Project updates‎ > ‎

New Xstack Programming Challenges, Runtime Systems and Tools FOA Announced

posted Nov 22, 2011, 6:04 PM by Eric Van Hensbergen   [ updated Nov 22, 2011, 6:23 PM ]
"The Office of Advanced Scientific Computing Research (ASCR) of the Office of Science (SC), 
U.S. Department of Energy (DOE), hereby invites applications for basic research that represents 
significant advances in programming models, languages, compilers, runtime systems and tools 
that address fundamental challenges related to the system software stack for Exascale computing 
platforms (X-Stack).   

Programming models, languages, and related technologies that have sustained High Performance 
Computing (HPC) application software development for the past decade are inadequate for 
Exascale era computers. The significant increase in complexity of Exascale platforms due to 
energy-constrained, billion-way parallelism, with major changes to processor and memory 
architecture, requires new energy-efficient and resilient programming techniques that are 
portable across multiple future machine generations. We expect to make research investments 
that address fundamental Exascale challenges, while offering a transition path for existing 
scientific applications to fully explore the challenges and rewards of Exascale platforms. 
Exascale programming challenges and strategies were identified in the ASCR Exascale 
Programming Challenges Workshop and carefully captured in the workshop report
Challenges and strategies related to tools for Exascale platforms were identified in the ASCR 

Sought are complete solutions that will address multiple components of the system software 
stack and that will have the following characteristics: 
• Scalability: enable applications to strongly scale to Exascale levels of parallelism;  
• Programmability: clearly reduce the burden we are placing on high performance 
• Performance Portability:  eliminate or significantly minimize requirements for porting to 
future platforms; 
• Resilience: properly manage fault detection and recovery at all components of the 
software stack; and 
• Energy Efficiency:  maximally exploit dynamic energy saving opportunities, leveraging 
the tradeoffs between energy efficiency, resilience, and performance.  
We encourage solutions that involve radically new approaches to programming Exascale 
applications and algorithms. New approaches are required in order to address the complexities of 
Exascale systems.   It is important to demonstrate the viability of such solutions in a broad high 
performance programming context by showing how the proposed solution: 
• interoperates with existing programming environments based on the MPI+X model, so 
that a smooth migration path is enabled, and/or 
• enables the automatic transformation of applications (possibly with users in the loop) 
from the “old” programming environment to the “new” one, such that the transformations 
are semantics and performance preserving." 

Computer Science Notices: (Link to the ASCR CS web page)
 •  2012 X-Stack: Programming Challenges, Runtime Systems, and Tools
   ·Issued: 11/22/2011
   ·Letter of Intent (Strongly Encouraged) Due Date: December 19, 2011
   ·Application Due Date: February 6, 2012, 11:59pm Eastern Time
   ·Universities, Industry, etc. Announcement: DE-FOA-0000619.  
   ·Laboratories Announcement: LAB 12-619