Increasing HPC cluster sizes and large-scale simulations that produce petabytes of data per run, create massive IO and storage challenges for analysis. Deep learning-based techniques, in particular, make use of these amounts of domain data to extract patterns that help build scientific understanding. Here, we demonstrate a streaming workflow in which simulation data is streamed directly to a machine-learning (ML) framework, circumventing the file system bottleneck. Data is transformed in transit, asynchronously to the simulation and the training of the model. With the presented workflow, data operations can be performed in common and easy-to-use programming languages, freeing the application user from adapting the application output routines. As a proof-of-concept we consider a GPU accelerated particle-in-cell (PIConGPU) simulation of the Kelvin- Helmholtz instability (KHI). We employ experience replay to avoid catastrophic forgetting in learning from this non-steady process in a continual manner. We detail challenges addressed while porting and scaling to Frontier exascale system.

Dr. Michael Bussmann (DE) kicked off the session with his pre-recorded presentation "Basics of explainable artificial intelligence with applications in PCa". He stated that AI currently mimics human intelligence but to advance, explainable AI needs lots of data to create value. Diverse, big and high-quality data is key to successful AI building, this data needs to be complex and unstructured. According to Dr. Bussmann, great prospects of explainable AI in PCa include, but are not limited to classification of prostate tumours with MRI, PCa detection, Gleason Score grading, risk stratification, lesion detection, biochemical recurrence, and robotic surgery. Dr. Bussmann also shed light on the big potential of synthetic data.

Artificial intelligence is coming to banking -- scratch that, it's already here, but customers may not have noticed. AI is already playing a role in consumers' lives, whether they know it or not. Talking to Siri, looking at recommendations from Amazon or Netflix, or chatting with Google Home about the temperature -- AI is all around us, and we're growing more comfortable with it all the time. That's good, says Arif Ahmed, senior vice president of payments innovation for U.S. Bank, because AI is set to help customers in important ways, and in the not-too-distant future. "Emerging artificial intelligence will improve the customer experience without compromising privacy," Ahmed told Bank Innovation.

Artificial intelligence is coming to banking -- scratch that, it's already here, but customers may not have noticed. AI is already playing a role in consumers' lives, whether they know it or not. Talking to Siri, looking at recommendations from Amazon or Netflix, or chatting with Google Home about the temperature -- AI is all around us, and we're growing more comfortable with it all the time. That's good, says Arif Ahmed, senior vice president of payments innovation for U.S. Bank, because AI is set to help customers in important ways, and in the not-too-distant future. "Emerging artificial intelligence will improve the customer experience without compromising privacy," Ahmed told Bank Innovation.

Asked if a computer will ever be able to give better investment advice than a human, Oliver Bussmann does not hesitate. "I believe it's possible," said Bussmann, who until March was the chief information officer of UBS. Banks' wealth management departments and other investment firms are starting to adopt artificial intelligence. This is different from robo advisers. Those have simplistic, rules-based models -- you give them your age, risk tolerance, goals, and so on and they select a basket of ETFs for you.

Asked if a computer will ever be able to give better investment advice than a human, Oliver Bussmann does not hesitate. "I believe it's possible," said Bussmann, who until March was the chief information officer of UBS. Banks' wealth management departments and other investment firms are starting to adopt artificial intelligence. This is different from the robo-advisers you've probably heard about. Those have simplistic, rules-based models -- you give them your age, risk tolerance, goals, and so on and they select a basket of exchange-traded funds for you.