MCP-S11-01 - LECTURE 1 - INTRODUCTION TO MULTICORE...

Info iconThis preview shows pages 1–8. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: LECTURE 1 - INTRODUCTION TO MULTICORE PROGRAMMING NYU Multicore Programming Class, Spring 2011 lerner@cs.nyu.edu 1 How Single-Chip Multiprocessors (SCM) came to be How (and when) they affect the way we write programs Dynamics of this class Demo and Lab 1 discussion In t r o C h a l le n g e s D Todays agenda Questions are welcome, anytime. D y n a m i c s D e m o / L a b NYU Multicore Programming, Spring 2011 2 Computation t 1 t 2 t N CPU In t r o Programs as sequence of instructions More speed or processing power are always welcome. How to obtain it? More instructions per cycle. Instructions (* diagrams from Blaise Barneys Introduction to Parallel Computing, an excellent read on the topic) NYU Multicore Programming, Spring 2011 3 Parallelizable Computation CPU In t r o Engaging more CPUs in a problems instructions Breaking the problem apart and generating more than one sequence of instructions. Or CPU NYU Multicore Programming, Spring 2011 4 Parallelizable Computation CPU In t r o We can also break data apart That way the same instructions can be run on each portion of the data. CPU CPU NYU Multicore Programming, Spring 2011 5 Parallelism is when several instructions are being executed at the same time Concurrency is when different instructions are interleaved Does parallel imply concurrent? Underlying architecture What is the relationship between instructions and data ow PUs are communicate In t r o A Bit of Terminology How CPUs are to communicate Programming models Hows a problem expressed in a parallelizable computation NYU Multicore Programming, Spring 2011 6 More on architecture and programming model on next lecture but , back to the point, its all about IPC Fetch Decode Operand Execute Result In t r o Instruction Execution If the CPU could execute faster, the computation would do so too. An idea: pipelining!...
View Full Document

This note was uploaded on 09/09/2011 for the course COMPUTER S MCP-S11 taught by Professor Robertgrimm during the Spring '11 term at NJIT.

Page1 / 22

MCP-S11-01 - LECTURE 1 - INTRODUCTION TO MULTICORE...

This preview shows document pages 1 - 8. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online