Future Generation Computer Systems 22 (2006) 1032–1039 www.elsevier.com/locate/fgcs Real-time multi-scale brain data acquisition, assembly, and analysis using an end-to-end OptIPuter Rajvikram Singh b,∗ , Nicholas Schwarz b , Nut Taesombut c , David Lee a , Byungil Jeong b , Luc Renambot b , Abel W. Lin a , Ruth West a , Hiromu Otsuka e , Sei Naito f , Steven T. Peltier a , Maryann E. Martone a , Kazunori Nozaki d , Jason Leigh b , Mark H. Ellisman a a National Center for Microscopy and Imaging Research, University of California, San Diego, United States b Electronic Visualization Laboratory, University of Illinois at Chicago, United States c Department of Computer Science and Engineering, University of California, San Diego, United States d Cybermedia Center, Osaka University, Japan e KDDI Corporation, Garden Air Tower, 10-10, Iidabashi 3-chome, Chiyoda-ku, Tokyo 102-8460, Japan f KDDI Labs, 2-1-15 Ohara, Fujimino, Saitama, 356-8502, Japan Available online 9 May 2006 Abstract At iGrid 2005 we demonstrated the transparent operation of a biology experiment on a test-bed of globally distributed visualization, storage, computational, and network resources. These resources were bundled into a unified platform by utilizing dynamic lambda allocation, high band- width protocols for optical networks, a Distributed Virtual Computer (DVC) [N. Taesombut, A. Chien, Distributed Virtual Computer (DVC): Simplifying the development of high performance grid applications, in: Proceedings of the Workshop on Grids and Advanced Networks, GAN 04, Chicago, IL, April 2004 (held in conjunction with the IEEE Cluster Computing and the Grid (CCGrid2004) Conference)], and applications running over the Scalable Adaptive Graphics Environment (SAGE) [L. Renambot, A. Rao, R. Singh, B. Jeong, N. Krishnaprasad, V. Vishwanath, V. Chandrasekhar, N. Schwarz, A. Spale, C. Zhang, G. Goldman, J. Leigh, A. Johnson, SAGE: The Scalable Adaptive Graphics Environment, in: Proceedings of WACE 2004, 23–24 September 2004, Nice, France, 2004]. Using these layered technologies we ran a multi-scale correlated microscopy experiment [M.E. Maryann, T.J. Deerinck, N. Yamada, E. Bushong, H. Ellisman Mark, Correlated 3D light and electron microscopy: Use of high voltage electron microscopy and electron tomography for imaging large biological structures, Journal of Histotechnology 23 (3) (2000) 261–270], where biologists imaged samples with scales ranging from 20X to 5000X in progressively increasing magnification. This allows the sci- entists to zoom in from entire complex systems such as a rat cerebellum to individual spiny dendrites. The images used spanned multiple modalities of imaging and specimen preparation, thus providing context at every level and allowing the scientists to better understand the biological structures. This demonstration attempts to define an infrastructure based on OptIPuter components which would aid the development and design of collabo- rative scientific experiments, applications and test-beds and allow the biologists to effectively use the high resolution real estate of tiled displays. c  2006 Published by Elsevier B.V. Keywords: Graphics clusters; Multi-scale correlated microscopy; Montage images; HDTV streaming; Telescience; Tile displays; Tiled displays; Optical networks; Scientific visualization; Remote instrumentation 1. Introduction 1.1. The challenge Researchers interested in assessing brain tissue at multiple resolutions are faced with a well-known problem when ∗ Corresponding author. Tel.: +1 312 404 4058. E-mail address: rsingh@ncmir.ucsd.edu (R. Singh). traversing scales: as investigations increase in resolution they typically decrease in scope. This gap between dimensional scales makes it difficult to understand how higher order structures are constructed from finer building blocks. A particular challenge for the nervous system is the need to bridge the dimensional range of 100s of microns to nanometers. This range is called “mesoscale” and encompasses cellular networks, dendritic and axonal architectures, synaptic connectivity and macromolecular constituents. These structures represent the 0167-739X/$ - see front matter c  2006 Published by Elsevier B.V. doi:10.1016/j.future.2006.03.017