Computing has evolved through a series of distinct architectural eras: mainframes, distributed computing, client-server, Internet, and cloud. Each offered a new way of organizing information and connecting people. However, the volume (and velocity and variety) of information is exploding. Today a PC generates only about 90 megabytes of network data per day, but in the not too distant future, a connected autonomous vehicle will produce many terabytes and a connected automated factory will generate over a petabyte. We need more than simply new ways to organize and connect to information, we need new ways to uncover the hidden insights within and to harness the full potential of machines. These paradigms are leading us to the next big wave of computing, which is analytics and artificial intelligence (AI).
Artificial intelligence isn’t a new concept, but the hype it generated in the past didn’t deliver on promises by early proponents. However, AI techniques are being increasingly embedded within many applications, and we expect this trend to accelerate going forward. We are experiencing a convergence of forces that make AI necessary and enable its success: the explosion of data, the computational capacity to handle it, and business and societal demand to drive it forward. Today I want to tell you about what we’re doing at Intel in three critical areas to fuel this computing transformation.
First, Intel has the broad portfolio of products required to deliver end-to-end AI solutions, delivering breakthrough AI computational capability from datacenter to the edge of the network. In the data center, Intel brings to bear massive amounts of computing power specifically tuned to AI applications. To reduce time to train machine algorithms, the Intel® Xeon Phi™ processor is a bootable x86 host CPU that “on-loads” machine learning processing right onto the processor and further boosts performance by scaling out efficiently to many cores and nodes. When it comes time to analyzing new data using a trained model, commonly known as inference, the Intel® Xeon™ E5 processor family is a high throughput processor that can assess data in real-time. Both processors are general purpose CPUs, which reduce TCO, because they can run any other x86 workload in the datacenter. In addition, at the network’s edge—where the data resides and decisions must often happen in real-time – Intel’s portfolio of CPU products enables the seamless distribution of these machine models for inference.
Second, Intel is reducing the complexity that stands in the way of adoption, in order to make AI more accessible to businesses and institutions. One aspect is to create software tools and libraries that let data scientists focus more on generating insights and less on the drudgery of data preprocessing and tuning. Intel provides an optimized version of the Intel® Math Kernel Library (including an open source Deep Neural Network version called MKL-DNN) and is in the process of optimizing leading machine and deep learning frameworks for CPU performance. Intel also recently made available the Intel® Deep Learning SDK to help to set up, tune and run deep learning algorithms, and is leading the Trusted Analytics Platform project that simplifies analytics and traditional machine learning techniques. Finally, Intel is training record numbers of developers on AI, in order to broaden access and grow the talent pool that can deploy these powerful tools going forward.
Finally, we recognize that accelerating the promise of AI takes a village—an open ecosystem in which many minds participate, collaborate and contribute. As a leading sponsor of open software, Intel is contributing heavily to open-source communities in this field, including optimizations to aforementioned libraries, frameworks and the Trusted Analytics Platform (TAP).
The information explosion is leading to amazing new possibilities through analytics and AI, and the potential impact on industry and society is immense. At Intel, we believe in accelerating progress and making the power of technology more accessible to all. We are committed to fueling the analytics and AI computing era, and we hope you’ll join us.
To learn more about AI at Intel visit: http://www.intel.com/ai
To learn more about Analytics at Intel visit: http://www.intel.com/analytics/