Publications
Benchmarking data analysis and machine learning applications on the Intel KNL many-core processor
September 12, 2017
Conference Paper
Published in:
HPEC 2017: IEEE Conf. on High Performance Extreme Computing, 12-14 September 2017.
Topic:
R&D area:
R&D group:
Summary
Knights Landing (KNL) is the code name for the second-generation Intel Xeon Phi product family. KNL has generated significant interest in the data analysis and machine learning communities because its new many-core architecture targets both of these workloads. The KNL many-core vector processor design enables it to exploit much higher levels of parallelism. At the Lincoln Laboratory Supercomputing Center (LLSC), the majority of users are running data analysis applications such as MATLAB and Octave. More recently, machine learning applications, such as the UC Berkeley Caffe deep learning framework, have become increasingly important to LLSC users. Thus, the performance of these applications on KNL systems is of high interest to LLSC users and the broader data analysis and machine learning communities. Our data analysis benchmarks of these application on the Intel KNL processor indicate that single-core double-precision generalized matrix multiply (DGEMM) performance on KNL systems has improved by ~3.5x compared to prior Intel Xeon technologies. Our data analysis applications also achieved ~60% of the theoretical peak performance. Also a performance comparison of a machine learning application, Caffe, between the two different Intel CPUs, Xeon E5 v3 and Xeon Phi 7210, demonstrated a 2.7x improvement on a KNL node.
Summary
Knights Landing (KNL) is the code name for the second-generation Intel Xeon Phi product family. KNL has generated significant interest in the data analysis and machine learning communities because its new many-core architecture targets both of these workloads. The KNL many-core vector processor design enables it to exploit much higher...
READ MORE
A cloud-based brain connectivity analysis tool
September 12, 2017
Conference Paper
Published in:
HPEC 2017: IEEE Conf. on High Performance Extreme Computing, 12-14 September 2017.
Summary
With advances in high throughput brain imaging at the cellular and sub-cellular level, there is growing demand for platforms that can support high performance, large-scale brain data processing and analysis. In this paper, we present a novel pipeline that combines Accumulo, D4M, geohashing, and parallel programming to manage large-scale neuron connectivity graphs in a cloud environment. Our brain connectivity graph is represented using vertices (fiber start/end nodes), edges (fiber tracks), and the 3D coordinates of the fiber tracks. For optimal performance, we take the hybrid approach of storing vertices and edges in Accumulo and saving the fiber track 3D coordinates in flat files. Accumulo database operations offer low latency on sparse queries while flat files offer high throughput for storing, querying, and analyzing bulk data. We evaluated our pipeline by using 250 gigabytes of mouse neuron connectivity data. Benchmarking experiments on retrieving vertices and edges from Accumulo demonstrate that we can achieve 1-2 orders of magnitude speedup in retrieval time when compared to the same operation from traditional flat files. The implementation of graph analytics such as Breadth First Search using Accumulo and D4M offers consistent good performance regardless of data size and density, thus is scalable to very large dataset. Indexing of neuron subvolumes is simple and logical with geohashing-based binary tree encoding. This hybrid data management backend is used to drive an interactive web-based 3D graphical user interface, where users can examine the 3D connectivity map in a Google Map-like viewer. Our pipeline is scalable and extensible to other data modalities.
Summary
With advances in high throughput brain imaging at the cellular and sub-cellular level, there is growing demand for platforms that can support high performance, large-scale brain data processing and analysis. In this paper, we present a novel pipeline that combines Accumulo, D4M, geohashing, and parallel programming to manage large-scale neuron...
READ MORE
Streaming graph challenge: stochastic block partition
September 12, 2017
Conference Paper
Published in:
HPEC 2017: IEEE Conf. on High Performance Extreme Computing, 12-14 September 2017.
Summary
An important objective for analyzing real-world graphs is to achieve scalable performance on large, streaming graphs. A challenging and relevant example is the graph partition problem. As a combinatorial problem, graph partition is NP-hard, but existing relaxation methods provide reasonable approximate solutions that can be scaled for large graphs. Competitive benchmarks and challenges have proven to be an effective means to advance state-of-the-art performance and foster community collaboration. This paper describes a graph partition challenge with a baseline partition algorithm of sub-quadratic complexity. The algorithm employs rigorous Bayesian inferential methods based on a statistical model that captures characteristics of the real-world graphs. This strong foundation enables the algorithm to address limitations of well-known graph partition approaches such as modularity maximization. This paper describes various aspects of the challenge including: (1) the data sets and streaming graph generator, (2) the baseline partition algorithm with pseudocode, (3) an argument for the correctness of parallelizing the Bayesian inference, (4) different parallel computation strategies such as node-based parallelism and matrix-based parallelism, (5) evaluation metrics for partition correctness and computational requirements, (6) preliminary timing of a Python-based demonstration code and the open source C++ code, and (7) considerations for partitioning the graph in streaming fashion. Data sets and source code for the algorithm as well as metrics, with detailed documentation are available at GraphChallenge.org.
Summary
An important objective for analyzing real-world graphs is to achieve scalable performance on large, streaming graphs. A challenging and relevant example is the graph partition problem. As a combinatorial problem, graph partition is NP-hard, but existing relaxation methods provide reasonable approximate solutions that can be scaled for large graphs. Competitive...
READ MORE
Static graph challenge: subgraph isomorphism
September 12, 2017
Conference Paper
Published in:
HPEC 2017: IEEE Conf. on High Performance Extreme Computing, 12-14 September 2017.
Summary
The rise of graph analytic systems has created a need for ways to measure and compare the capabilities of these systems. Graph analytics present unique scalability difficulties. The machine learning, high performance computing, and visual analytics communities have wrestled with these difficulties for decades and developed methodologies for creating challenges to move these communities forward. The proposed Subgraph Isomorphism Graph Challenge draws upon prior challenges from machine learning, high performance computing, and visual analytics to create a graph challenge that is reflective of many real-world graph analytics processing systems. The Subgraph Isomorphism Graph Challenge is a holistic specification with multiple integrated kernels that can be run together or independently. Each kernel is well defined mathematically and can be implemented in any programming environment. Subgraph isomorphism is amenable to both vertex-centric implementations and array-based implementations (e.g., using the Graph-BLAS.org standard). The computations are simple enough that performance predictions can be made based on simple computing hardware models. The surrounding kernels provide the context for each kernel that allows rigorous definition of both the input and the output for each kernel. Furthermore, since the proposed graph challenge is scalable in both problem size and hardware, it can be used to measure and quantitatively compare a wide range of present day and future systems. Serial implementations in C++, Python, Python with Pandas, Matlab, Octave, and Julia have been implemented and their single threaded performance have been measured. Specifications, data, and software are publicly available at GraphChallenge.org.
Summary
The rise of graph analytic systems has created a need for ways to measure and compare the capabilities of these systems. Graph analytics present unique scalability difficulties. The machine learning, high performance computing, and visual analytics communities have wrestled with these difficulties for decades and developed methodologies for creating challenges...
READ MORE
Performance measurements of supercomputing and cloud storage solutions
September 12, 2017
Conference Paper
Published in:
HPEC 2017: IEEE Conf. on High Performance Extreme Computing, 12-14 September 2017.
Summary
Increasing amounts of data from varied sources, particularly in the fields of machine learning and graph analytics, are causing storage requirements to grow rapidly. A variety of technologies exist for storing and sharing these data, ranging from parallel file systems used by supercomputers to distributed block storage systems found in clouds. Relatively few comparative measurements exist to inform decisions about which storage systems are best suited for particular tasks. This work provides these measurements for two of the most popular storage technologies: Lustre and Amazon S3. Lustre is an opensource, high performance, parallel file system used by many of the largest supercomputers in the world. Amazon's Simple Storage Service, or S3, is part of the Amazon Web Services offering, and offers a scalable, distributed option to store and retrieve data from anywhere on the Internet. Parallel processing is essential for achieving high performance on modern storage systems. The performance tests used span the gamut of parallel I/O scenarios, ranging from single-client, single-node Amazon S3 and Lustre performance to a large-scale, multi-client test designed to demonstrate the capabilities of a modern storage appliance under heavy load. These results show that, when parallel I/O is used correctly (i.e., many simultaneous read or write processes), full network bandwidth performance is achievable and ranged from 10 gigabits/s over a 10 GigE S3 connection to 0.35 terabits/s using Lustre on a 1200 port 10 GigE switch. These results demonstrate that S3 is well-suited to sharing vast quantities of data over the Internet, while Lustre is well-suited to processing large quantities of data locally.
Summary
Increasing amounts of data from varied sources, particularly in the fields of machine learning and graph analytics, are causing storage requirements to grow rapidly. A variety of technologies exist for storing and sharing these data, ranging from parallel file systems used by supercomputers to distributed block storage systems found in...
READ MORE
SoK: cryptographically protected database search
May 22, 2017
Conference Paper
Published in:
IEEE Symp. on Security and Privacy, 22-26 May 2017.
Summary
Protected database search systems cryptographically isolate the roles of reading from, writing to, and administering the database. This separation limits unnecessary administrator access and protects data in the case of system breaches. Since protected search was introduced in 2000, the area has grown rapidly, systems are offered by academia, start-ups, and established companies. However, there is no best protected search system or set of techniques. Design of such systems is a balancing act between security, functionality, performance, and usability. This challenge is made more difficult by ongoing database specialization, as some users will want the functionality of SQL, NoSQL, or NewSQL databases. This database evolution will continue, and the protected search community should be able to quickly provide functionality consistent with newly invented databases. At the same time, the community must accurately and clearly characterize the tradeoffs between different approaches. To address these challenges, we provide the following contributions:(1) An identification of the important primitive operations across database paradigms. We find there are a small number of base operations that can be used and combined to support a large number of database paradigms.(2) An evaluation of the current state of protected search systems in implementing these base operations. This evaluation describes the main approaches and tradeoffs for each base operation. Furthermore, it puts protected search in the context of unprotected search, identifying key gaps in functionality.(3) An analysis of attacks against protected search for different base queries.(4) A roadmap and tools for transforming a protected search system into a protected database, including an open-source performance evaluation platform and initial user opinions of protected search.
Summary
Protected database search systems cryptographically isolate the roles of reading from, writing to, and administering the database. This separation limits unnecessary administrator access and protects data in the case of system breaches. Since protected search was introduced in 2000, the area has grown rapidly, systems are offered by academia, start-ups...
READ MORE
SIAM data mining "brings it" to annual meeting
February 17, 2017
Project Report
Published in:
MIT Lincoln Laboratory Report 113081
Summary
The Data Mining Activity Group is one of SIAM's most vibrant and dynamic activity groups. To better share our enthusiasm for data mining with the broader SIAM community, our activity group organized six minisymposia at the 2016 Annual Meeting. These minisymposia included 48 talks organized by 11 SIAM members.
Summary
The Data Mining Activity Group is one of SIAM's most vibrant and dynamic activity groups. To better share our enthusiasm for data mining with the broader SIAM community, our activity group organized six minisymposia at the 2016 Annual Meeting. These minisymposia included 48 talks organized by 11 SIAM members.
READ MORE
Learning by doing, High Performance Computing education in the MOOC era
January 17, 2017
Journal Article
Published in:
J. Parallel Distrib. Comput., Vol. 105, July 2017, pp. 105-15.
Topic:
R&D area:
R&D group:
Summary
The High Performance Computing (HPC) community has spent decades developing tools that teach practitioners to harness the power of parallel and distributed computing. To create scalable and flexible educational experiences for practitioners in all phases of a career, we turn to Massively Open Online Courses (MOOCs). We detail the design of a unique self-paced online course that incorporates a focus on parallel solutions, personalization, and hands-on practice to familiarize student-users with their target system. Course material is presented through the lens of common HPC use cases and the strategies for parallelizing them. Using personalized paths, we teach researchers how to recognize the alignment between scientific applications and traditional HPC use cases, so they can focus on learning the parallelization strategies key to their workplace success. At the conclusion of their learning path, students should be capable of achieving performance gains on their HPC system.
Summary
The High Performance Computing (HPC) community has spent decades developing tools that teach practitioners to harness the power of parallel and distributed computing. To create scalable and flexible educational experiences for practitioners in all phases of a career, we turn to Massively Open Online Courses (MOOCs). We detail the design...
READ MORE
Enhancing HPC security with a user-based firewall
September 13, 2016
Conference Paper
Published in:
HPEC 2016: IEEE Conf. on High Performance Extreme Computing, 13-15 September 2016.
Topic:
R&D area:
R&D group:
Summary
High Performance Computing (HPC) systems traditionally allow their users unrestricted use of their internal network. While this network is normally controlled enough to guarantee privacy without the need for encryption, it does not provide a method to authenticate peer connections. Protocols built upon this internal network, such as those used in MPI, Lustre, Hadoop, or Accumulo, must provide their own authentication at the application layer. Many methods have been employed to perform this authentication, such as operating system privileged ports, Kerberos, munge, TLS, and PKI certificates. However, support for all of these methods requires the HPC application developer to include support and the user to configure and enable these services. The user-based firewall capability we have prototyped enables a set of rules governing connections across the HPC internal network to be put into place using Linux netfilter. By using an operating system-level capability, the system is not reliant on any developer or user actions to enable security. The rules we have chosen and implemented are crafted to not impact the vast majority of users and be completely invisible to them. Additionally, we have measured the performance impact of this system under various workloads.
Summary
High Performance Computing (HPC) systems traditionally allow their users unrestricted use of their internal network. While this network is normally controlled enough to guarantee privacy without the need for encryption, it does not provide a method to authenticate peer connections. Protocols built upon this internal network, such as those used...
READ MORE
Benchmarking the Graphulo processing framework
September 13, 2016
Conference Paper
Published in:
HPEC 2016: IEEE Conf. on High Performance Extreme Computing, 13-15 September 2016.
Topic:
R&D area:
R&D group:
Summary
Graph algorithms have wide applicability to a variety of domains and are often used on massive datasets. Recent standardization efforts such as the GraphBLAS are designed to specify a set of key computational kernels that hardware and software developers can adhere to. Graphulo is a processing framework that enables GraphBLAS kernels in the Apache Accumulo database. In our previous work, we have demonstrated a core Graphulo operation that performs large scale multiplication operations of database tables called TableMult. In this article, we present results of scaling the Graphulo engine to larger problems and scalablity when using greater number of resources. Specifically, we present the results of two experiments that demonstrate Graphulo scaling performance as linear with the number of available resources. The first experiment demonstrates cluster processing rates through Graphulo's TableMult operator on two large graphs, scaled between 2^17 and 2^19 vertices. The second experiment uses TableMult to extract a random set of rows from a large graph (2^19 nodes) to simulate a cued graph analytic. These benchmarking results are of relevance to Graphulo users who wish to apply Graphulo to their graph problems.
Summary
Graph algorithms have wide applicability to a variety of domains and are often used on massive datasets. Recent standardization efforts such as the GraphBLAS are designed to specify a set of key computational kernels that hardware and software developers can adhere to. Graphulo is a processing framework that enables GraphBLAS...
READ MORE