Christian Bell

Learn More
This paper demonstrates the one-sided communication used in languages like UPC can provide a significant performance advantage for bandwidth-limited applications. This is shown through communication microbenchmarks and a case-study of UPC and MPI implementations of the NAS FT benchmark. Our optimizations rely on aggressively overlapping communication with(More)
This paper proposes a new memory registration strategy for supporting Remote DMA (RDMA) operations over pinning-based networks, as existing approaches are insufficient for efficiently implementing Global Address Space (GAS) languages. Although existing approaches often maximize bandwidth, they require levels of synchronization that discourage one-sided(More)
The Cray X1 was recently introduced as the first in a new line of parallel systems to combine high-bandwidth vector processing with an MPP system architecture. Alongside capabilities such as automatic fine-grained data parallelism through the use of vector instructions, the X1 offers hardware support for a transparent global-address space (GAS), which makes(More)
Process analytical technology (PAT), the regulatory initiative for incorporating quality in pharmaceutical manufacturing, is an area of intense research and interest. If PAT is effectively applied to bioprocesses, this can increase process understanding and control, and mitigate the risk from substandard drug products to both manufacturer and patient. To(More)
High-end supercomputers are increasingly built out of commodity components, and lack tight integration between the processor and network. This often results in inefficiencies in the communication subsystem, such as high software overheads and/or message latencies. In this paper we use a set of microbenchmarks to quantify the cost of this commoditization,(More)
One of the main trends in modern high performance computing is the use of clusters of SMPs instead of the large monolithic machines to solve scientific codes. Clusters have recently received a lot of traction in the community through the availability of low-cost high performance interconnects. These interconnects allow not only high performance, but good(More)
This paper proposes to improve a memory registration strategy for Remote DMA operations over pinningbased networks in the context of Clusters of Multiprocessors (CLUMPS). Although existing approaches focus primarily on bandwidth as a metric for evaluating the cost of DMA page registration (or pinning), there are many levels of host synchronization that hide(More)
Looking is one of the most basic and fundamental goaldirected behaviors. The neural circuitry that generates gaze shifts towards target objects is adaptive and compensates for changes in the sensorimotor plant. Here, we present a neural-dynamic architecture, which enables an embodied agent to direct its gaze towards salient objects in its environment. The(More)
The Berkeley UPC Compiler is an open source, high performance and portable implementation of Unified Parallel C (UPC), an SPMD global-address space language extension of ISO C. In previous work, we have experimented our compiler on a variety of high-performance networks and parallel architectures, including distributed memory machines and clusters of SMPs.(More)
Recent academic assessments of the potential for secure Internet-based voting have been extremely pessimistic. A panel assembled to review the Department of Defense’s SERVE (Secure Electronic Registration and Voting Experiment) project– a system designed to allow foreign-based members of the armed services to vote online in Federal elections, which was(More)
  • 1