A Novel GPU Algorithm for Indexing Columnar Databases with Column Imprints

Total Page:16

File Type:pdf, Size:1020Kb

A Novel GPU Algorithm for Indexing Columnar Databases with Column Imprints A Novel GPU Algorithm for Indexing Columnar Databases with Column Imprints A THESIS SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL OF THE UNIVERSITY OF MINNESOTA BY Manaswi Mannem IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE Eleazar Leal July 2020 c Manaswi Mannem 2020 Acknowledgements I would like to thank my advisor, Dr. Eleazar Leal, who gave me the opportunity to work on this research. He has guided and supported me throughout my degree program, including this research. I have really learnt a lot from him. Additionally, I would like to thank Dr. Arshia Khan and Dr. Desineni Subbaram Naidu for being on my defense committee and for taking the time to go over this research work with me. Finally, many thanks to all the Masters of Computer Science students from the graduating classes of 2018, 2019 and 2020 for being an integral part of my growth here at UMD. i Dedication I dedicate this thesis to my parents, Syam Prasanna and Kamala Mannem, and my brother, Manasseh for being the most remarkable role models to me throughout my life. Their constant support and guidance has always pushed me to pursue my dreams and to excel at them. I would also like to dedicate this thesis to the incredible faculty of the Department of Computer Science at UMD. Working and interacting with each one of them as a student, a researcher and a teaching assistant has been the most rewarding learning experience of my graduate school tenure. ii Abstract Columnar database management systems (CDBMS) are specialized database sys- tems that store data in column-major order, i.e. where all the values of each attribute are stored consecutively in storage, as opposed to row-major order, which stores all the values of each row consecutively and is the most commonly strategy used by relational database systems such as Oracle, SQLServer, PostgreSQL, etc. The column-major order approach makes CDBMS more appropriate for data warehouses because query workloads for the latter systems usually involve retrieving all the values of a small subset of attributes. Just like in relational database management systems, query response time per- formance in CDBMS can greatly benefit from the existence of specially designed data structures, called indexes, that can help avoid exhaustively searching the entire database. However, existing indexing techniques for CDBMS like bitmaps, zonemaps etc. have been shown to result in large storage overhead and memory traffic between storage and CPU. Column imprints are an indexing approach for CDBMS that deals with these two issues by compressing the data in storage and by storing data in such a way that it reduces expensive data transfers between Random Access Mem- ory (RAM) and Central Processing Unit (CPU) caches. However, data compression and decompression, which are necessary for query processing in CDBMS imposes a significant computational burden. To deal with this problem, parallel architectures, such as Graphical Processing Units (GPU) can be used. GPUs are co-processors that have been shown to outperform CPUs for highly parallel tasks. Besides this, GPUs are highly energy efficient, affordable, and available in many types of machines, from mobile devices to supercomputers. Despite their advantages, no work has focused on the use of GPUs for column imprints indexing. iii To address the gap mentioned before, this thesis introduces the first GPU algo- rithm, named GPUImprints, for indexing columnar databases with column imprints. We also performed the first experimental study, using one real-world dataset and one synthetic dataset, on the use of GPUs versus CPUs for column imprints in terms of their index creation time, query response times and index space requirements. Our experiments showed that GPUImprints can speedup the column imprints construc- tion time by a factor of 20X and can speedup the query processing times of range queries by a factor of 3.7X when compared to the CPU imprints algorithm. iv Contents 1 Introduction1 1.1 Motivation.................................3 1.2 Contributions...............................5 1.3 Outline...................................6 2 Background7 2.1 Columnar Databases...........................7 2.1.1 Overview.............................7 2.1.2 Indexing Techniques.......................9 2.1.3 Applications............................ 15 2.2 GPUs................................... 16 2.2.1 Overview............................. 16 2.2.2 CUDA............................... 17 2.2.3 Applications............................ 20 2.3 Why use GPUs for Columnar Databases?................ 20 3 Proposed Algorithm 22 3.1 Overview.................................. 22 3.2 Stage 1: Creation of Imprint Index................... 22 v 3.3 Stage 2: Imprints Compression..................... 27 3.4 Stage 3: Query Evaluation........................ 29 4 Experimental Results 32 4.1 Challenges of Designing GPU Algorithm for Column Imprints.... 32 4.2 Hardware................................. 33 4.3 Datasets and Query Workload...................... 33 4.4 Parameters................................ 34 4.5 Analysis of GPUImprint Index...................... 35 4.6 Performance Analysis of Range Queries................. 39 5 Conclusions and Future Work 43 5.1 Conclusions................................ 43 5.2 Future Work................................ 44 References 45 vi List of Tables 2.1 Example of bitmap indexing........................ 11 vii List of Figures 1.1 Examples of OLAP and OLTP......................1 1.2 Example of row and columnar databases................2 2.1 Example of database cracking......................9 2.2 Example of byte-aligned bitmap code (BBC) compression...... 12 2.3 Example of WAH compression..................... 13 2.4 Example of zonemaps.......................... 15 2.5 APIs of CUDA.............................. 18 2.6 CUDA stack................................ 19 3.1 Example of GPU column imprints index................ 29 4.1 Comparison between imprints creation time of CPU and GPU for 2- byte integer data type (TPC-H).................... 35 4.2 Comparison between imprints creation time of CPU and GPU for 4- byte data type (Skyserver)........................ 36 4.3 Comparison between imprints creation time of CPU and GPU for 8- byte data type (Skyserver)........................ 37 4.4 GPU imprints index creation time per 10,000 values for different value type with varying widths of GPU imprint vector............ 38 viii 4.5 GPU imprints index creation time per 1000 values for different value type with varying cache line size..................... 39 4.6 Comparison of imprints index size for different column sizes for CPU and GPU................................. 40 4.7 Comparison of query response times for CPU and GPU........ 41 4.8 Query response time for different width GPU imprint vectors with varying cache line size.......................... 42 ix 1 Introduction Database management systems are used to record daily activities. They store data in tables either in row-wise (row databases) or column-wise (columnar databases) fashion. Most of the systems like order entry, retail sales, and financial transaction systems use databases managed by Online Transaction Processing Systems (OLTP) to record their daily updates [33]. However, for decision support and strategic planning the organizations also need historical data which is aggregated over many years [31]. The systems that store historical data are referred to data warehouses, also called Online Analytical Processing Systems (OLAP). Some of the examples of OLAP and OLTP systems are shown in Figure 1.1. In a row-oriented database, the data is stored row by row, that is after storing the Figure 1.1: Examples of OLAP and OLTP 1 entire first row, the second row is stored. Similarly, in columnar databases, data is stored column-wise, that is all the data associated with column is stored adjacently. In Figure 1.2, for row-oriented database, data is organized by row, keeping all of the data associated with a row next to each other in memory. Similarly in column- oriented databases, data is organized by column, keeping all the data associated with a column next to each other in memory. Some examples of columnar databases are MonetDB [20], MariaDB [17], CrateDB [9], Apache Hbase [3] etc. MonetDB is a main-memory database management system that stores data in columns. The archi- tecture of MonetDB has three layers: the front-end layer(top), back-end layer(middle) and the database kernel(bottom) [20]. The two indexing mechanisms used in Mon- etDB are Database Cracking [14] and Column Imprints [27]. Figure 1.2: Example of row and columnar databases 2 1.1 Motivation Column-oriented databases can scan the entire database faster than row-oriented databases [1]. This is because in column-oriented databases, scan operation is per- formed by traversing an array with the use of loops. This traversing can be en- hanced by exploiting indexing properly. According to [1], column-oriented database indexes are faster compared to traditional row-oriented database indexes and used as a standard for storing and querying large data warehouses. The main feature for the columnar architecture is that compression [35] can be efficiently performed. As most of the values in the columns are repetitive and have the same data type, com- pression algorithm can achieve higher compression ratio as they compress similar and consecutive values into one. An index is a data structure which helps to speed up query search. One example of index is an online catalog
Recommended publications
  • IBM DB2 for I Indexing Methods and Strategies Learn How to Use DB2 Indexes to Boost Performance
    IBM DB2 for i indexing methods and strategies Learn how to use DB2 indexes to boost performance Michael Cain Kent Milligan DB2 for i Center of Excellence IBM Systems and Technology Group Lab Services and Training July 2011 © Copyright IBM Corporation, 2011 Table of contents Abstract........................................................................................................................................1 Introduction .................................................................................................................................1 The basics....................................................................................................................................3 Database index introduction .................................................................................................................... 3 DB2 for i indexing technology .................................................................................................................. 4 Radix indexes.................................................................................................................... 4 Encoded vector indexes .................................................................................................... 6 Bitmap indexes - the limitations ....................................................................................................... 6 EVI structure details......................................................................................................................... 7 EVI runtime usage...........................................................................................................................
    [Show full text]
  • Improved Query Performance with Variant Indexes
    Improved Query Performance with Variant Indexes Patrick O’Neil Dallan Quass Department of Mathematics and Computer Science Department of Computer Science University of Massachusetts at Boston Stanford University- Boston, MA 02125-3393 Stanford, CA 94305 [email protected] [email protected] Abstract: The read-mostly environment of data warehousing hand. The Sybase IQ product currently provides both variant in- makes it possible to use more complex indexes to speed up dex types [EDEL95, FREN95], and recommendsmultiple indexes queries than in situations where concurrent updates are present. per column in some cases. The current paper presents a short review of current indexing technology, including row-set representation by Bitmaps, and Late in the paper, we introduce a new indexing approach to sup- then introduces two approaches we call Bit-Sliced indexing and port OLAP-type queries, commonly used in Data Warehouses. Projection indexing. A Projection index materializes all values Such queries are called Datncube queries in [GBLP96]. OLAP of a column in RID order, and a Bit-Sliced index essentially takes query performance depends on creating a set of summary tables an orthogonal bit-by-bit view of the same data. While some of to efficiently evaluate an expected set of queries. The summary these concepts started with the MODEL 204 product, and both tables pre-materialize needed aggregates, an approach that is Bit-Sliced and Projection indexing are now fully realized in possible only when the expected set of queries is known in ad- Sybase IQ, this is the first rigorous examination of such index- vance. Specifically, the OLAP approach addressesqueries that ing capabilities in the literature.
    [Show full text]
  • Using Bitmap Index for Joint Queries on Structured and Text Data∗
    Using Bitmap Index for Joint Queries on Structured and Text Data∗ Kurt Stockinger‡ John Cieslewicz§ Kesheng Wu‡ Doron Rotem‡ Arie Shoshani‡ Abstract The database and the information retrieval communities have been working on separate sets of techniques for querying structured data and text data, but there is a growing need to handle these types of data together. In this paper, we present a strategy to efficiently answer joint queries on both types of data. By using an efficient compression algorithm, our compressed bitmap indexes, called FastBit, are compact even when they contain millions of bitmaps. Therefore FastBit can be applied effectively on hundreds of thousands of terms over millions of documents. Bitmap indexes are designed to take advantage of data that only grows but does not change over time (append-only data), and thus are just as effective with append-only text archives. In a performance comparison against a commonly used database system with a full-text index, MySQL, we demonstrate that our indexes answer queries 50 times faster on average. Furthermore, we demonstrate that integrating FastBit with a open source database system, called MonetDB, yields similar performance gains. Since the integrated MonetDB/FastBit system provides the full SQL functionality, the overhead of supporting SQL is not the main reason for the observed performance differences. Therefore, using FastBit in other database systems can offer similar performance advantages. 1 Introduction The records in data warehouses are usually extracted from other database systems and therefore contain only what is known as structured data [10, 8, 29]. In these cases, most vendors are reusing existing database techniques to perform analysis tasks.
    [Show full text]
  • (Spring 2017) :: OLAP Indexes
    15-721 DATABASE SYSTEMS Lecture #09 – OLAP Indexes Andy Pavlo // Carnegie Mellon University // Spring 2016 @Andy_Pavlo // Carnegie Mellon University // Spring 2017 2 TODAY’S AGENDA OLAP Schemas Projection/Columnar Indexes (MSSQL) Bitmap Indexes CMU 15-721 (Spring 2017) 3 BIFURCATED ENVIRONMENT Extract Transform Load OLTP Databases OLAP Database CMU 15-721 (Spring 2017) 4 DECISION SUPPORT SYSTEMS Applications that serve the management, operations, and planning levels of an organization to help people make decisions about future issues and problems by analyzing historical data. Star Schema vs. Snowflake Schema CMU 15-721 (Spring 2017) 5 STAR SCHEMA PRODUCT_DIM CUSTOMER_DIM CATEGORY_NAME ID CATEGORY_DESC FIRST_NAME PRODUCT_CODE SALES_FACT LAST_NAME PRODUCT_NAME EMAIL PRODUCT_DESC PRODUCT_FK ZIP_CODE TIME_FK LOCATION_FK CUSTOMER_FK LOCATION_DIM TIME_DIM COUNTRY PRICE YEAR STATE_CODE QUANTITY DAY_OF_YEAR STATE_NAME MONTH_NUM ZIP_CODE MONTH_NAME CITY DAY_OF_MONTH CMU 15-721 (Spring 2017) 5 STAR SCHEMA PRODUCT_DIM CUSTOMER_DIM CATEGORY_NAME ID CATEGORY_DESC FIRST_NAME PRODUCT_CODE SALES_FACT LAST_NAME PRODUCT_NAME EMAIL PRODUCT_DESC PRODUCT_FK ZIP_CODE TIME_FK LOCATION_FK CUSTOMER_FK LOCATION_DIM TIME_DIM COUNTRY PRICE YEAR STATE_CODE QUANTITY DAY_OF_YEAR STATE_NAME MONTH_NUM ZIP_CODE MONTH_NAME CITY DAY_OF_MONTH CMU 15-721 (Spring 2017) 6 CAT_LOOKUP SNOWFLAKE SCHEMA CATEGORY_ID CATEGORY_NAME CATEGORY_DESC CUSTOMER_DIM PRODUCT_DIM ID CATEGORY_FK SALES_FACT FIRST_NAME PRODUCT_CODE LAST_NAME PRODUCT_NAME PRODUCT_FK EMAIL PRODUCT_DESC TIME_FK ZIP_CODE LOCATION_FK LOCATION_DIM CUSTOMER_FK TIME_DIM COUNTRY YEAR STATE_FK DAY_OF_YEAR ZIP_CODE PRICE MONTH_FK CITY DAY_OF_MONTH QUANTITY STATE_LOOKUP MONTH_LOOKUP STATE_ID MONTH_NUM STATE_CODE MONTH_NAME STATE_NAME MONTH_SEASON CMU 15-721 (Spring 2017) 7 STAR VS. SNOWFLAKE SCHEMA Issue #1: Normalization → Snowflake schemas take up less storage space. → Denormalized data models may incur integrity and consistency violations. Issue #2: Query Complexity → Snowflake schemas require more joins to get the data needed for a query.
    [Show full text]
  • SBH: Super Byte-Aligned Hybrid Bitmap Compression
    Information Systems 62 (2016) 155–168 Contents lists available at ScienceDirect Information Systems journal homepage: www.elsevier.com/locate/infosys SBH: Super byte-aligned hybrid bitmap compression Sangchul Kim, Junhee Lee, Srinivasa Rao Satti n, Bongki Moon n Department of Computer Science and Engineering, Seoul National University, Seoul 08826, Republic of Korea article info abstract Article history: Bitmap indexes are commonly used in data warehousing applications such as on-line Received 9 February 2016 analytic processing (OLAP). Storing the bitmaps in compressed form has been shown to be Received in revised form effective not only for low cardinality attributes, as conventional wisdom would suggest, 24 June 2016 but also for high cardinality attributes. Compressed bitmap indexes, such as Byte-aligned Accepted 8 July 2016 Bitmap Compression (BBC), Word-Aligned Hybrid (WAH) and several of their variants have Recommended by: D. Shasha fi Available online 16 July 2016 been shown to be ef cient in terms of both time and space, compared to traditional database indexes. In this paper, we propose a new technique for compressed bitmap Keywords: indexing, called Super Byte-aligned Hybrid (SBH) bitmap compression, which improves Database Indexing upon the current state-of-the-art compression schemes. In our empirical evaluation, the Bitmap Index query processing time of SBH was about five times faster than that of WAH, while the size Bitmap compression Byte-based Bitmap Code of its compressed bitmap indexes was retained nearly close to that of BBC. Word-Aligned Hybrid & 2016 Elsevier Ltd. All rights reserved. 1. Introduction A bitmap index is a collection of bit vectors created for each distinct value in the indexed column.
    [Show full text]
  • Efficient Indexing of Hashtags Using Bitmap Indices
    Efficient Indexing of Hashtags using Bitmap Indices Lawan Thamsuhang Subba Christian Thomsen Torben Bach Pedersen Aalborg University Aalborg University Aalborg University [email protected] [email protected] [email protected] ABSTRACT workloads (OLAP). In order to read only relevant data, column- The enormous amounts of data being generated regularly means oriented storage formats like Orc and Parquet support predicate that rapidly accessing relevant data from data stores is just as pushdown, where the search predicates using =, <, >, <=, >= or != important as its storage. This study focuses on the use of a dis- are pushed down to the storage level and are evaluated against its tributed bitmap indexing framework to accelerate query execu- aggregate based indices holding the minimum and maximum val- tion times in distributed data warehouses. Previous solutions for ues for each column in each block. Such aggregate based indices bitmap indexing at a distributed scale are rigid in their implemen- work fine when the data is ordered, but when the data appears tation, use a single compression algorithm, and provide their own unordered or is skewed, they are prone to false positive results. mechanisms to store, distribute and retrieve the indices. Users To alleviate this problem [20] proposed columnar imprints which are locked to their implementations even when other alterna- scans the entire column to create bit vectors for every cache line tives for compression and index storage are available or desirable. of data. A bit is set within a bit vector if at least one value occurs We provide an open source, lightweight, and flexible distributed within the corresponding bin.
    [Show full text]
  • Bitmap Index Design Choices and Their Performance Implications
    LBNL-62756 Bitmap Index Design Choices and Their Performance Implications Elizabeth O’Neil and Patrick O’Neil Kesheng Wu University of Massachusetts at Boston Lawrence Berkeley National Laboratory {eoneil, poneil}@cs.umb.edu [email protected] ABSTRACT Historically, bitmap indexing has provided an important database capability to accelerate queries. However, only a few database systems have implemented these indexes because of the difficulties of modifying fundamental assumptions in the low-level design of a database system and in the expectations of customers, both of which have developed in an environment that does not support bitmap indexes. Another problem that arises, and one that may more easily be addressed by a research article, is that there is no definitive design for bitmap indexes; bitmap index designs in Oracle, Sybase IQ, Vertica and MODEL 204 are idiosyncratic, and some of them were designed for older machine architectures. To investigate an efficient design on modern processors, this paper provides details of the Set Query benchmark and a comparison of two research implementations of bitmap indexes. One, called RIDBit, uses the N-ary storage model to organize table rows, and implements a strategy that gracefully switches between the well-known B-tree RID-list structure and a bitmap structure. The other, called FastBit is based on vertical organization of the table data, where all columns are individually stored. It implements a compressed bitmap index, with a linear organization of the bitmaps to optimize disk accesses. Through this comparison, we evaluate the pros and cons of various design choices. Our analysis adds a number of subtleties to the conventional indexing wisdom commonly quoted in the database community.
    [Show full text]
  • Cognitive Computing Featuring the IBM Power System AC922
    Front cover Cognitive Computing Featuring the IBM Power System AC922 Ivaylo Bozhinov Boran Lee Gustavo Santos Redpaper IBM Redbooks Cognitive Computing Featuring the IBM Power System AC922 October 2019 REDP-5555-00 Note: Before using this information and the product it supports, read the information in “Notices” on page v. First Edition (October 2019) This edition applies to IBM Power System AC922 models GTH and GTX for Cognitive Solutions. © Copyright International Business Machines Corporation 2019. All rights reserved. Note to U.S. Government Users Restricted Rights -- Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp. Contents Notices . .v Trademarks . vi Preface . vii Authors. vii Now you can become a published author, too . viii Comments welcome. viii Stay connected to IBM Redbooks . ix Chapter 1. Introduction to cognitive computing . 1 1.1 Definition of cognitive computing . 2 1.2 What is IBM cognitive computing . 3 1.3 IBM cognitive solutions . 3 1.3.1 Watson Machine Learning . 4 1.3.2 IBM PowerAI Vision . 5 1.4 Third party cognitive solutions. 6 1.5 Power AC922 end-to-end . 6 Chapter 2. IBM Power System AC922 for cognitive computing . 9 2.1 Key hardware components . 10 2.2 Software supported on the Power AC922. 11 2.3 Outstanding features. 12 Chapter 3. Cognitive solutions . 15 3.1 Cognitive solutions from IBM . 16 3.1.1 IBM Watson Machine Learning Accelerator . 16 3.2 Third-party cognitive solutions . 30 3.2.1 H2O Driverless AI and Power AC922 . 30 3.2.2 SQream DB and Power AC922. 31 3.2.3 Kinetica and Power AC922 .
    [Show full text]
  • Modern B-Tree Techniques Contents
    Foundations and TrendsR in Databases Vol. 3, No. 4 (2010) 203–402 c 2011 G. Graefe DOI: 10.1561/1900000028 Modern B-Tree Techniques By Goetz Graefe Contents 1 Introduction 204 1.1 Perspectives on B-trees 204 1.2 Purpose and Scope 206 1.3 New Hardware 207 1.4 Overview 208 2 Basic Techniques 210 2.1 Data Structures 213 2.2 Sizes, Tree Height, etc. 215 2.3 Algorithms 216 2.4 B-trees in Databases 221 2.5 B-trees Versus Hash Indexes 226 2.6 Summary 230 3 Data Structures and Algorithms 231 3.1 Node Size 232 3.2 Interpolation Search 233 3.3 Variable-length Records 235 3.4 Normalized Keys 237 3.5 Prefix B-trees 239 3.6 CPU Caches 244 3.7 Duplicate Key Values 246 3.8 Bitmap Indexes 249 3.9 Data Compression 253 3.10 Space Management 256 3.11 Splitting Nodes 258 3.12 Summary 259 4 Transactional Techniques 260 4.1 Latching and Locking 265 4.2 Ghost Records 268 4.3 Key Range Locking 273 4.4 Key Range Locking at Leaf Boundaries 280 4.5 Key Range Locking of Separator Keys 282 4.6 Blink-trees 283 4.7 Latches During Lock Acquisition 286 4.8 Latch Coupling 288 4.9 Physiological Logging 289 4.10 Non-logged Page Operations 293 4.11 Non-logged Index Creation 295 4.12 Online Index Operations 296 4.13 Transaction Isolation Levels 300 4.14 Summary 304 5 Query Processing 305 5.1 Disk-order Scans 309 5.2 Fetching Rows 312 5.3 Covering Indexes 313 5.4 Index-to-index Navigation 317 5.5 Exploiting Key Prefixes 324 5.6 Ordered Retrieval 327 5.7 Multiple Indexes for a Single Table 329 5.8 Multiple Tables in a Single Index 333 5.9 Nested Queries and Nested Iteration 334
    [Show full text]
  • The Dimension-Join: a New Index for Data Warehouses Pedro Bizarro and Henrique Madeira University of Coimbra, Portugal Dep
    The Dimension-Join: A New Index for Data Warehouses Pedro Bizarro and Henrique Madeira University of Coimbra, Portugal Dep. Engenharia Informática – CISUC 3030-397, Coimbra – Portugal [email protected], [email protected] Abstract There are several auxiliary pre-computed access structures that allow faster answers by reading less base data. Examples are materialized views, join indexes, B-tree and bitmap indexes. This paper proposes dimension-join, a new type of index especially suited for data warehouses. The dimension-join borrows ideas from several concepts. It is a bitmap index, it is a multi-table join and when being used one of the tables is not read to improve performance. It is a multi-table join because it holds information belonging to two tables, which is similar to the join index proposed by Valduriez. However, instead of being composed by the tables’ primary keys, the dimension-join index is a bitmap index over the fact table using values from a dimension column. The dimension-join index is very useful when selecting facts depending on dimension tables belonging to snowflakes. The dimension-join represents a direct connection between the fact table and a table in the snowflake that can avoid several joins and produce enormous performance improvements. This paper also evaluates experimentally the dimension-join indexes using the TPC-H benchmark and shows that this new index structure can dramatically improve the performance for some queries. 1. Introduction Data warehouses (DWs) often grow to sizes of gigabytes or terabytes of information, which makes the performance of the queries issued by decision support tools one of the most important issues in Data Warehousing.
    [Show full text]
  • Gpu-Accelerated Applications Gpu‑Accelerated Applications
    GPU-ACCELERATED APPLICATIONS GPU-ACCELERATED APPLICATIONS Accelerated computing has revolutionized a broad range of industries with over five hundred applications optimized for GPUs to help you accelerate your work. CONTENTS 1 Computational Finance 2 Climate, Weather and Ocean Modeling 2 Data Science & Analytics 4 Deep Learning and Machine Learning 8 Defense and Intelligence 9 Manufacturing/AEC: CAD and CAE COMPUTATIONAL FLUID DYNAMICS COMPUTATIONAL STRUCTURAL MECHANICS DESIGN AND VISUALIZATION ELECTRONIC DESIGN AUTOMATION 17 Media & Entertainment ANIMATION, MODELING AND RENDERING COLOR CORRECTION AND GRAIN MANAGEMENT COMPOSITING, FINISHING AND EFFECTS EDITING ENCODING AND DIGITAL DISTRIBUTION ON-AIR GRAPHICS ON-SET, REVIEW AND STEREO TOOLS WEATHER GRAPHICS 24 Medical Imaging 27 Oil and Gas 28 Research: Higher Education and Supercomputing COMPUTATIONAL CHEMISTRY AND BIOLOGY NUMERICAL ANALYTICS PHYSICS SCIENTIFIC VISUALIZATION 39 Safety and Security 42 Tools and Management Computational Finance APPLICATION NAME COMPANY/DEVELOPER PRODUCT DESCRIPTION SUPPORTED FEATURES GPU SCALING Accelerated Elsen Secure, accessible, and accelerated back- • Web-like API with Native bindings for Multi-GPU Computing Engine testing, scenario analysis, risk analytics Python, R, Scala, C Single Node and real-time trading designed for easy • Custom models and data streams are integration and rapid development. easy to add Adaptiv Analytics SunGard A flexible and extensible engine for fast • Existing models code in C# supported Multi-GPU calculations of a wide variety of pricing transparently, with minimal code Single Node and risk measures on a broad range of changes asset classes and derivatives. • Supports multiple backends including CUDA and OpenCL • Switches transparently between multiple GPUs and CPUS depending on the deal support and load factors.
    [Show full text]
  • Gpu-Accelerated Applications Gpu‑Accelerated Applications
    GPU-ACCELERATED APPLICATIONS GPU-ACCELERATED APPLICATIONS Accelerated computing has revolutionized a broad range of industries with over five hundred applications optimized for GPUs to help you accelerate your work. CONTENTS 1 Computational Finance 2 Climate, Weather and Ocean Modeling 2 Data Science and Analytics 4 Artificial Intelligence DEEP LEARNING AND MACHINE LEARNING 8 Federal, Defense and Intelligence 10 Design for Manufacturing/Construction: CAD/CAE/CAM COMPUTATIONAL FLUID DYNAMICS COMPUTATIONAL STRUCTURAL MECHANICS DESIGN AND VISUALIZATION ELECTRONIC DESIGN AUTOMATION INDUSTRIAL INSPECTION 19 Media & Entertainment ANIMATION, MODELING AND RENDERING COLOR CORRECTION AND GRAIN MANAGEMENT Test Drive the COMPOSITING, FINISHING AND EFFECTS World’s Fastest EDITING ENCODING AND DIGITAL DISTRIBUTION Accelerator – Free! ON-AIR GRAPHICS ON-SET, REVIEW AND STEREO TOOLS Take the GPU Test Drive, a free and WEATHER GRAPHICS easy way to experience accelerated computing on GPUs. You can run 29 Medical Imaging your own application or try one of the preloaded ones, all running on a 30 Oil and Gas remote cluster. Try it today. 31 Research: Higher Education and Supercomputing www.nvidia.com/gputestdrive COMPUTATIONAL CHEMISTRY AND BIOLOGY NUMERICAL ANALYTICS PHYSICS SCIENTIFIC VISUALIZATION 47 Safety and Security 49 Tools and Management Computational Finance APPLICATION NAME COMPANY/DEVELOPER PRODUCT DESCRIPTION SUPPORTED FEATURES GPU SCALING Accelerated Elsen Secure, accessible, and accelerated back- • Web-like API with Native bindings for Multi-GPU Computing Engine testing, scenario analysis, risk analytics Python, R, Scala, C Single Node and real-time trading designed for easy • Custom models and data streams are integration and rapid development. easy to add Adaptiv Analytics SunGard A flexible and extensible engine for fast • Existing models code in C# supported Multi-GPU calculations of a wide variety of pricing transparently, with minimal code Single Node and risk measures on a broad range of changes asset classes and derivatives.
    [Show full text]