If you build it, will they buy it?

Slides:



Advertisements
Similar presentations
Introduction Companion slides for The Art of Multiprocessor Programming by Maurice Herlihy & Nir Shavit TexPoint fonts used in EMF. Read the TexPoint manual.
Advertisements

Computer Organization. The Nature of Data Spreadsheets give us a way to create processes that manipulate data We can think of a piece of data as a quantity.
Multiprocessors— Large vs. Small Scale Multiprocessors— Large vs. Small Scale.
Lecture 6: Multicore Systems
Computer Systems Nat 4/5 Computing Science Types of Computer and Performance.
Dr. Alexandra Fedorova August 2007 Introduction to Systems Research at SFU.
Hit or Miss ? !!!.  Cache RAM is high-speed memory (usually SRAM).  The Cache stores frequently requested data.  If the CPU needs data, it will check.
CSCE101 – 4.2, 4.3 October 17, Power Supply Surge Protector –protects from power spikes which ruin hardware. Voltage Regulator – protects from insufficient.
Many-Core Programming with GRAMPS Jeremy Sugerman Kayvon Fatahalian Solomon Boulos Kurt Akeley Pat Hanrahan.
 States that the number of transistors on a microprocessor will double every two years.  Current technology is approaching physical limitations. The.
ECE 510 Brendan Crowley Paper Review October 31, 2006.
Dr. Gheith Abandah, Chair Computer Engineering Department The University of Jordan 20/4/20091.
Energy Model for Multiprocess Applications Texas Tech University.
Single-Chip Multi-Processors (CMP) PRADEEP DANDAMUDI 1 ELEC , Fall 08.
Operating Systems Should Manage Accelerators Sankaralingam Panneerselvam Michael M. Swift Computer Sciences Department University of Wisconsin, Madison,
LOGO Multi-core Architecture GV: Nguyễn Tiến Dũng Sinh viên: Ngô Quang Thìn Nguyễn Trung Thành Trần Hoàng Điệp Lớp: KSTN-ĐTVT-K52.
Multi-Threading and Load Balancing Compiled by Paul TaylorCSE3AGR Stolen mainly from Orion Granatir
Last Time Performance Analysis It’s all relative
Parallel and Distributed Systems Instructor: Xin Yuan Department of Computer Science Florida State University.
CS101 Four Concepts Four Important Concepts. Concept 1: Computers are stupid! However, they are getting smarter! Cognitive Computing.
CSE 691: Energy-Efficient Computing Lecture 7 SMARTS: custom-made systems Anshul Gandhi 1307, CS building
Sogang University Advanced Computing System Chap 1. Computer Architecture Hyuk-Jun Lee, PhD Dept. of Computer Science and Engineering Sogang University.
Dragged, Kicking and Screaming: Multicore Architecture and Video Games.
Outline  Over view  Design  Performance  Advantages and disadvantages  Examples  Conclusion  Bibliography.
OLE Slide No. 1 Object Linking and Embedding H OLE H definition H add other information to documents H copy.
Central Processing Unit (CPU)
Shouqing Hao Institute of Computing Technology, Chinese Academy of Sciences Processes Scheduling on Heterogeneous Multi-core Architecture.
CS101 Parts of a Computer. What we know! To be a computer it needs to do four things: –1 –2 –3 –4 We might already know a lot about three of these requirements.
Multi-core, Mega-nonsense. Will multicore cure cancer? Given that multicore is a reality –…and we have quickly jumped from one core to 2 to 4 to 8 –It.
CS203 – Advanced Computer Architecture
Processor Level Parallelism 2. How We Got Here Developments in PC CPUs.
Computer Stuff By Raj and Brad. Virtual Memory This takes place if your computer is running low on RAM. It stimulates more RAM when your computer reaches.
INTRODUCTION TO HIGH PERFORMANCE COMPUTING AND TERMINOLOGY.
Hardware refers to the tangible parts of computer systems and typically includes support for processing, storage, input, and output. Hardware Processing.
SPRING 2012 Assembly Language. Definition 2 A microprocessor is a silicon chip which forms the core of a microcomputer the concept of what goes into a.
Directions in Linux OpenGL
Conclusions on CS3014 David Gregg Department of Computer Science
High Performance Computer Architecture:
Multiple processor systems
Lynn Choi School of Electrical Engineering
Anshul Gandhi 347, CS building
Introduction.
CS6401- OPERATING SYSTEMS L T P C
Design Process You already know the Scientific Method process. Today we are going to learn about the process engineers use when inventing new products.
Central Processing Unit- CPU
Distributed Databases
The University of Adelaide, School of Computer Science
Nat 4/5 Computing Science Types of Computer and Performance
How We Think Of Computers
Architecture & Organization 1
iMovie Mac software for creating movies from video
EE 193: Parallel Computing
Nat 4/5 Computing Science Types of Computer and Performance
Saturday, 10 November 2018 Binary
Introduction.
Architecture & Organization 1
CS101 Parts of a Computer.
Computer Based Technology:
Chapter 1 Introduction.
Dr. Tansel Dökeroğlu University of Turkish Aeronautical Association Computer Engineering Department Ceng 442 Introduction to Parallel.
Design Process You already know the Scientific Method process. Today we are going to learn about the process engineers use when inventing new products.
What is Computer Architecture?
August 22 Bellringer What is engineering?.
Hardware Main memory 26/04/2019.
Chapter-5 Traffic Engineering.
Lecture 20 Parallel Programming CSE /27/2019.
ALF Amdhal’s Law is Forever
CSE378 Introduction to Machine Organization
Course Code 114 Introduction to Computer Science
Logic Gates and Memory.
Presentation transcript:

If you build it, will they buy it? Processors will be multicore It’s the only way the chip companies know how to use all those transistors! (pesky Moore’s law) But, does the average computer user need it? Probably not – the average user probably already has way too much compute power! But, nobody has ever looked smart predicting that computers are as powerful as they need to be…

Hetero vs. Homogeneous The easy path to Multicore is simply tile some old cores you have sitting around Old being the operative word Intel says – “Don’t worry, if you’re burning too much power, you can just shut off most of the cores.” Heterogeneous makes more sense! Only turn things on when you need them When does Multicore become System on Chip? Already the m.o. for embedded stuff

Pick your apps carefully! Hmmm… How about hardware for Ray Tracing? Loads of completely independent threads Read-only database Write-only output to frame buffer Woo Hoo! Tiling these processors might actually make sense Beware the memory, pin, power, wire, etc. issues! Just because it’s embarrassing doesn’t mean it’s easy!

How about Asynchrony? I’m a member of a strange religion - the Asynchronous Systems cult With lots of (heterogeneous?) cores on the same chip, how will they deal with communication, coordination, cooperation, contention, consumption, chronology? How about asynchronous protocols? Hmmm… It’s so crazy, it just might work!