SLAC's Ryan Coffee Talks Ultrafast Science

Editor’s note: This is part of a series profiling researchers advancing science in the expanding universe of high performance computing.

Ryan Coffee, senior research scientist at SLAC National Accelerator Laboratory at Stanford, blows things up for a living.

Well, almost. He studies particle physics. But Coffee jokes, “This sounds like fancy X-ray diffraction and molecular shape and motion, but really, we’re blowing [molecules] up really fast.”

Coffee spoke with AI Podcast host Noah Kravitz about his work at SLAC as well as his upcoming GTC Digital session, EdgeAI: Turning Data into Information from the Point of Production Toward an HPC Ecosystem.

Key Points From This Episode:

In his GTC Digital talk, Coffee will discuss how the next generation of X-ray free electron lasers will produce massive quantities of information, which will need to be processed as close to the sensor as possible.
Coffee took the time to answer questions from Kravitz’s friends and family about everything from data reduction to the implications of capturing and processing data on a scale of terabytes per second.

Tweetables:

“We don’t have time to run full-blown calculations and simulations, but we do have time to run inference engines.” — Ryan Coffee [16:58]

“As our detectors become very flexible and can handle errors in interesting ways, and they can start to interpret what humans are doing — I think this is a much brighter future of a partnership between artificial or augmented intelligence and human creativity.” — Ryan Coffee [33:50]

Astronomers Turn to AI as New Telescopes Come Online

Astronomers are seeing further into space than ever before, but they’re also seeing more data than any human can handle. To turn vast quantities of data into world-changing scientific discoveries, Brant Robertson, a visiting professor at the Institute for Advanced Study in Princeton and an associate professor of astronomy at UC Santa Cruz, is turning to AI.

How Deep Learning Can Accelerate the Quest for Cheap, Clean Fusion Energy

This episode’s guest, William Tang, is a principal research physicist at the Princeton Plasma Physics Laboratory. He explains how deep learning is enabling big-data-driven discovery in endeavors such as the quest for clean, cheap fusion energy.

Making “Iron Man” Interface Real: AI-Based Virtualitics Demystifies Data Science with VR

Virtualitics, an AI-based analytics platform, is bringing creativity to data science through machine learning and immersive virtualization. Virtualitics Machine Learning Projects Head Aakash Indurkhya speaks about how VR can be instrumental in the field.