---
title: "ORNL - FP64 Emulation in HPC: Performance, Precision, and Practical Trade-Offs"
image: https://webinars.techstronglearning.com/hubfs/AMD%20-%20ORNL%20-%201080x1080.jpg
---

FP64 Emulation in HPC: Performance, Precision, and Practical Trade-Offs

No video selected

Select a video type in the sidebar.

<iframe allow="encrypted-media" allowfullscreen frameborder="0" height="315" src="https://www.bigmarker.com/recordings/e1d27b3f4e7c?height=315&amp;width=700" width="700"></iframe>

No video selected

Select a video type in the sidebar.

Guy Currier, Chief Analyst of Futurum Media, moderates a discussion with Nick Malaya, Senior Fellow in HPC and Sovereign AI at AMD, and Bronson Messer, distinguished Staff Scientist and Director of Science at the Oak Ridge Leadership Computing Facility, about emulating 64-bit floating-point (FP64) math on GPUs built mainly for low-precision AI workloads. As AI dominates the processor market and chip real estate shifts toward reduced-precision units, emulation methods such as Ozaki have drawn attention for matching or beating native FP64 on certain large matrix operations. The panelists argue, however, that their applicability is narrow: most scientific codes rely on fused multiply-adds and transcendental functions rather than the tensor contractions emulation handles well.

Accuracy is the deeper concern. Scientific problems often hinge on nearly canceling terms, emulated methods carry finite error relative to IEEE standards, and reproducing IEEE safeguards would take considerable work. The conversation also touches on growing hardware heterogeneity, the value of high-precision simulation data for training AI models, and the HPC community’s habit of defaulting to double precision everywhere. Both guests close by calling on the community to determine rigorously where high precision is truly needed, framing it as a grand challenge for both HPC and AI.

![Nick Malaya AMD Headshot (1)-modified](https://webinars.techstronglearning.com/hs-fs/hubfs/Nick%20Malaya%20AMD%20Headshot%20(1)-modified.png?width=905&height=905&name=Nick%20Malaya%20AMD%20Headshot%20(1)-modified.png "Nick Malaya AMD Headshot (1)-modified")

### Nicholas Malaya

###### Senior Fellow in High Performance Computing and Sovereign AI - AMD

Nicholas Malaya is an AMD Senior Fellow in High Performance Computing and Sovereign AI. He focuses on enabling scientific and AI workloads to run efficiently on the world’s largest supercomputers from Day-0 of system deployment. He served as AMD’s application performance technical lead for Frontier and El Capitan, the first two exascale supercomputers in the United States.

His research interests include high-performance computing, computational fluid dynamics, Bayesian inference, and AI for Science, particularly agentic performance tuning, surrogate models, and automated laboratories. He earned his PhD in computational science from the University of Texas at Austin and previously double-majored in physics and mathematics at Georgetown University, where he received the Treado Medal.

![Bronson Messer-modified](https://webinars.techstronglearning.com/hs-fs/hubfs/Bronson%20Messer-modified.png?width=399&height=399&name=Bronson%20Messer-modified.png "Bronson Messer-modified")

### Bronson Messer

###### Director of Science - Oak Ridge Leadership Computing Facility (OLCF)

As Director of Science, his primary job is to ensure that the resources provided through the OLCF’s allocation programs are effectively used to produce high-impact science and to communicate those impacts to all stakeholders, including the public. He is also Joint Faculty Professor in the Department of Physics & Astronomy at the University of Tennessee. His primary research interests are related to the explosion mechanisms and phenomenology of supernovae (both thermonuclear and core-collapse), especially neutrino transport and signatures, dense matter physics, and the details of turbulent nuclear combustion.

In addition, he has worked on the application of machine learning algorithms to the analysis of galaxy merger simulations and on performance modeling and prediction for high-performance computing architectures.

![Guy Currier-modified-2](https://webinars.techstronglearning.com/hs-fs/hubfs/Guy%20Currier-modified-2.png?width=447&height=447&name=Guy%20Currier-modified-2.png "Guy Currier-modified-2")

### Guy Currier

###### Research Director - The Futurum Group

Guy is responsible for positioning, go-to-market, and sales guidance across technologies and markets. He has decades of field experience describing technologies, their business and community value, and how they are evaluated and acquired. Guy’s specialty areas include AI, DevOps/cloud-native, enterprise applications, application integration, Big Data, governance-risk-compliance, containerization, virtualization, HPC, CPUs-GPUs-xPUs, and systems lifecycle management.

Guy started his technology career as a research director for technology media company Ziff Davis, with stints at PC Magazine, eWeek, and CIO Insight. Prior to joining Visible Impact, he worked at Dell, including postings in marketing, product, and technical marketing groups for a wide range of products, including engineered systems, cloud infrastructure, enterprise software, and mission-critical cloud services. He lives and works in Austin, TX.  

![Techstrong Group logo](https://webinars.techstronglearning.com/hubfs/Techstrong%20Group%20logo.png "Techstrong Group logo")

![TechstrongLearning-Colored-2](https://webinars.techstronglearning.com/hubfs/TechstrongLearning-Colored-2.png "TechstrongLearning-Colored-2")

![DO.com R+B RGB Logo-Feb-22-2024-07-06-28-4406-PM](https://webinars.techstronglearning.com/hs-fs/hubfs/DO.com%20R+B%20RGB%20Logo-Feb-22-2024-07-06-28-4406-PM.png?width=1363&height=267&name=DO.com%20R+B%20RGB%20Logo-Feb-22-2024-07-06-28-4406-PM.png "DO.com R+B RGB Logo-Feb-22-2024-07-06-28-4406-PM")

![Security boulevard logo blue](https://webinars.techstronglearning.com/hs-fs/hubfs/Security%20boulevard%20logo%20blue.png?width=2000&height=360&name=Security%20boulevard%20logo%20blue.png "Security boulevard logo blue")

![CNN Blue and White-4](https://webinars.techstronglearning.com/hs-fs/hubfs/CNN%20Blue%20and%20White-4.png?width=1400&height=400&name=CNN%20Blue%20and%20White-4.png "CNN Blue and White-4")

![Techstrong.ai blue logo](https://webinars.techstronglearning.com/hs-fs/hubfs/Techstrong.ai%20blue%20logo.png?width=2000&height=295&name=Techstrong.ai%20blue%20logo.png "Techstrong.ai blue logo")

© 2026 [Techstrong Inc.](https://techstronggroup.com/) All Rights Reserved. [Contact Support.](mailto:jared@techstronggroup.com)