[hpc-announce] [CFP] [Deadline Apr 16th] FlexScience at HPDC'26 (16th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures)

[Bogdan Nicolae]

======================================================
CALL FOR PAPERS
======================================================

16th Workshop on AI and Scientific Computing at Scale using Flexible
Computing Infrastructures
***  FlexScience 2026  ***
Held in conjunction with ACM HPDC 2026
July 16, 2026
Cleveland, OH, USA

======================================================
https://urldefense.us/v3/__https://sites.google.com/view/flexscience__;!!G_uCfscf7eWS!feIZ6o00LMfpKXxiLZJYn3ZQPQ6WP2Lzvi4KFlIpAzNSiZGkVlGH8vkJ9xTsRT2YK6yikr-AeYHkUIVKFbJhCaXj3KIc5w$ 
======================================================

The 16th workshop on AI and Scientific Computing at Scale using
Flexible Computing Infrastructures (FlexScience) will provide the
scientific community a dedicated forum for discussing new research,
development, and deployment efforts in running scientific computing
workloads that increasingly make use of agentic AI patterns in
flexible ecosystems that favor elasticity and dynamicity, which is not
sufficiently addressed by the current generation of supercomputers and
dedicated data centers. As the boundaries between traditional
simulation and modern artificial intelligence blur, FlexScience seeks
to explore the architectural and algorithmic innovations required to
support this convergence. We invite researchers and practitioners to
submit original work addressing the challenges of orchestrating
high-performance workloads across heterogeneous, distributed, and
often volatile environments.

FlexScience solicits high-quality 8-page full papers or 5-page short
papers describing original research in the following areas:

TOPICS OF INTEREST:
-------------------
Complex workflows at the intersection of HPC, Big Data and AI
Agentic AI workflows (shared agentic memory, tool invocation using
MCP, reasoning strategies, etc)
Experimental evaluations of porting HPC/AI applications to clouds and
reconfigurable data centers
Techniques to federate workflows and resources across distributed data
centers, HPC machines and the edge
Scalable, Secure and reliable federated learning in distributed
environments (HPC systems, edge devices)
Elastic infrastructures that combine HPC data centers and/or clouds
(bursting, data sharing)
Performance portability and related abstractions to hide the
heterogeneity of resources
Scalability and fine-tuning of high-performance AI and deep learning
frameworks for elastic use of resources (e.g., Tensorflow, PyTorch,
Horovod on a variable number of GPUs.)
Virtualization, containers, and dynamic provisioning
Elastic I/O, storage and data management services and architectures
(caching, prefetching, adaptations of building blocks such as NoSQL
databases and parallel file systems, etc.)
AI data pipelines, AI model repositories, AI storage techniques (async
checkpointing, incremental transfer learning with frozen weights,
provenance metadata, etc.)
Scalable Inferences and model serving (retrieval augmented generation,
batched inferences, KV caching, etc.)
Fault tolerance and reliability under dynamic provisioning of resources
Analysis of management complexity, cost, and variability of
heterogeneous resources

SCHEDULE AND SUBMISSION PROCEDURE:
----------------------------------
Paper submission deadline: April 17th, 2026
Author notifications: May 6, 2026
Camera ready papers: May 16, 2026
Workshop date: July 16, 2026

Authors are invited to submit papers describing unpublished, original
research. All submitted manuscripts should be formatted using the ACM
Master Template with sigconf  format (please be sure to use the
current version). All necessary documentation can be found at:
https://urldefense.us/v3/__https://www.acm.org/publications/proceedings-template__;!!G_uCfscf7eWS!feIZ6o00LMfpKXxiLZJYn3ZQPQ6WP2Lzvi4KFlIpAzNSiZGkVlGH8vkJ9xTsRT2YK6yikr-AeYHkUIVKFbJhCaXmhjZm1A$ . Workshop papers
can be either short (max 5 pages) or regular (max 8 pages). Both types
will receive equal consideration. We use single-blind reviewing
process, so please keep the authors names, publications, etc., in the
text.
Papers conforming to these guidelines should be submitted through
HotCRP: https://urldefense.us/v3/__https://flexscience2026.hotcrp.com__;!!G_uCfscf7eWS!feIZ6o00LMfpKXxiLZJYn3ZQPQ6WP2Lzvi4KFlIpAzNSiZGkVlGH8vkJ9xTsRT2YK6yikr-AeYHkUIVKFbJhCaXHjdlO1A$ 

WORKSHOP CHAIRS:
----------------
Alexandru Costan, IRISA / INSA Rennes, France (alexandru.costan at irisa.fr)
Bogdan Nicolae, Argonne National Laboratory, USA (bogdan.nicolae at acm.org)
Kento Sato, RIKEN Center, Japan (kento.sato at riken.jp)