DOE/LANL Jurisdiction Fire Danger Rating:
Hrc Banner
  1. LANL Home
  2. Engage with LANL
  3. Lab Organizations
  4. ALDSCT
  5. high-performance-computing
  6. Ultrascale Systems Research Center

EMC3

Efficient Mission Centric Computing Consortium

EMC3 employs various tools and resources to investigate ultra-scale computing architectures, systems, and environments that can achieve higher efficiencies in extreme scale mission-centric computing.

The primary focus is on the most demanding multi-physics applications involving largely unstructured/sparse problems that require a balance of compute, memory size, memory bandwidth, memory latency, network, and I/O. Working together, consortium members from the US Industrial Base “Users” to HPC component and system developers, will collaborate together to attack this challenging problem space. Achieving higher efficiency in ultra-scale computation supports National Security interests and the US Industrial Base needs

Get involved

There are a few ways you can get involved:

  • Co-Fund research and development which can include funding/co-funding and mentorship to computer science and mathematician graduate students on applicable research topics of interest. May also include lab test and evaluation elements (LANSCE).
  • Co-Locate Staff at Ultrascale Systems Research Center (USRC) or co-fund USRC staff to conduct research on topics of interest.
  • Sabbatical location for staff to work on desired projects.
  • Technical Interchange meetings (TIMS) or Focus area discussions. These can range from quarterly technical meetings, to casual technical discussions or special topic days.
  • Participation in forthcoming “Efficient Mission Computing” workshops and conferences.

Projects

Publications

  • Accelerating Scientific Data Analytics with SK hynix KV-CSD
  • API Requirement for LANL's Next-Gen KV-Based Storage

Collaboration Areas

Areas include, but are not limited to:

  • Storage and Network: data management, data movement, data manipulation and library middleware packages to enable all of these.
  • Compute: processor/memory complex for focused balanced application performance, enhancements to communication libraries. (MPI, OpenSHMEM, OpenUCX)
  • Simulation, Optimization and Benchmarking: scaling efficiencies.
  • Efficient System Management: next generation data driven management/job execution, anomaly analysis, system and environmentally aware scheduling, system resiliency improvements—all focused on creating more reliable and sustainable HPC systems.