BCS MPI: N new approach i the system software design for large-scale parallel computers

Conference · Wed Jan 01 00:00:00 EST 2003

DOI:https://doi.org/10.1145/1048935.1050207· OSTI ID:976663

Fernández, J C ^[1]; Petrini, F ^[2]; Frachtenberg, E ^[3]

---

1. Juan C.
2. Fabrizio
3. Eitan

Buffered Co-Scheduled (BCS) MPI proposes a new approach to design the communication libraries for large-scale parallel machines. The emphasis of BCS MPI is on the global coordination of a large number of processes rather than in the traditional optimization of the local performance of a pair of communicating processes. BCS MPI delays the interprocessor communication in order to schedule globally the communication pattern and it is designed on top of a minimal set of collective communication primitives. In this paper we describe a prototype implementation of BCS MDI and its Communication protocols. The experimental results, executed on a set of scientific applications representative of the ASCI workload, show that BCS MPI is only marginally slower than the production-level MPI, but much simpler to implement, debug and analyze.

• Research Organization: Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
• Sponsoring Organization: USDOE
• OSTI ID: 976663
• Report Number(s): LA-UR-03-3207; TRN: US201017%%807
• Resource Relation: Conference: Submitted to: Supercomputing 2003, Phoenix, AZ, November 2003
• Country of Publication: United States
• Language: English

Related Subjects

97 MATHEMATICAL METHODS AND COMPUTING
COMMUNICATIONS
COMPUTERS
DESIGN
IMPLEMENTATION
OPTIMIZATION
PERFORMANCE
SCHEDULES
SUPERCOMPUTERS