Pico-kernel Adaptive and Scalable Operating Systems Phase II (PICASO2)

• funded by: Securboration, Inc.
• funding level: $250,000
• duration: 06/01/2013 - 05/31/2015

The objective of this work is to continue on our path to design and evaluate novel system and program abstractions for combined performance and scalability paving the path into a future of operating system supporting a massive number of cores on a single chip. In Phase II, NCSU will collaborate with SecurBoration and will contribute effort to the following areas: (1) include the use of shared memory where appropriate, potentially within communication among related nodes; (2) improve power awareness, to include scheduling and communication considerations; (3) intelligently map tasks to cores; (4) optimize micro kernel interaction with other multicore elements, such as the micro controller; and (5) continue to ensure vendor neutrality. The results of the Phase II development will be a complete OS kernel design that is scalable and portable to multiple vendor many-core platforms.

Publications:

• Hybrid MPI/OpenMP Programming on the Tilera Manycore Architecture by Vishwanathan Chandu, Frank Mueller in International Conference on High Performance Computing & Simulation (HPCS), Jul 2016.
• Efficient and Predictable Group Communication for Manycore NoCs by Karthik Yagna, Onkar Patil, Frank Mueller in International Supercomputing Conference (ISC), Jun 2016.
• Distributed Job Allocation for Large-Scale Manycores by Subramanian Ramachandran, Frank Mueller in International Supercomputing Conference (ISC), Jun 2016.
• Reducing NoC and Memory Contention for Manycores by V. Chandru, F. Mueller in Architecture of Computing Systems (ARCS), Apr 2016.
• NoCMsg: A Scalable Message Passing Abstraction for Network-on-Chips by Christopher Zimmer, Frank Mueller in ACM Transactions on Architecture and Code Optimization (TACO), Vol. 12, No. 1, Apr 2015.
• NoCMsg: Scalable NoC-Based Message Passing by Christopher Zimmer, Frank Mueller in International Symposium on Cluster, Cloud and Grid Computing, May 2014.

• "Collective Communication for Multi-core NOC Interconnects" by Karthik Yagna, M.S. Thesis, North Carolina State University, May 2013 (last known position: Riverbed technologies, CA)
• "Distributed Job Allocation for Large-Scale Many-cores" by Subramanian Ramachandran, M.S. Thesis, North Carolina State University, May 2014 (last known position: Riverbed, CA)
• "Effcient and Lightweigth Inter-process Collective Operations for Massive Multi-core Architectures" by Onkar Patil, M.S. Thesis, North Carolina State University, Jun 2014 (last known position: NetApp, NC)
• "Analysis of Memory Performance and Execution Models for Large-Scale Manycores" by Vishwanathan Chandru, M.S. Thesis, North Carolina State University, Aug 2015 (last known position: Intel, IL)