Mark Hereld

Introduction to building projection-based tiled display systems

This tutorial introduces the concepts and technologies needed to build projector-based display systems. Tiled displays offer scalability, high resolution, and large formats for various applications. Tiled displays are an emerging technology for constructing semi-immersive visualization environments capable of presenting high-resolution images from scientific simulation. The largest impact may well arise from using large-format tiled displays as one of possibly multiple displays in building information or active spaces that surround the user with diverse ways of interacting with data and multimedia information flows. These environments may prove the ultimate successor to the desktop metaphor for information technology work. Several fundamental technological problems must be addressed to make tiled displays practical. These include: the choice of screen materials and support structures; choice of projectors, projector supports, and optional fine positioners; techniques for integrating image tiles into a seamless whole; interface devices for interaction with applications; display generators and interfaces; and the display software environment.

Toward simulation-time data analysis and I/O acceleration on leadership-class systems

The performance mismatch between computing and I/O components of current-generation HPC systems has made I/O the critical bottleneck for scientific applications. It is therefore critical to make data movement as efficient as possible, and, to facilitate simulation-time data analysis and visualization to reduce the data written to storage. These will be of paramount importance to enabling us to glean novel insights from simulations. We present our work in GLEAN, a flexible framework for data-analysis and I/O acceleration at extreme scale. GLEAN leverages the data semantics of applications, and fully exploits the diverse system topologies and characteristics. We discuss the performance of GLEAN for simulation-time analysis and I/O acceleration with simulations at scale on leadership class systems

ParalleX: A Study of A New Parallel Computation Model

This paper proposes the study of a new computation model that attempts to address the underlying sources of performance degradation (e.g. latency, overhead, and starvation) and the difficulties of programmer productivity (e.g. explicit locality management and scheduling, performance tuning, fragmented memory, and synchronous global barriers) to dramatically enhance the broad effectiveness of parallel processing for high end computing. In this paper, we present the progress of our research on a parallel programming and execution model - mainly, ParalleX. We describe the functional elements of ParalleX, one such model being explored as part of this project. We also report our progress on the development and study of a subset of ParalleX

Emergent epileptiform activity in neural networks with weak excitatory synapses

the LITL-X at University of Delaware. We then present a novel architecture model - Gilgamesh II - as a ParalleX processing architecture. A design point study of Gilgamesh II and the architecture concept strategy are presented.

Visualizing Large, Heterogeneous Data in Hybrid-Reality Environments

Brain electrical activity recorded during an epileptic seizure is frequently associated with rhythmic discharges in cortical networks. Current opinion in clinical neurophysiology is that strongly coupled networks and cellular bursting are prerequisites for the generation of epileptiform activity. Contrary to expectations, we found that weakly coupled cortical networks can create synchronized cellular activity and seizure-like bursting. Evaluation of a range of synaptic parameters in a detailed computational model revealed that seizure-like activity occurs when the excitatory synapses are weakened. Guided by this observation, we confirmed experimentally that, in mouse neocortical slices, a pharmacological reduction of excitatory synaptic transmission elicited sudden onset of repetitive network bursting. Our finding provides powerful evidence that onset of seizures can be associated with a reduction in synaptic transmission. These results open a new avenue to explore network synchrony and may ultimately lead to a rational approach to treatment of network pathology in epilepsy.

The star formation histories of low surface brightness galaxies

Constructing integrative visualizations that simultaneously cater to a variety of data types is challenging. Hybrid-reality environments blur the line between virtual environments and tiled display walls. They incorporate high-resolution, stereoscopic displays, which can be used to juxtapose large, heterogeneous datasets while providing a range of naturalistic interaction schemes. They thus empower designers to construct integrative visualizations that more effectively mash up 2D, 3D, temporal, and multivariate datasets.

THE SOUTH POLE NEAR INFRARED SKY BRIGHTNESS

We have performed deep imaging of a diverse sample of 26 low surface brightness galaxies (LSBGs) in the optical and the near-infrared. Using stellar population synthesis models, we find that it is possible to place constraints on the ratio of young to old stars (which we parametrize in terms of the average age of the galaxy), as well as the metallicity of the galaxy, using optical and near-infrared colours. LSBGs have a wide range of morphologies and stellar populations, ranging from older, high-metallicity earlier types to much younger and lower-metallicity late-type galaxies. Despite this wide range of star formation histories, we find that colour gradients are common in LSBGs. These are most naturally interpreted as gradients in mean stellar age, with the outer regions of LSBGs having lower ages than their inner regions. In an attempt to understand what drives the differences in LSBG stellar populations, we compare LSBG average ages and metallicities with their physical parameters. Strong correlations are seen between an LSBG’s star formation history and its K-band surface brightness, K-band absolute magnitude and gas fraction. These correlations are consistent with a scenario in which the star formation history of an LSBG primarily correlates with its surface density and its metallicity correlates with both its mass and its surface density.

Topology-aware data movement and staging for I/O acceleration on Blue Gene/P supercomputing systems

We report our finding that the South Pole is the darkest known Earth-based site for near infrared astronomical observations. For this reason it has great potentail for the most sensitive surveys of distant or faint objects. We find that the south polar sky background is substantially darker in the standard near infrared J, H, and K filters, and in an optimized KDARK filter centered at 2.36 microns. In particular, the KDARK background at the South Pole is only 162 ± 67 mu-Jy arcsec-2 at the zenith. This is consistent with the results described in an accompanying paper by Ashley et al. 1996, and is comparable to the sky brightness measured by high altitude balloon in the 2.4 micron (Matsumoto et al. 1994).

Examples of in transit visualization

There is growing concern that I/O systems will be hard pressed to satisfy the requirements of future leadership-class machines. Even current machines are found to be I/O bound for some applications. In this paper, we identify existing performance bottlenecks in data movement for I/O on the IBM Blue Gene/P (BG/P) supercomputer currently deployed at several leadership computing facilities. We improve the I/O performance by exploiting the network topology of BG/P for collective I/O, leveraging data semantics of applications and incorporating asynchronous data staging. We demonstrate the efficacy of our approaches for synthetic benchmark experiments and for application-level benchmarks at scale on leadership computing systems.

Accelerating I/O Forwarding in IBM Blue Gene/P Systems

One of the most pressing issues with petascale analysis is the transport of simulation results data to a meaningful analysis. Traditional workflow prescribes storing the simulation results to disk and later retrieving them for analysis and visualization. However, at petascale this storage of the full results is prohibitive. A solution to this problem is to run the analysis and visualization concurrently with the simulation and bypass the storage of the full results. One mechanism for doing so is in transit visualization in which analysis and visualization is run on I/O nodes that receive the full simulation results but write information from analysis or provide run-time visualization. This paper describes the work in progress for three in transit visualization solutions, each using a different transport mechanism.