Combining AI+HPC for Real-World Breakthroughs

Webinar: July 13th, 2021 @ 11:00am PT / 2:00pm ET - Register Here

In scientific research and new product development, AI and HPC are converging as a promising strategy for unlocking new discoveries and business value. GPUs have now expanded into the world of accelerated computing due to their utility for high throughput and scale when compared to traditional HPC CPU architectures.

Specific techniques like deep learning combined with purpose-built hardware are accelerating computing breakthroughs and producing higher fidelity simulations and models.

Rescale and Google Cloud Platform have partnered together to offer the widest range of Nvidia GPU classes and the CPUs needed to instantly scale up compute intensive workloads. Whether you need to perform computer vision or drug discovery, AI, ML, and HPC practitioners use Rescale for a fast and simple onramp to powerful computing on GCP infrastructure.

Agenda:

• HPC+AI: What is it and what new capabilities does it unlock? [10 min]
• State of HPC+AI today: What are some industry use-cases in production today? [15 min]
• Imagining what’s possible: Promising use-cases for the future [10 min]
• Getting started with Rescale and Google Cloud Platform [10 min]
• Q&A [10 min]

Doğuş Akaydın, Rescale

Doğuş manages a team of engineers responsible for ML workflows and GPU-accelerated virtual desktop solutions. Before joining Rescale he was a research scientist at NASA Ames Research Center for several years, contributing to development of next generation air and spacecraft. He has a PhD degree in mechanical engineering from CUNY and has postdoctoral training experience at Stanford University.

Massimo Mascaro, Google

Massimo is part of the Office of the CTO for Google Cloud and, in this role, he helps Google Cloud customers reimagine their businesses by leveraging the power of AI and the products from Google that enable it. Massimo started in theoretical physics and has a PhD in Computational Neuroscience. In his research work he studied the dynamics of large recurrent neural networks, their connections with attention and the ability of networks to mesh signals from different modalities.