Intel sponsored this post.
Machine learning models are designed to be resilient, flexible and meet business goals but often, engineers who build the product and algorithms face obstacles to ensure that it works reliably, quickly and at scale. Frameworks are not easy to use. As many of the world’s leading organizations embrace approaches to scaling machine learning, Intel is offering the tools, applications and hardware to make it easier for developers to build, deploy and manage artificial intelligence and machine learning models that can be used by tens of thousands of people instead of just a few.
Introduced in 2018, Intel launched oneAPI, an open, standards-based, unified programming model that makes it easier for developers to write software that can take advantage of all the unique capabilities presented by different chip architectures. It’s a way to standardize central processing units (CPUs), graphic processing units (GPUs), field-programming gate arrays (FPGAs) or any class of processors. This model, together with recently launched Intel oneAPI toolkits, are making it easier for machine learning engineers to invoke multiple classes of processors that can accelerate application workloads with large data sets and high complexity.
Join us on Feb. 10 at 9 a.m. PT for a live Day of Machine Learning with Intel discussion, where we’ll dive deeper into oneAPI. We’ll explore the software at scale issues with machine learning and the hardware needed for it. We’ll look at the tools and the infrastructure that is used for developing, deploying and managing the algorithms. We’ll also dive into questions around how Intel’s oneAPI toolkit is a way to resolve problems that teams face, and how oneAPI fits with existing frameworks such as PyTorch and TensorFlow.
The Line Up:
9 a.m. PT: What is oneAPI? The Software Tool Gap: A Roundtable Discussion
The first session will feature a roundtable discussion with Saumya Satish, AI product manager for architecture graphics and software at Intel, Rachel Roumeliotis, vice president of AI and data at O’Reilly Media and Janakiram MSV, analyst, advisor and architect at Janakiram & Associates. We’ll explore how you get better performance out of your machine learning frameworks, and how to optimize the pipeline to increase performance. We’ll also discuss the latest market trends and key drivers of cross architectural development.
10:30 a.m. PT: How Google Health Uses Machine Learning With Intel
The second session will feature an insightful discussion with AG Ramesh, Principal Engineer, Deep Learning Software at Intel. In this 10-minute lightning discussion, we’ll hear about how Intel works with Google in the areas of machine learning, and how Google uses that work with their customers in areas like Google Health.
11:30 a.m. PT: Democratize AI with OneAPI
In this third session with Raymond Lo, OpenVINO Edge AI Software Evangelist at Intel, we’ll look at the complexities of hardware and software platforms versus vertically integrated models to scale machine learning projects. We’ll explore how oneAPI and open source machine learning technologies such as OpenVINO can be applied to address real-world problems such as climate modeling, economics, and health care.
2 p.m. PT: Machine Learning and the Future of the Data Center
Join us in the afternoon for a discussion with Meena Arunachalam, Principal Engineer, at Intel focused on machine learning performance, and Paul Teich, Director of Product Management-Infrastructure at Equinix and as we look at scaling machine learning in the new modern data center. We’ll look at how machine learning has historically been accomplished, the challenges of scaling and the different approaches and roles within an organization.
Want to catch up on Intel’s machine learning journey? Check out this article on KDNuggets about performance gains that can be achieved through software optimization: Showcasing the Benefits of Software Optimizations for AI Workloads on Intel® Xeon® Scalable Platforms.
What questions would you like to explore in our upcoming Day with Intel? Tweet us your questions in advance to @TheNewStack.
The New Stack is a wholly owned subsidiary of Insight Partners. TNS owner Insight Partners is an investor in the following companies: Real.