TechTalks.Tv beta

Although Petaflop systems have only recently become a reality, scientists and governments are eagerly anticipating Exascale capabilities. There is a significant effort to achieve an Exaflop within the decade. However, unlike Petascale systems, Exascale systems will not be straightforward extrapolations from predecessors, and success is not a foregone conclusion. The chief combatant is power consumption: systems must become enormously more energy-efficient to make Exascale practical. General-purpose cores are too power-hungry, and thus more specialized forms of parallelism must be exploited. However, it is highly likely that this will require changes in applications and programming models, especially considering that these systems will require tens of millions of cores. Complicating matters further, error detection and resiliency are already major issues for Petascale systems, and improving fault tolerance typically requires more energy. We will discuss innovations in technology, architecture, and software tha

Opening Session Remarks

Note: This is the part-1 of the tutorial. Please see Mapreduce Part-2 for the rest of the presentation. Abstract: Apache Hadoop has become the platform of choice for developing large-scale data-intensive applications. In this tutorial, we will discuss design philosophy of Hadoop, describe how to design and develop Hadoop applications and higher-level application frameworks to crunch several terabytes of data, using anywhere from four to 4,000 computers. We will discuss solutions to common problems encountered in maximizing Hadoop application performance. We will also describe several frameworks and utilities developed using Hadoop that increase programmer-productivity and application-performance.

MapReduce Programming with Apache Hadoop

This presentation is recorded in the 2nd half of this video. Please excuse our delay, we will split and put this as a separate presentation soon.

Analyzing and Adjusting User Runtime Estimates to Improve Job Scheduling on the Blue Gene/P

This presentation is recorded in the 2nd half of this video. Please excuse our delay, we will split and put this as a separate presentation soon.

The new era in genomics: Opportunities and challenges for high performance computing

NVIDIA's CUDA architecture provides a powerful platform for writing highly parallel programs. By providing simple abstractions for hierarchical thread organization, memories, and synchronization, the CUDA programming model allows programmers to write scalable programs without the burden of learning a multitude of new programming constructs. The CUDA architecture can support many languages and programming environments, including C, Fortran, OpenCL, and DirectX Compute. In this tutorial, I will provide an overview of modern GPU processor design and its implications for successful parallel programming models. I will present the programming model defined by the CUDA architecture, and demonstrate how this is exposed in the C/C++ language. Finally, I will sketch some techniques for implementing common data-parallel algorithms in the CUDA model.

Where Is Your Dog's Belly Button? or IC-Scheduling Theory: A New Scheduling Paradigm for Task-Hungry Platforms