HPE DISCOVER: Explore the promises and practicalities of AI

[Curt_Hopkins]

Ashkan Seyedi, Natalia Vassilieva, Sergey Serebryakov, and Thomas Van Vaerenbergh

By Curt Hopkins, Managing Editor, Hewlett Packard Labs

Hewlett Packard Labs’ AI scientists will be offering a number of sessions around the topic of artificial intelligence, all of which are worth attending.

What is the right hardware for (the next generation of) Artificial Intelligence? (T4787)

Labs’ head of AI, Natalia Vassilieva, Ashkan Seyedi, photonics research scientist, and Thomas Van Vaerenbergh, photonics research engineer, will discuss hardware requirements for AI applications of today and tomorrow. Do you need a specialized AI chip today? Will you need one tomorrow? If so, what should it look like? As AI has become a must-have, the obstacles to proper implementation have become more obvious.

“We’ll talk about what algorithms we have today and what their computational and hardware requirements are,” says Labs senior research manager Natalia Vassilieva.

Van Vaerenbergh will discuss HPE’s research program on unconventional accelerators and explain how to leverage optical accelerators to solve different optimization problems (training of neural networks being one of them). Seyedi will explain how optical interconnects can fit into the overall picture of specialized hardware for AI.

What is HPE Deep Learning Cookbook and how do you use it? (B4786)

Vassilieva, Seyed Mirsepassi, head of AI for HPE, and Glyn Bowden, chief architect for AI and data science practice lead for HPE Pointnext, will introduce attendees to the HPE Deep Learning Cookbook – a set of tools to drive optimal hardware/software stacks for deep learning workloads.

“We’ll illustrate what an AI journey for a customer could look like,” says Vassilieva, “as well as explaining how HPE’s Pointnext consultancy compliments the Cookbook toolset developed at Labs.” 

Both technical experts and managers will learn how the complexity of applying AI to specific goals can be mitigated with Pointnext advisory services and the Cookbook.

HPE Deep Learning Cookbook: Get your technology recipe to build a deep learning system (DF4912)

Labs’ senior researcher Sergey Serebryakov give attendees a hands-on introduction to the HPE Deep Learning Cookbook. He will explain how different deep learning models have different requirements and how you can employ the DLC to sketch out the best strategy for your goals.