Horst D. Simon

Lawrence Berkeley National Laboratory | LBL · Directorate

Neuroscience initiatives aim to develop new technologies and tools to measure and manipulate neuronal circuits. To deal with the massive amounts of data generated by these tools, the authors envision the co-location of open data repositories in standardized formats together with high-performance computing hardware utilizing open source optimized an...

Opportunities offered by new neuro-technologies are threatened by lack of coherent plans to analyze,manage, and understand the data. High-performance computing will allow exploratory analysis of massive datasets stored in standardized formats, hosted in open repositories, and integrated with simulations.

Abstract. In recent years the idea of emerging superintelligence has been discussed widely by popular media, and many experts voiced grave warnings about its possible consequences. This talk will use an analysis of progress in supercomputer performance to examine the gap between current technology and reaching the capabilities of the human brain. I...

Conventional architectures fail to performance-scale due to Moore and Dennard effects. This effort will apply machine learning techniques/ capability by augmenting platforms with neuromorphic accelerator functionality. The programming model will be Berkeley Caffe + specialized layers (e.g. Tea-CPE for TrueNorth) + accelerator API (IBM CAPI, Illinoi...

For more than two decades, the TOP500 list has enjoyed incredible success as a metric for supercomputing performance and as a source of data for identifying technological trends. The project's editors reflect on its usefulness and limitations for guiding large-scale scientific computing into the exascale era.

This poster outlines a proposed effort to incorporate hardware deep learning ASICs within a mainstream cluster environment, to develop and evaluate an appropriate programmer interface and API, and to port one or more HPC applications to accelerate adoption of this new computing paradigm.

High Frequency Trading is pervasive across all electronic financial markets. As algorithms replace an increasing number of tasks previously performed by humans, cascading effects similar to the Flash Crash of May 6th 2010 become more likely. In this study, we bring together a number of different data analysis tools to improve our understanding of n...

One of the fundamental shortcoming of the popular analysis tools for time series is that they require the data to be taken at uniform time intervals. However, the real-world time series, such as those from financial markets, are mostly from irregular time intervals. It is a common practice to resample the irregular time series into a regular one, b...

VPIN (Volume synchronized Probability of Informed trading) is a leading indicator of liquidity-induced volatility. It is best known for having produced a signal more than hours before the Flash Crash of 2010. On that day, the market saw the biggest one-day point decline in the Dow Jones Industrial Average, which culminated to the market value of $1...

The development of an exascale computing capability with machines capable of executing O(1018) operations per second by the end of the decade will be characterized by significant and dramatic changes in computing hardware architecture from current (2014) petascale high-performance computers. From the perspective of computational science, this will...

The development of an exascale computing capability with machines capable of executing O(1018) operations per second by the end of the decade will be characterized by significant and dramatic changes in computing hardware architecture from current (2014) petascale high-performance computers. From the perspective of computational science, this will...

VPIN (Volume synchronized Probability of Informed trading) is a leading indicator of liquidity-induced volatility. It is best known for having produced a signal more than hours before the Flash Crash of 2010. On that day, the market saw the biggest one-day point decline in the Dow Jones Industrial Average, which culminated to the market value of $1...

Electricity is a vital part of our daily life; therefore it is important to avoid irregularities such as the California Electricity Crisis of 2000 and 2001. In this work, we seek to predict anomalies using advanced machine learning algorithms, more specifically a Change Point Detection (CPD) algorithm on the electricity prices during the California...

The development of an exascale computing capability with machines capable of executing O(1018) operations per second by the end of the decade will be characterized by significant and dramatic changes in computing hardware architecture from current (2012) petascale high-performance computers. From the perspective of computational science, this will...

Inspired by the function, power, and volume of the organic brain, we are developing TrueNorth, a novel modular, non-von Neumann, ultra-low power, compact architecture. TrueNorth consists of a scalable network of neurosynaptic cores, with each core containing neurons, dendrites, synapses, and axons. To set sail for TrueNorth, we developed Compass, a...

Two-point Correlation Function (TPCF) is widely used in astronomy to characterize the distribution of matter/energy in the Universe, and help derive the physics that can trace back to the creation of the universe. However, it is prohibitively slow for current sized datasets, and would continue to be a critical bottleneck with the trend of increasin...

Among the algorithms that are likely to play a major role in future exascale computing, the fast multipole method (fmm) appears as a rising star. Our previous recent work showed scaling of an fmm on gpu clusters, with problem sizes of the order of billions ...

In January 2011, the International Center for Computational Science (ICCS) at LBNL and UC Berkeley organized a workshop in Berkeley, CA, USA on the role of emerging many-core and accelerator-based architectures in science and technology. The focus of the workshop was to usher experts and enthusiasts; from disparate backgrounds in academia and indus...

In the quest for cognitive computing, we have built a massively parallel cortical simulator, C2, that incorporates a number of innovations in computation, memory, and communication. Using C2 on LLNL's Dawn Blue Gene/P su-percomputer with 147, 456 CPUs and 144 TB of main memory , we report two cortical simulations – at unprecedented scale – that eff...

The analysis of Cosmic Microwave Background (CMB) observations is a long-standing computational challenge, driven by the exponential growth in the size of the data sets being gathered. Since this growth is projected to continue for at least the next decade, it will be critical to extend the analysis algorithms and their implementations to peta-scal...

The Thick-Restart Lanczos (TRLan) method is an effective method for solving large-scale Hermitian eigenvalue problems. However, its performance strongly depends on the dimension of the projection subspace. In this paper, we propose an objective function to quantify the effectiveness of a chosen subspace dimension, and then introduce an adaptive sch...

In the quest for cognitive computing, we have built a massively parallel cortical simulator, C2, that incorporates a number of innovations in computation, memory, and communication. Using C2 on LLNL’s Dawn Blue Gene/P supercomputer with 147, 456 CPUs and 144 TB of main memory, we report two cortical simulations – at unprecedented scale – that effec...

In this paper we analyze major recent trends and changes in High Performance Computing (HPC). The introduction of vector computers started the area of “Supercomputing”. The initial success of vector computers in the seventies was driven by raw performance. Massive Parallel Processors (MPPs) became successful in the early nineties due to their bette...

Designed for senior undergraduate and first-year graduate students, Grid Computing: Techniques and Applications shows professors how to teach this subject in a practical way. Extensively classroom-tested, it covers job submission and scheduling, Grid security, Grid computing services and software tools, graphical user interfaces, workflow editors,...

By almost any measure, the SciDAC community has come a long way since DOE launched the SciDAC program back in 2001. At the time, we were grappling with how to efficiently run applications on terascale systems (the November 2001 TOP500 list was led by DOE's ASCI White IBM system at Lawrence Livermore achieving 7.2 teraflop/s). And the results stemmi...

In the quest for cognitive computing, we have built a mas- sively parallel cortical simulator, C2, that incorporates a number of innovations in computation, memory, and com- munication. Using C2 on LLNL’s Dawn Blue Gene/P su- percomputer with 147, 456 CPUs and 144 TB of main mem- ory, we report two cortical simulations – at unprecedented scale – th...

In the quest for cognitive computing, we have built a massively parallel cortical simulator, C2, that incorporates a number of innovations in computation, memory, and communication. Using C2 on LLNL's Dawn Blue Gene/P su-percomputer with 147, 456 CPUs and 144 TB of main memory , we report two cortical simulations – at unprecedented scale – that eff...

Thanks to imperatives for limiting waste heat, maximizing performance, and controlling operating cost, energy efficiency has been a driving force in the evolution of supercomputers. The challenge going forward will be to extend these gains to offset the steeply rising demands for computing services and performance.

The original software package TRLan, [TRLan User Guide], page 24, implements the thick restart Lanczos method, [Wu and Simon 2001], page 24, for computing eigenvalues λ and their corresponding eigenvectors v of a symmetric matrix A: Av = λv. Its effectiveness in computing the exterior eigenvalues of a large matrix has been demonstrated, [LBNL-429...

Green's function for the Laplace operator represents the propagation of inuence of point sources and is the foun- dation for solving many physics problems. On a graph of pairwise similarities, the Green's function is the inverse of the combinatorial Laplacian; we resolve the zero-mode dif- cult y by showing its physical origin as the consequence of...

This paper describes a restarted Lanczos algorithm that particularly suitable for implementation on distributed machines. The only communication operation is requires outside of the matrix-vector multiplication is a global sum. For most large eigenvalue problems, the global sum operation takes a small fraction of the total execution time. The major...

Scientific computing has often been called the third approach to scientific discovery, emerging as a peer to experimentation and theory. Historically, the synergy between theory and experimentation has been well understood. Experiments give insight into possible theories, theories inspire experiments, experiments reinforce or invalidate theories, a...

25th Edition of TOP500 List of World's Fastest Supercomputers Released: DOE/L LNL BlueGene/L and IBM gain Top Positions MANNHEIM, Germany; KNOXVILLE, Tenn.; BERKELEY, Calif. In what has become a closely watched event in the world of high-performance computing, the 25th edition of the TOP500 list of the world's fastest supercomputers was released to...

NERSC has developed a five-year strategic plan focusing on three components: Science-Driven Systems, Science-Driven Services, and Science-Driven Analytics. (1) Science-Driven Systems: Balanced introduction of the best new technologies for complete computational systems--computing, storage, networking, visualization and analysis--coupled with the ac...

The last decade has witnessed a rapid proliferation of superscalar cache-based microprocessors to build high-end capability and capacity computers primarily because of their generality, scalability, and cost effectiveness. However, the constant degradation of superscalar sustained performance, has become a well-known problem in the scientific compu...

We provide a systematic analysis of nonnegative matrix factorization (NMF) relating to data clustering. We generalize the usual X = FG{sup T} decomposition to the symmetric W = HH{sup T} and W = HSH{sup T} decompositions. We show that (1) W = HH{sup T} is equivalent to Kernel K-means clustering and the Laplacian-based spectral clustering. (2) X = F...

Over the past several years, computational scientists have observed a frustrating trend of stagnating application performance despite dramatic increases in peak performance of high performance computers. In 2002, researchers at Lawrence Berkeley National Laboratory, Argonne National Laboratory, and IBM pro-posed a new process to reverse this situat...

In a recent paper, Gordon Bell and Jim Gray (2002) put forth a view of the past, present, and future of high-performance computing (HPC) that is both insightful and thought provoking. Identifying key trends with a grace and candor rarely encountered in a single work, the authors describe an evolutionary past drawn from their vast experience and pro...

In this paper we analyze major recent trends and changes in the High Performance Computing (HPC) market place. The introduction of vector computers started the area of ‘Supercomputing’. The initial success of vector computers in the seventies was driven by raw performance. Massive parallel systems (MPP) became successful in the early nineties due t...

evelopment of parallel vector systems offers the potential to reduce this gap for many computational science codes and deliver a substantial increase in computing capabilities. This project examines the performance of the cacheless vector Earth Simulator (ES) and compares it to superscalar cache-based IBM Power3 system. Results demonstrate that the...

23rd Edition of TOP500 List of World's Fastest Supercomputers Released: Japan's Earth Simulator Enters Third Year in Top Position MANNHEIM, Germany; KNOXVILLE, Tenn.;&BERKELEY, Calif. In what has become a closely watched event in the world of high-performance computing, the 23rd edition of the TOP500 list of the world's fastest supercomputers was r...

Latent Semantic Indexing (LSI) uses the singular value decomposition to reduce noisy dimensions and improve the performance of text retrieval systems. Preliminary results have shown modest improvements in retrieval accuracy and recall, but these have mainly explored small collections. In this paper we investigate text retrieval on a larger document...

Lawrence Berkeley National Laboratory (Berkeley Lab) proposes to create a National Facility for Advanced Computational Science (NFACS) and to establish a new partnership between the American computer industry and a national consortium of laboratories, universities, and computing facilities. NFACS will provide leadership-class scientific computing c...

In [SIAM J. Matrix Anal. Appl., 23 (2002), pp. 706--727], we developed numerical algorithms for computing sparse low-rank approximations of matrices, and we also provided a detailed error analysis of the proposed algorithms together with some numerical experiments. The low-rank approximations are constructed in a certain factored form with the degr...

The goal of data clustering can be formally stated as a min-max clustering principle: data points are grouped into clusters such that (a) between-cluster similarities are minimized while (b) within-cluster similar-ities are maximized. Existing methods typically focus on one of the above requirements. Here we propose a new method that emphasizes bot...

ProteinShop Gives Researchers a Hands-On Tool for Manipulating, Visualizing Protein Structures. The Human Genome Project and other biological research efforts are creating an avalanche of new data about the chemical makeup and genetic codes of living organisms. But in order to make sense of this raw data, researchers need software tools which let t...

Last year's paper by Bell and Gray [1] examined past trends in high performance computing and asserted likely future directions based on market forces. While many of the insights drawn from this perspective have merit and suggest elements governing likely future directions for HPC, there are a number of points put forth that we feel require further...

21st Edition of TOP500 List of World's Fastest Supercomputers Released MANNHEIM, Germany; KNOXVILLE, Tenn.;&BERKELEY, Calif. In what has become a much-anticipated event in the world of high-performance computing, the 21st edition of the TOP500 list of the world's fastest supercomputers was released today (June 23, 2003). The Earth Simulator superco...

this paper, we describe the basic features of TOP/DOMDEC and highlight their application to the parallel solution of computational fluid and solid mechanics problems

In this paper we present the findings and recommendations that emerged from a one-day workshop held at Lawrence Berkeley National Laboratory (LBNL) on June 5, 2002, in conjunction with the National Energy Research Scientific Computing (NERSC) User Group (NUG) Meeting. The motivation for this workshop was to solicit direct input from the application...

We believe that it is critical for the future of high end computing in the United States to bring into existence a new class of computational capability that is optimal for science. In recent years scientific computing has increasingly become dependent on hardware that is designed and optimized for commercial applications. Science in this country h...

22nd Edition of TOP500 List of World s Fastest Supercomputers Released MANNHEIM, Germany; KNOXVILLE, Tenn.; BERKELEY, Calif. In what has become a much-anticipated event in the world of high-performance computing, the 22nd edition of the TOP500 list of the worlds fastest supercomputers was released today (November 16, 2003). The Earth Simulator supe...

With the exponential growth of information on the World Wide Web, there is great demand for developing efficient methods for effectively organizing the large amount of retrieved information. Document clustering plays an important role in information retrieval and taxonomy management for the Web. In this paper we examine three clustering methods: K-...

Two popular webpage ranking algorithms are HITS and PageRank. HITS emphasizes mutual reinforcement between authority and hub webpages, while PageRank emphasizes hyperlink weight normalization and web surfing based on random walk models. We systematically generalize /combine these concepts into a unified framework. The ranking framework contains a l...

20th Edition of TOP500 List of World's Fastest Supercomputers Released MANNHEIM, Germany; KNOXVILLE, Tenn.;&BERKELEY, Calif. In what has become a much-anticipated event in the world of high-performance computing, the 20th edition of the TOP500 list of the world's fastest supercomputers was released today (November 15, 2002). The Earth Simulator sup...

It is well-known that for high dimensional data clustering, standard algorithms such as EM and the K-means are often trapped in local minimum. Many initialization methods were proposed to tackle this problem , but with only limited success. In this paper we propose a new approach to resolve this problem by repeated dimension reductions such that K-...

This document proposes a multi-site strategy for creating a new class of computing capability for the U.S. by undertaking the research and development necessary to build supercomputers optimized for science in partnership with the American computer industry.

Latent Semantic Indexing (LSI) uses the Singular Value Decomposition to reduce noisy dimensions and improve the performance of text retrieval systems. Preliminary results have shown modest improvements in retrieval accuracy and recall, but these have mainly explored small collections. In this paper we investigate text retrieval on a large document...

Abstract We demonstrate that data clustering amounts to a dynamic process of self-aggregation in which data objects move,towards each other to form clusters, revealing the inherent pattern of similarity. Self-aggregation is governed by connectivity and occurs in a space obtained by a nonlinear scaling of principal component,analysis (PCA). The meth...

NERSC, the National Energy Research Scientific Computing Center, is a leading scientific computing facility for unclassified research, and has had a significant impact on computational physics in the U.S. Here we will summarize the recent experience at NERSC, and present the four key elements of our strategic plan for the next five years. Significa...

19th Edition of TOP500 List of World's Fastest Supercomputers Released MANNHEIM, Germany; KNOXVILLE, Tenn.;&BERKELEY, Calif. In what has become a much-anticipated event in the world of high-performance computing, the 19th edition of the TOP500 list of the worlds fastest supercomputers was released today (June 20, 2002). The recently installed Earth...

Ranking the tens of thousands of retrieved webpages for a user query on a Web search engine such that the most informative webpages are on the top is a key information retrieval technology. A popular ranking algorithm is the HITS algorithm which explores the mutual reinforcement between authority and hub webpages based on hyperlink structure of the...

The popular K-means clustering partitions a data set by minimizing a sum-of-squares cost function. A coordinate descend method is then used to find local minima. In this paper we show that the minimization can be reformulated as a trace maximization problem associated with the Gram matrix of the data vectors. Furthermore, we show that a relaxed ver...

It is well-known that for high dimensional data clustering, standard algorithms such as EM and K-means are often trapped in a local minimum. Many initialization methods have been proposed to tackle this problem, with only limited success. In this paper we propose a new approach to resolve this problem by repeated dimension reductions such that K-me...

Two popular webpage ranking algorithms are HITS and PageRank. HITS emphasizes mutual reinforcement be- tween authority and hub webpages, while PageRank em- phasizes hyperlink weight normalization and web surf- ing based on random walk models. We systemati- cally generalize/combine these concepts into a unied framework. The ranking framework contain...

18th Edition of TOP500 List of World's Fastest Supercomputers Released MANNHEIM, GERMANY; KNOXVILLE, TENN.; BERKELEY, CALIF. In what has become a much-anticipated event in the world of high-performance computing, the 18th edition of the TOP500 list of the world's fastest supercomputers was released today (November 9, 2001). The latest edition of th...

The Singular Value Decomposition (SVD) has attracted much interest of late as a technique for improving the performance of text retrieval systems (also called Latent Semantic Indexing). Preliminary results have shown modest improvements in retrieval accuracy, but these have mainly explored small collections. In this paper we investigate using the L...

An important application of graph partitioning is data clustering using a graph model | the pairwise similarities between all data objects form a weighted graph adjacency matrix that contains all necessary information for clustering. Here we propose a new algorithm for graph partition with an objective function that follows the min-max clustering p...

Grouping webpages into distinct topics is one way to organize the large amount of retrieved information on the web. In this paper, we report that based on similarity metric which incorporates textual information, hyperlink structure and co-citation relations, an unsupervised clustering method can automatically and effectively identify relevant topi...

In this paper we consider k-way graph clustering and k-way bipartite graph clustering. Many data types arising from data mining applications can be modeled as graphs and bipartite graphs. Examples include webpage links on the world-wide web, terms and documents in a text corpus, customers and purchasing items in market basket analysis and reviewers...

Automatic grouping of voluminous data into meaningful structures is a challenging task frequently encountered in broad areas of science, engineering and information processing. These data clustering tasks are frequently performed in Euclidean space or a subspace chosen from principal component analysis (PCA). Here we describe a space obtained by a...

Many data types arising from data mining applications can be modeled as bipartite graphs, examples include terms and documents in a text corpus, customers and purchasing items in market basket analysis and reviewers and movies in a movie recommender system. In this paper, we propose a new data clustering method based on partitioning the underlying...

Many data types arising from data mining applications can be modeled as bipartite graphs, examples include terms and documents in a text corpus, customers and purchasing items in market basket analysis and reviewers and movies in a movie recommender system. In this paper, we propose a new data clustering method based on partitioning the underlying...

In 1993 for the first time a list of the top 500 supercomputer sites worldwide has been made available. The TOP500 list allows a much more detailed and well founded analysis of the state of high performance computing. Previously data such as the number and geographical distribution of supercomputer installations were difficult to obtain, and only a...