PPT-Floating-Point Data Compression

at 75 Gbs on a GPU Molly A ONeil and Martin Burtscher Department of Computer Science Introduction Scientific simulations on HPC clusters Run on interconnected compute nodes P roduce and transfer lots of . - . Nalini. Kumar. . Gaurav. . Chitroda. . Komal. . Kasat. OUTLINE. Introduction. Prior Art. Hw/. Sw. Partitioning Of Floating Point Applications. Floating Point To Fixed Point Conversion. Fixed Point To Floating Point Conversion. HPEC Conference, Sept 12 2012. Michael Parker. Altera Corp. Dan Pritsker . Altera Corp. 2. 28-nm . DSP Architecture on Stratix V FPGAs. User-programmable variable-precision . signal processing. Optimized for single- and double-precision . Convert Unsigned Floating-Point Binary 10111101.101110112 to Decimal Solution: The integer portion is evaluated as an unsigned integer: 101111012 Original Integer decimal = 1*27 + 0*26 + 1*25 + in . Computer . Systems. Chapter 2. 2. Chapter 2 Objectives. Understand the fundamentals of numerical data representation and manipulation in digital computers.. Master the skill of converting between various radix systems.. 15-213: Introduction to Computer Systems. Recitation 3: Monday, Sept. . 9. th. , 2013. Ian Hartwig. Section E. Welcome to Recitation. Recitation is a place for interaction. If you have questions, please ask.. openacc. Ebad. . Salehi. , Ahmad . Lashgar. and . Amirali. . Baniasadi. Electrical and Computer Engineering Department. University of Victoria. This Work. Motivation: Effective memory bandwidth usage is needed to achieve high performance in GPUs.. Weirdness. . Overflow. Each data type has a limited range. Depends on platform/compiler. Going past boundary wraps around. Data Types. Integral Types. Name. Size. Range. short. 16 bits. –2. 15. (-32,768) to 2. Outline. Fixed-point Numbers. Floating Point Numbers. Superscalar Processors. Multithreading. Homogeneous Multiprocessing. Heterogeneous Multiprocessing. 1. 3.141592653589793238462643383…. Fixed-point Numbers. The data is usually described as a certain type. This type determines what you can do with the data and how the data is stored. In Python basic types are integers (. int. ), floating point (float), strings and Booleans. Rounding. Floating-Point Operations. Mathematical Properties. CS 105. “Tour of the Black Holes of Computing!”. Floating-Point Puzzles. For each of the following C expressions, either:. Argue that it is true for all argument values. Instructor:. . Mark Wyse. Teaching Assistants:. Kevin Bi Parker, . DeWilde. , Emily . Furst. ,. Sarah House, Waylon Huang, Vinny . Palaniappan. http://xkcd.com/571/. . Administrivia. Lab 1 due Friday (1/19). Rounding. Floating-Point Operations. Mathematical Properties. CS 105. “Tour of the Black Holes of Computing!”. Floating-Point Puzzles. For each of the following C expressions, either:. Argue that it is true for all argument values. IEEE 754 types of real number values. IEEE uses three different encodings to represent real values:. 1. . Denormalized values: . These are . extremely. small values, very close to 0, including 0. If such a value becomes too small, it can cause . Marks 32. Visit for more Learning Resources. Features of Pentium Processor. Separate instruction and Data caches.. Dual integer pipelines i.e. U-pipeline and V-Pipeline.. Branch prediction using the branch target buffer (BTB)..