Douglas R. Miller

TL;DR: A multi-petascale Highly Efficient Parallel Supercomputer of 100 petaOPS-scale computing, at decreased cost, power and footprint, allows for a maximum packaging density of processing nodes from an interconnect point of view.

TL;DR: In this paper, the authors present a method and computer system for efficiently accessing resources in a multiprocessor computer system, where multiple resources are grouped into nodes and system resource descriptors are maintained as data structures.

TL;DR: This paper describes an extension to MPI that introduces communication endpoints as a means to relax the one-to-one relationship between processes and threads, and illustrates their potential for additional performance and computation management benefits through the decoupling of ranks from processes.

TL;DR: The MPI endpoints extension is described, which relaxes the longstanding one-to-one relationship between MPI processes and ranks and enables threads to be addressable in MPI operations, enhancing interoperability between MPi and other programming models.

TL;DR: In this paper, the authors propose a data communications model in a parallel computer that includes compute nodes that each include main memory and a messaging unit, the messaging unit including computer memory and coupling compute nodes for data communications, in which each compute node at compute node boot time allocates, in the message unit's computer memory, a predefined number of message buffers.