IXPUG Workshop
"Application  Performance on Intel Xeon Phi –
Being Prepared for KNL and Beyond"

Location: Frankfurt am Main, Germany
Date: Thursday, June 23, 2016, 8:30am-6:00pm
Venue: Marriott Frankfurt Hotel

The workshop will bring together software developers and technology experts to share challenges, experiences and best-practice methods for the optimization of HPC workloads on the Intel Xeon Phi. The workshop will cover application performance and scalability challenges at all levels - from single processor, to moderately-scaled cluster, up to large HPC configurations with many Xeon Phi devices.

The keynote will present recent information about the KNL processor. The submitted talks cover optimization and scalability topics in real-world HPC applications, e.g. data layouts and code restructuring for efficient SIMD operation, work distribution and thread management. Aspects related to KNL features (e.g. high-bandwidth memory) are of particular interest. The usability of tools for development, debugging and performance analysis will be covered.

A planned panel session provides an opportunity to discuss optimization strategies for Phi and to provide feedback to the toolchain developers.

Important Note: The workshop is held in conjunction with the ISC'16 in Frankfurt (Main). For attending the workshop, you have to register for ISC Workshops (registration is open since March 1st). More information is on the ISC'16 Conference site. 

Program

Program Committee

Organizers and Contacts

Please contact one of the organizers if you have any questions.

• Richard A. Gerber
  National Energy Research Scientific Computing Center, Lawrence Berkeley National Lab. (NERSC)
  This email address is being protected from spambots. You need JavaScript enabled to view it.
• Kent Milfeld
  Texas Advanced Computing Center (TACC)
  This email address is being protected from spambots. You need JavaScript enabled to view it.
• Chris J. Newburn
  Intel Corp.
  This email address is being protected from spambots. You need JavaScript enabled to view it.
• Thomas Steinke
  Zuse Institute Berlin (ZIB)
  This email address is being protected from spambots. You need JavaScript enabled to view it.

Call for Papers

Many-core and multicore processor technologies have put more demand on memory and interconnects. These processor technologies also require a more detailed understanding of the interaction between the memory, interconnect and SIMD units for users to design optimal algorithms at scale on a node, as well as at scale on a system level.

The IXPUG workshop is about sharing ideas, implementations, and experiences that will help users take advantage of new technologies such as AVX512 operations, high-bandwidth memory (HBM) and OmniPath. These architectural advances in Vectorization, Memory, and Communications on the Intel Xeon Phi platform will help boost adoption of many-core architecture in HPC as well as other computational spaces.

In the workshop you will experience an open forum with fellow application programmers, Intel Phi architecture designers, and compiler and tool experts.

In addition to the technical paper presentations, the program will include a morning keynote on Intel microprocessors and an afternoon presentation on memory performance, and will conclude with a panel discussion.

IXPUG welcomes paper submissions on innovative work from KNC and KNL users in academia, industry and government labs, describing original discoveries and experiences that will promote and prescribe efficient use of many-core and multicore systems.

Topics of interest are (but not limited to):

• Vectorization: Data layout in cache for efficient SIMD operations, SIMD directives and operations, and 2-core tiling with 2D interconnected mesh
• Memory: Data layout in memory for efficient access (data preconditioning), access latency concerns (prefetch, streams, costs for HBM), partitioning of DDR and HBM for applications (memory policies)
• Communication, including early experiences with OmniPath
• Thread and Process Management: Process and thread affinity issues, SMT (simultaneous multi-threading, in core), balancing processes and threads
• Programming Modells: OpenMP 4.x, hStreams, using MPI 3 on Xeon Phi, hybrid programming (MPI/OpenMP, others)
• Algorithms and Methods: , including scalable and vectorizable algorithms
• Software Environments and Tools
• Benchmarking & Profiling Tools
• Visualization

Paper Submission

Paper Format: Papers will be published through the Springer Publishing, using their template (see Information for Authors below). It will also provide consistency for the reviewers in layout, page limit and font size.

• PDF files only from LaTeX or MS Word docs.
• Page limits: minimum 4, maximum 10 pages excluding reference list and acknowledgement.
• Submission URL: EasyChair IXPUG Workshop ISC16.  Please select workshop, under topics.

The authors of the best scored papers are invite by the Program Committee for publication in the ISC’16 Workshop volume published within the Springer LNCS series.

Information for Authors

General information for authors can be found on Springers LNCS “Information for Authors of Computer Science Publications”. It includes links to MS Word templates. For LaTeX, a simplified LNCS template and Quick Start can be found in the LNCS Repository on GitHub.

Presentation Submission

A limited number of presentations will be accepted in the area of late-breaking technology  and subjects that may contain less content than a paper (for a lightning talk).
Make sure you check the "presentation" box in the submission form. 

• Presentation Format: IXPUG PowerPoint template (with suggestions; specific format not required)
• Page limits: 10 pages for lightning talks; 20 pages for late-breaking talks
• Submission URL: EasyChair IXPUG Workshop ISC16

The authors of the best scored papers are invite by the Program Committee for presentation in the ISC’16 Workshop.

Important Due Dates

Reviewers are expected to make judgment on what was available at the time reviews were assigned (May 3).  Subsequent updates to content may or may not be considered by the program committee as part of the selection decision.  We encourage authors to exercise the freedom to use the time up until presentation and camera ready copy to provide the highest-quality product. 

Review Process

All submitted papers will be reviewed. We apply a standard single-blind review process, i.e., the authors will be known to reviewers. All submissions within the scope of the workshop will be peer-reviewed and will need to demonstrate quality of the results, originality and new insights, technical strength, and correctness. The submitted papers may not be published in or be in preparation for other conferences, workshops or journals.