DOC PREVIEW
Berkeley COMPSCI 61C - Lecture 42 – Parallel Computing

This preview shows page 1-2-24-25 out of 25 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 25 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 25 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 25 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 25 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 25 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

CS61C L42 Parallel Computing (1)Carle, Spring 2005 © UCBAndy [email protected]/~cs61cCS61C : Machine StructuresLecture #42 – Parallel Computing2005-05-09The California legislature is currently working on a bill to ban “remote hunting via the internet” after the incorporation of a Texas company specializing in a unique combination of robotics, web cameras, and weapons. Years of Counter Strike practice and I can’t even get a meal out of it…CS61C L42 Parallel Computing (2)Carle, Spring 2005 © UCBScientific Computing° Traditional Science1) Produce theories and designs on “paper”2) Perform experiments or build systems• Has become difficult, expensive, slow, and dangerous for fields on the leading edge° Computational Science• Use ultra-high performance computers to simulate the system we’re interested in° Acknowledgement• Many of the concepts and some of the content of this lecture were drawn from Prof. Jim Demmel’s CS 267 lecture slides which can be found at http://www.cs.berkeley.edu/~demmel/cs267_Spr05/CS61C L42 Parallel Computing (3)Carle, Spring 2005 © UCBExample Applications° Science• Global climate modeling• Biology: genomics; protein folding; drug design• Astrophysical modeling• Computational Chemistry• Computational Material Sciences and Nanosciences° Engineering• Semiconductor design• Earthquake and structural modeling• Computation fluid dynamics (airplane design)• Combustion (engine design)• Crash simulation° Business• Financial and economic modeling• Transaction processing, web services and search engines° Defense• Nuclear weapons -- test by simulations• CryptographyCS61C L42 Parallel Computing (4)Carle, Spring 2005 © UCBPerformance Requirements° Terminology• Flop – Floating point operation• Flops/second – standard metric for expressing the computing power of a system° Global Climate Modeling• Divide the world into a grid (e.g. 10 km spacing)• Solve fluid dynamics equations to determine what the air has done at that point every minute- Requires about 100 Flops per grid point per minute• This is an extremely simplified view of how the atmosphere works, to be maximally effective you need to simulate many additional systems on a much finer gridCS61C L42 Parallel Computing (5)Carle, Spring 2005 © UCBPerformance Requirements (2)° Computational Requirements• To keep up with real time (i.e. simulate one minute per wall clock minute): 8 Gflops/sec• Weather Prediction (7 days in 24 hours): 56 Gflops/sec• Climate Prediction (50 years in 30 days): 4.8 Tflops/sec• Climate Prediction Experimentation (50 years in 12 hours): 288 Tflops/sec° Perspective• Pentium 4 1.4GHz, 1GB RAM, 4x100MHz FSB- ~320 Mflops/sec, effective- Climate Prediction would take ~1233 yearsReference:http://www.tc.cornell.edu/~lifka/Papers/SC2001.pdfCS61C L42 Parallel Computing (6)Carle, Spring 2005 © UCBWhat Can We Do?°Wait• Moore’s law tells us things are getting better; why not stall for the moment?°Parallel Computing!CS61C L42 Parallel Computing (7)Carle, Spring 2005 © UCBProhibitive Costs° Rock’s Law• The cost of building a semiconductor chip fabrication plant that is capable of producing chips in line with Moore’s law doubles every four yearsCS61C L42 Parallel Computing (8)Carle, Spring 2005 © UCBHow fast can a serial computer be?° Consider a 1 Tflop/sec sequential machine:• Data must travel some distance, r, to get from memory to CPU• To get 1 data element per cycle, this means 1012times per second at the speed of light, c = 3x108m/s. Thus r < c/1012 = 0.3 mm- So all of the data we want to process must be stored within 0.3 mm of the CPU° Now put 1 Tbyte of storage in a 0.3 mm x 0.3 mm area:• Each word occupies about 3 square Angstroms, the size of a very small atom• Maybe someday, but it most certainly isn’t going to involve transistors as we know themCS61C L42 Parallel Computing (9)Carle, Spring 2005 © UCBWhat is Parallel Computing?°Dividing a task among multiple processors to arrive at a unified (meaningful) solution• For today, we will focus on systems with many processors executing identical code°How is this different from Multiprogramming (which we’ve touched on some in this course)?°How is this different from Distributed Computing?CS61C L42 Parallel Computing (10)Carle, Spring 2005 © UCBRecent History° Parallel Computing as a field exploded in popularity in the mid-1990s° This resulted in an “arms race” between universities, research labs, and governments to have the fastest supercomputer in the worldSource: top500.orgCS61C L42 Parallel Computing (11)Carle, Spring 2005 © UCBCurrent ChampionsBlueGene/L – IBM/DOERochester, United States32768 Processors, 70.72 Tflops/sec0.7 GHz PowerPC 440Columbia – NASA/AmesMountain View, United States10160 Processors, 51.87 Tflops/sec1.5 GHz SGI AltixEarth Simulator – Earth Simulator Ctr.Yokohama, Japan5120 Processors, 35.86 Tflops/secSX6 VectorData Source: top500.orgCS61C L42 Parallel Computing (12)Carle, Spring 2005 © UCBAdministrivia° HKN evaluations on Monday° Last semester’s final + solutions online° Final exam review• Sunday, 2005-05-08 @ 2pm in 10 Evans° Final exam• Tuesday, 2005-05-14 @ 12:30-3:30pm in 220 Hearst Gym• Same rules as Midterm, except you get 2 double-sided handwritten review sheets (1 from your midterm, 1 new one) + green sheet [Don’t bring backpacks]+ swim trunks (TAs only)CS61C L42 Parallel Computing (13)Carle, Spring 2005 © UCBParallel Programming°Processes and Synchronization°Processor Layout°Other Challenges• Locality• Finding parallelism• Parallel Overhead• Load BalanceCS61C L42 Parallel Computing (14)Carle, Spring 2005 © UCBProcesses°We need a mechanism to intelligently split the execution of a program°Fork:int main(…){ int pid = fork();if (pid == 0) printf(“I am the child.”);if (pid != 0) printf(“I am the parent.”);return 0;}°What will this print?CS61C L42 Parallel Computing (15)Carle, Spring 2005 © UCBProcesses (2)°We don’t know! Two potential orderings:• I am the child.I am the parent.• I am the parent.I am the child.• This situation is a simple race condition.This type of problem can get far more complicated…°Modern parallel compilers and runtime environments hide the details of actually calling fork() and moving the processes to individual processors, but the complexity of synchronizationremainsCS61C L42 Parallel


View Full Document

Berkeley COMPSCI 61C - Lecture 42 – Parallel Computing

Documents in this Course
SIMD II

SIMD II

8 pages

Midterm

Midterm

7 pages

Lecture 7

Lecture 7

31 pages

Caches

Caches

7 pages

Lecture 9

Lecture 9

24 pages

Lecture 1

Lecture 1

28 pages

Lecture 2

Lecture 2

25 pages

VM II

VM II

4 pages

Midterm

Midterm

10 pages

Load more
Download Lecture 42 – Parallel Computing
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Lecture 42 – Parallel Computing and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Lecture 42 – Parallel Computing 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?