Horovod is a distributed training framework for TensorFlow. The goal of Horovod is to make distributed Deep Learning fast and easy to use. The primary motivation for this project is to make it easy to take a single-GPU TensorFlow program and successfully train it on many GPUs faster. Internally at Uber we found that it's much easier for people to understand an MPI model that requires minimal changes to source code than to understand how to set up regular Distributed TensorFlow. If none of these things makes sense to you - don't worry, you don't have to learn them if you use Horovod.
Facebook has figured out a new way to train computer vision models that will massively accelerate the company's work with artificial intelligence. Using the new technique, the company can train an image classification model in an hour while maintaining its accuracy. At peak performance, the new system Facebook laid out in a paper today can train 40,000 images per second using 256 GPUs, without sacrificing the quality of the resulting model. It's an achievement that should help improve the quality of future research by helping data scientists test their hypotheses faster. Accelerated machine vision training is incredibly important for Facebook, which sees augmented reality and machine learning as key to its future business.
Intel Capital, the investment arm of the processor giant, is today announcing its latest tranche of investments, a total of nearly $60 million going in to 15 startups that are working on solving different problems in the bigger area of big data (with a full rundown below). The investments come on the back of a big year for the group: In 2017 so far, Intel says that it's invested $566 million in startups in its portfolio. The focus on big data in this latest group of startups comes out of a new turn for Intel and how it's been making strategic investments in recent times. Intel Capital is one of the bigger names when it comes corporate tech investing. In total, it has invested $12.2 billion in 1,500 companies since 1991.
This is a community blog and effort from the engineering team at John Snow Labs, explaining their contribution to an open-source Apache Spark Natural Language Processing (NLP) library. Apache Spark is a general-purpose cluster computing framework, with native support for distributed SQL, streaming, graph processing, and machine learning. Now, the Spark ecosystem also has an Spark Natural Language Processing library. Get it on GitHub or begin with the quickstart tutorial. The John Snow Labs NLP Library is under the Apache 2.0 license, written in Scala with no dependencies on other NLP or ML libraries.
In a blog post today, Intel (NASDAQ:INTC) CEO Brian Krzanich announced the Nervana Neural Network Processor (NNP). The Intel Nervana NNP promises to revolutionize AI computing across myriad industries. Using Intel Nervana technology, companies will be able to develop entirely new classes of AI applications that maximize the amount of data processed and enable customers to find greater insights – transforming their businesses... We have multiple generations of Intel Nervana NNP products in the pipeline that will deliver higher performance and enable new levels of scalability for AI models. This puts us on track to exceed the goal we set last year of achieving 100 times greater AI performance by 2020.
Artificial intelligence now fits in our daily lives and is deployed in more and more business sectors, hustling human expertise. Artificial intelligence should transform one job over two, but does not necessarily represent a threat. In fact, these jobs should be redirected to less repetitive tasks, with more added value, discusses According to a PwC study from March 2017, 70% of the jobs in the energy sector and 65% of the jobs in the consumer sector could be automated through artificial intelligence. This new technology involves a necessary change in the value chain and, if it opens the way to new skills like cybersecurity, it also represents a major challenge and opportunity for these businesses. Managers and Top-Levels are directly involved in the facing of this challenge, by accompanying the teams through this mutation: vanquish fears, embracing innovation, transforming businesses, training teams.
In my previous post on the recent Linley Processor Conference, I wrote about the ways that semiconductor companies are developing heterogeneous systems to reach higher levels of performance and efficiency than with traditional hardware. One of the areas where this is most urgently needed is vision processing, a challenge that got a lot of attention at this year's conference. The obvious application here is autonomous vehicles. One of the dirty secrets of self-driving cars is that today's test vehicles rely on a trunk full of electronics (see Ford's latest Fusion Hybrid autonomous development vehicle below). Sensors and software tend to be the big focus, but it still requires a powerful CPU and multiple GPUs burning hundreds of watts to process all this data and make decisions in real-time.
From robot workers to drone deliveries, Amazon is known for using innovative and futuristic technologies. But the latest patent suggests that the firm wants to take things further, and use technology to track its workers. The patent reveals designs for ultrasonic wristbands that could be used to monitor workers' performance. The latest patent suggests that Amazon wants to take things further, and use technology to track its workers. The patent reveals designs for ultrasonic wristbands that could be used to monitor workers' performance The patent describes'ultrasonic tracking of a worker's hands' that would be used to'monitor performance of assigned tasks.'
Switching to a new language is always a big step, especially when only one of your team members has prior experience with that language. Early this year, we switched Stream's primary programming language from Python to Go. This post will explain some of the reasons why we decided to leave Python behind and make the switch to Go. The performance is similar to that of Java or C . For our use case, Go is typically 30 times faster than Python.
Infosys, a global leader in technology services and consulting, is aiming to reinvent the way people consume sport using extensive player data. The Indian firm, which had revenues of $9.5 billion in its last financial year, demonstrated its'Infosys Information Platform (IIP)' during the recent ATP Tennis tournament in London, of which it was a headline sponsor. Speaking to Access AI, the firm's head of energy and services for Europe Mohamed Anis, who joined in 2000, said Infosys uses machine learning to analyse historical data on player performance, which in turn is able to predict behaviour, shot selection, and even a probabilistic outcome of the match itself. Anis (pictured) said the data is delivered in real time and can be used to help spectators view the game/match on an entirely different level – comparable to that of the coach. "Tennis has been around for a very long time," explained Anis.