CSE 305 Computer Architecture Computer Architecture n Course

CSE 305 Computer Architecture

Computer Architecture n Course Teachers n n n Textbook n n Rifat Shahriyar (rifat 1816@gmail. com) Johra Muhammad Moosa Computer Organization and Design (The Hardware/Software Interface) th Ed. ) n David A. Patterson and John L. Hennessy (5 Homepage n http: //rifatshahriyar. github. io/CSE 305. html

Chapter 1 Computer Abstractions and Technology

n Progress in computer technology n n Makes novel applications feasible n n n Underpinned by Moore’s Law § 1. 1 Introduction The Computer Revolution Computers in automobiles Cell phones Human genome project World Wide Web Search Engines Computers are pervasive Chapter 1 — Computer Abstractions and Technology — 4

Moore’s Law n n The number of transistors in a dense integrated circuit doubles approximately every two years Named after Gordon Moore, the co-founder of Intel n n n 1965 - described a doubling every year 1975 - revised the forecast to doubling every two years The period is often quoted as 18 months Chapter 1 — Computer Abstractions and Technology — 5

Moore’s Law Chapter 1 — Computer Abstractions and Technology — 6

Classes of Computers n Personal computers n n n General purpose, variety of software Subject to cost/performance tradeoff Server computers n n n Network based High capacity, performance, reliability Range from small servers to building sized Chapter 1 — Computer Abstractions and Technology — 7

Classes of Computers n Supercomputers n n n High-end scientific and engineering calculations Highest capability but represent a small fraction of the overall computer market Embedded computers n n Hidden as components of systems Stringent power/performance/cost constraints Chapter 1 — Computer Abstractions and Technology — 8

The Post. PC Era Chapter 1 — Computer Abstractions and Technology — 9

The Post. PC Era n Personal Mobile Device (PMD) n n n Battery operated Connects to the Internet Hundreds of dollars Smart phones, tablets, electronic glasses Cloud computing n n Warehouse Scale Computers (WSC) Software as a Service (Saa. S) Portion of software run on a PMD and a portion run in the Cloud Amazon and Google Chapter 1 — Computer Abstractions and Technology — 10

What You Will Learn n How programs are translated into the machine language n n n The hardware/software interface What determines program performance n n n And how the hardware executes them And how it can be improved How hardware designers improve performance What is parallel processing Chapter 1 — Computer Abstractions and Technology — 11

Understanding Performance n Algorithm n n Programming language, compiler, architecture n n Determine number of machine instructions executed per operation Processor and memory system n n Determines number of operations executed Determine how fast instructions are executed I/O system (including OS) n Determines how fast I/O operations are executed Chapter 1 — Computer Abstractions and Technology — 12

n Design for Moore’s Law n Use abstraction to simplify design n Make the common case fast n Performance via parallelism n Performance via pipelining n Performance via prediction n Hierarchy of memories n Dependability via redundancy § 1. 2 Eight Great Ideas in Computer Architecture Eight Great Ideas Chapter 1 — Computer Abstractions and Technology — 13

n Application software n n Written in high-level language System software n n Compiler: translates HLL code to machine code Operating System: service code n n § 1. 3 Below Your Program Handling input/output Managing memory and storage Scheduling tasks & sharing resources Hardware n Processor, memory, I/O controllers Chapter 1 — Computer Abstractions and Technology — 15

Levels of Program Code n High-level language n n n Assembly language n n Level of abstraction closer to problem domain Provides for productivity and portability Textual representation of instructions Hardware representation n n Binary digits (bits) Encoded instructions and data Chapter 1 — Computer Abstractions and Technology — 16

The BIG Picture n Same components for all kinds of computer n n Desktop, server, embedded § 1. 4 Under the Covers Components of a Computer Input/output includes n User-interface devices n n Storage devices n n Display, keyboard, mouse Hard disk, CD/DVD, flash Network adapters n For communicating with other computers Chapter 1 — Computer Abstractions and Technology — 17

Anatomy of a Computer Output device Network cable Input device Chapter 1 — Computer Abstractions and Technology — 18

Anatomy of a Mouse n Optical mouse n n n LED illuminates desktop Small low-res camera Basic image processor n n n Looks for x, y movement Buttons & wheel Supersedes roller-ball mechanical mouse Chapter 1 — Computer Abstractions and Technology — 19

Through the Looking Glass n LCD screen: picture elements (pixels) n Mirrors content of frame buffer memory Chapter 1 — Computer Abstractions and Technology — 20

Touchscreen n Post. PC device Supersedes keyboard and mouse Resistive and Capacitive types n n Most tablets, smart phones use capacitive Capacitive allows multiple touches simultaneously Chapter 1 — Computer Abstractions and Technology — 21

Opening the Box Capacitive multitouch LCD screen 3. 8 V, 25 Watt-hour battery Computer board Chapter 1 — Computer Abstractions and Technology — 22

Inside the Processor (CPU) n n n Datapath: performs operations on data Control: sequences datapath, memory, I/O devices. . . Cache memory n Small fast SRAM memory for immediate access to data Chapter 1 — Computer Abstractions and Technology — 23

Inside the Processor n Apple A 5 Chapter 1 — Computer Abstractions and Technology — 24

Inside the Processor n AMD Barcelona: 4 processor cores Chapter 1 — Computer Abstractions and Technology — 25

Inside the Processor n Northbridge and Southbridge Chapter 1 — Computer Abstractions and Technology — 26

Inside the Processor n Northbridge and Southbridge Chapter 1 — Computer Abstractions and Technology — 27

Abstractions The BIG Picture n Abstraction helps us deal with complexity n n Instruction set architecture (ISA) n n The hardware/software interface Application binary interface n n Hide lower-level detail The ISA plus system software interface Implementation n The details underlying and interface Chapter 1 — Computer Abstractions and Technology — 28

Abstractions Problem Operating System Interface between Architecture Software and Hardware Circuits Chapter 1 — 29

A Safe Place for Data n Volatile main memory n n Loses instructions and data when power off Non-volatile secondary memory n n n Magnetic disk Flash memory Optical disk (CDROM, DVD) Chapter 1 — Computer Abstractions and Technology — 31

Networks n n Communication and resource sharing Local area network (LAN): Ethernet n n n Within a building Wide area network (WAN: the Internet Wireless network: Wi. Fi, Bluetooth Chapter 1 — Computer Abstractions and Technology — 32

n Electronics technology continues to evolve n n Increased capacity and performance Reduced cost DRAM capacity Year Technology Relative performance/cost 1951 Vacuum tube 1965 Transistor 1975 Integrated circuit (IC) 1995 Very large scale IC (VLSI) 2013 Ultra large scale IC 1 35 900 2, 400, 000 § 1. 5 Technologies for Building Processors and Memory Technology Trends 250, 000, 000 Chapter 1 — Computer Abstractions and Technology — 33

n Which airplane has the best performance? § 1. 6 Performance Defining Performance Chapter 1 — Computer Abstractions and Technology — 34

Response Time and Throughput n Response time n n n Throughput n n n How long it takes to do a task Individual computer users interest Total work done per unit time (tasks per hour) Datacenter managers interest How are response time and throughput affected by n n Replacing the processor with a faster version Adding additional processors to a system that uses multiple processors for separate tasks Chapter 1 — Computer Abstractions and Technology — 35

Relative Performance n Define Performance = 1/Execution Time “X is n time faster than Y” n Example: time taken to run a program n n 10 s on A, 15 s on B Execution Time. B / Execution Time. A = 15 s / 10 s = 1. 5 So A is 1. 5 times faster than B Chapter 1 — Computer Abstractions and Technology — 36

Measuring Execution Time n Wall Clock time/Elapsed time n Total response time, including all aspects n n n Processing, I/O, OS overhead, idle time Determines system performance CPU time n Time spent processing a given job n n n Discounts I/O time, other jobs shares Comprises user CPU time and system CPU time Different programs are affected differently by CPU and system performance Chapter 1 — Computer Abstractions and Technology — 37

CPU Clocking n Operation of digital hardware governed by a constant-rate clock Clock period Clock (cycles) Data transfer and computation Update state n Clock period: duration of a clock cycle n n e. g. , 250 ps = 0. 25 ns = 250× 10– 12 s Clock frequency (rate): cycles per second n e. g. , 4. 0 GHz = 4000 MHz = 4. 0× 109 Hz Chapter 1 — Computer Abstractions and Technology — 38

CPU Time n Performance improved by n n Reducing number of clock cycles Increasing clock rate Hardware designer must often trade off clock rate against cycle count Many techniques that decrease the number of clock cycles may also increase the clock cycle time Chapter 1 — Computer Abstractions and Technology — 39

CPU Time Example n n Computer A: 2 GHz clock, 10 s CPU time Designing Computer B n n n Aim for 6 s CPU time Can do faster clock, but causes 1. 2 × clock cycles How fast must Computer B clock be? Chapter 1 — Computer Abstractions and Technology — 40

Instruction Count and CPI n Instruction Count for a program n n Determined by program, ISA and compiler Average cycles per instruction n n Determined by CPU hardware If different instructions have different CPI n Average CPI affected by instruction mix Chapter 1 — Computer Abstractions and Technology — 41

CPI Example n n Computer A: Cycle Time = 250 ps, CPI = 2. 0 Computer B: Cycle Time = 500 ps, CPI = 1. 2 Same ISA Which is faster, and by how much? A is faster… …by this much Chapter 1 — Computer Abstractions and Technology — 42

CPI in More Detail n If different instruction classes take different numbers of cycles n Weighted average CPI Relative frequency Chapter 1 — Computer Abstractions and Technology — 43

CPI Example Chapter 1 — Computer Abstractions and Technology — 44

CPI Example n n Alternative compiled code sequences using instructions in classes A, B, C Class A B C CPI for class 1 2 3 IC in sequence 1 2 IC in sequence 2 4 1 1 Sequence 1: IC = 5 n n Clock Cycles = 2× 1 + 1× 2 + 2× 3 = 10 Avg. CPI = 10/5 = 2. 0 n Sequence 2: IC = 6 n n Clock Cycles = 4× 1 + 1× 2 + 1× 3 =9 Avg. CPI = 9/6 = 1. 5 Chapter 1 — Computer Abstractions and Technology — 45

Performance Summary The BIG Picture n Performance depends on n n Algorithm: affects IC, possibly CPI Programming language: affects IC, CPI Compiler: affects IC, CPI Instruction set architecture: affects IC, CPI, Tc Chapter 1 — Computer Abstractions and Technology — 46

Chapter 1 — Computer Abstractions and Technology — 47

n § 1. 7 The Power Wall Power Trends In CMOS IC technology Chapter 1 — Computer Abstractions and Technology — 48

Reducing Power n Suppose a new CPU has n n n The power wall n n n 85% of capacitive load of old CPU 15% voltage and 15% frequency reduction We can’t reduce voltage further We can’t remove more heat How else can we improve performance? Chapter 1 — Computer Abstractions and Technology — 49

Power vs. Performance Chapter 1 — Computer Abstractions and Technology — 50

§ 1. 8 The Sea Change: The Switch to Multiprocessors Uniprocessor Performance Constrained by power, instruction-level parallelism, memory latency Chapter 1 — Computer Abstractions and Technology — 51

Multiprocessors n Multicore microprocessors n n More than one processor per chip Requires explicitly parallel programming n Compare with instruction level parallelism n n n Hardware executes multiple instructions at once Hidden from the programmer Hard to do n n n Programming for performance Load balancing Optimizing communication and synchronization Chapter 1 — Computer Abstractions and Technology — 52

n Programs used to measure performance n n Standard Performance Evaluation Corp (SPEC) n n Supposedly typical of actual workload Develops benchmarks for CPU, I/O, Web, … SPEC CPU 2006 n Elapsed time to execute a selection of programs n n n § 1. 9 SPEC CPU Benchmark Negligible I/O, so focuses on CPU performance Normalize relative to reference machine Summarize as geometric mean of performance ratios n CINT 2006 (integer) and CFP 2006 (floating-point) Chapter 1 — Computer Abstractions and Technology — 53

CINT 2006 for Intel Core i 7 920 Chapter 1 — Computer Abstractions and Technology — 54

n n Improving an aspect of a computer and expecting a proportional improvement in overall performance Example: multiply accounts for 80 s/100 s n How much improvement in multiply performance to get 5× overall? n n § 1. 10 Fallacies and Pitfalls Pitfall: Amdahl’s Law Can’t be done! Corollary: make the common case fast Chapter 1 — Computer Abstractions and Technology — 55

Fallacy: Low Power at Idle n Look back at i 7 power benchmark n n Google data center n n n At 100% load: 258 W At 50% load: 170 W (66%) At 10% load: 121 W (47%) Mostly operates at 10% – 50% load At 100% load less than 1% of the time Consider designing processors to make power proportional to load Chapter 1 — Computer Abstractions and Technology — 56

Pitfall: MIPS as a Performance Metric n MIPS: Millions of Instructions Per Second n Doesn’t account for n n n Differences in ISAs between computers Differences in complexity between instructions CPI varies between programs on a given CPU Chapter 1 — Computer Abstractions and Technology — 57

Pitfall: MIPS as a Performance Metric n Computer A n n n Computer B n n 10 Billion instructions, 4 GHz Clock rate CPI = 1 8 Billion instructions, 4 GHz Clock rate CPI = 1. 1 1) Which has the highest MIPS rating? 2) Which is faster? Chapter 1 — Computer Abstractions and Technology — 58

n Cost/performance is improving n n Hierarchical layers of abstraction n In both hardware and software Instruction set architecture n n Due to underlying technology development § 1. 11 Concluding Remarks The hardware/software interface Execution time: the best performance measure Power is a limiting factor n Use parallelism to improve performance Chapter 1 — Computer Abstractions and Technology — 59