On June 17, Intel’s oneAPI Workshop concluded on a high note. AI/ML developers, data scientists, AI enthusiasts, AI researchers, and GPU and HPC programmers participated in the three-hour-long virtual hands-on workshop hosted by Analytics India Magazine on oneAPI AI Analytics Toolkit.
The highlight of this action-packed workshop was the breadth and depth of its content that focused on the programming challenges for multiple architectures. The session also provided first-hand exposure to Intel® oneAPI toolkits.
Topics covered in the workshop:
- Overview of Intel® oneAPI AI Analytics Toolkit
- Introduction to the Intel Dev Cloud for testing & porting applications
- Machine learning products & hands-on labs
- Deep learning products & hands-on labs
oneAPI workshop kicked off at 9:30 am with a welcome note by the Developer Marketing Manager- APJ at Intel — Kavita Aroor, explaining the workshop’s key elements and the rules and guidelines of the contests, audience polls, and dedicated network sessions. The welcome note was followed by an introduction to oneAPI by Lakshmi Narasimhan, the Technical Consultant Lead at Intel; Jing Xu, Senior Technical Consulting Engineer, working as an AI specialist within the Intel Software group; and Aditya Sirvaiya, an AI Technical Consulting Engineer in Developer Products support & consulting organisation within Intel Software group. The expert panel provided actionable knowledge on using Intel® oneAPI suite with a hands-on session.
With a brief oneAPI introduction, Lakshmi Narasimhan, the Technical Consultant Lead at Intel, talked about how the increasing workload diversity and heterogeneous computing create various challenges for machine learning developers — like a lack of common programming languages or API and the ability to reuse codes for an available accelerator. He reiterated that cross-architecture programming is becoming key for developers to improve their performance.
“When developers have a code that runs on CPU, it usually isn’t compatible with GPUs or FPGAs — making it a cumbersome process,” said Narasimhan. “oneAPI, built on Intel’s rich heritage of CPU tools expanded to XPUs and other accelerators, has been designed to address this issue.”
To know more about the workshop click here.
The toolkit equips data scientists, AI developers, and researchers with familiar Python tools and frameworks to accelerate their end-to-end data science and analytics pipelines on Intel® architectures. Not only it maximises performance from preprocessing through machine learning but also provides interoperability for efficient model development.
Explaining how oneAPI optimises data-centric workloads, Narasimhan said, it is a unified programming model that includes a programming language, Data-Parallel C++ and libraries to provide enhanced cross-architecture application performance.
Later, Jing Xu and Aditya Sirvaiya showcased how oneAPI AI analytics toolkit provides an alternative to single architecture programming approaches. While Jing introduced the audience to the machine learning products of Intel® oneAPI suite, Aditya provided a hands-on Modin and Intel oneDAL (oneAPI Data Analytics Library) using Census Workload via Scikit-Learn. The duo also spoke about shared and distributed computing tools on heterogeneous hardware platforms, including Intel CPUs, Intel hardware accelerators, and Intel discrete graphics solutions.
To check out the AI Analytics Workshop lab exercises click here.
To bring the attendees up to speed, the workshop also hosted a Live Q&A Session session on Discord, conducted by Intel technical experts. One of the major highlights of the session was the #oneAPI Workshop Contest, designed to test the participant’s understanding of the AI analytics toolkit. Attendees were asked to submit the screenshots of the performance results for both ML and Deep Learning Workflows using the optimised versions of oneDAL and Intel ® Optimization for Tensorflow demonstrated during the workshop on Intel® DevCloud, to stand a chance to win an Amazon Echo Show 5.
The workshop also had interesting polls and fun-filled engagements throughout the session, keeping the attendees glued to the platform. The session concluded with the lucky draw winners taking home Amazon Vouchers.
The oneAPI workshop provided attendees with an opportunity to network with the leading professionals from Intel. At the end of the workshop, all the attendees had the opportunity to share their feedback on the survey link.
Intel is also running an ongoing DevMesh Contest, where developers and creators need to submit their projects using Intel DevCloud & oneAPI toolkit by the 25th of June 2021. Participants need to submit their project on Intel’s DevMesh portal and email their project link to [email protected].
If the projects are using stock Python, stock TensorFlow, etc., it could be replaced by Intel’s distribution of Python, Intel Optimised Tensorflow, oneDAL and other relevant oneAPI components. After replacing the components, the performance can be compared with their stock counterparts, and the same can be submitted for Intel’s evaluation. Top 10 projects will stand a chance to win Amazon Vouchers worth INR 1,500.
Intel is also providing an amazing opportunity to learn the DPC++ Learning Paths. Click here to know more.
To know more about oneAPI AI Analytics Toolkit, click here.
In association with Intel, Analytics India Magazine announced the top five winners of the #oneAPI Workshop Contest.
The participants were carefully evaluated based on their submissions of the performance results for both ML and DL Workflows using the optimised versions of oneDAL and Intel® Optimization for Tensorflow.
Here are the top five winners of the #oneAPI Workshop Contest —
- Abilash C, Strategic Development Analyst at HP
- Ajay Gupta, ASA at Avaya
- Deepak Walia, Senior Engineer at Ericsson
- Deepika Goel, Data Scientist at Technocolabs
- Divyanka Latwal, Data Scientist at ExperienceFlow
We congratulate each winner on their achievement, pushing the boundaries of innovation with optimised ML algorithms and libraries that are part of the oneAPI AI Analytics toolkit.
Each of the winners will receive an Amazon Echo Show 5.