily basis.

We are witnessing a huge surge in efforts and interest in developing machine-learning (ML) based solutions for optimizing large-scale networked systems.

FaRM is an in-memory transactional object store for random-access latency-sensitive workloads, such as graphs and key-value stores.

Computation, unlike mathematics, is a physical process that takes time, energy, and space.

Advanced technology in software and hardware that interacts with humans can radically benefit from a team-based approach to the research and regulatory process.

Our ability to produce data continues to grow exponentially, and so does the computational power available in the world.

As algorithms reach ever more deeply into our daily lives, increasing concern that they be “fair” has resulted in an explosion of research in the theory and machine learning communities.

Raw bit errors are common in NAND flash memory and will increase in the future. These errors reduce flash reliability and limit the lifetime of a flash memory device.

Our results are: ** Introduce the problem of finding stable matchings that are robust to errors in the input.

Is matching in NC, i.e., is there a deterministic fast parallel algorithm for it?

Matrix completion is a well-studied problem with many machine learning applications. In practice, the problem is often solved by non-convex optimization algorithms.

Episode 40 | September 5, 2018 Data centers have a hard time keeping their cool. Literally.

Laura Haas, Dean, College of Information and Computer Sciences, UMass Amherst attended the Microsoft Research Faculty Summit 2018.

Computer networks can be represented by (marked) point processes communicating information between nodes.

In an ideal world, there would be infinite computing resources. These resources would be free and sustainable, with no impact on the future of our planet.

Data science techniques have an important role to play in the next generation of cyber-security defenses.

We give a constant-factor approximation algorithm for the asymmetric traveling salesman problem.

Edge computing is a natural evolution of cloud computing, where server resources, ranging from a credit-card size computers to micro data centers, are placed closer to data and information