1 / 18

Physics 413 Chapter 1

Physics 413 Chapter 1. Computer Architecture. What is a Digital Computer ? A computer is essentially a fast electronic calculating machine.

drew-hoover
Download Presentation

Physics 413 Chapter 1

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Physics 413Chapter 1

  2. Computer Architecture • What is a Digital Computer ? A computer is essentially a fast electronic calculating machine. • What is a program ? A program is a set of very simple instructions that a computer can understand. Examples of simple instructions are ADD two numbers, Is X > Y?, COPY A into the memory. • What is Machine Language ? The set of exceedingly simple and primitive instructions that a computer can understand is called the Instruction Set or Machine Language.

  3. Architecture and Organization • Computer Architecture deals with the design of those components of a computer that are accessible to a programmer. For instance, the Motorola 6800 does not have a MUL instruction but the Pentium does. • Computer Organization deals with the implementation of the architecture and may be transparent to the programmer. For instance, precisely how is MUL implemented in the hardware?

  4. Levels of Abstraction • L5 High‑level Language (C++, Java) • L4 Assembly Language • L3 Operating System • L2 Instruction Set Architecture (ISA) • L1 Microarchitecture Level • L0 Digital Logic Level • Transistors

  5. A Brief History of Computers • First Generation ‑ Vacuum Tubes ‑ 1945‑1955 ENIAC (Electronic Numerical Integrator and Computer) had 18,000 Vacuum Tubes, 1500 relays and weighed 30 tons • Second Generation ‑ Transistors ‑ 1955‑65 ‑ PDP ‑8 (DEC) • Third Generation ‑ IC ‑ 1965‑80 ‑ IBM 360 • Fourth Generation ‑ 1980‑ ? VLSI ‑ Pentium4 ‑PowerPC (G4) ‑ SPARC

  6. Intel and Motorola/IBM Microprocessors CPU Year Data/Address Clock Transistors 4004 1971 4 / 8 108 kHz 2,300 8088 1979 8 / 20 8 MHz 30,000 P4 2000 64 / 32 3 GHz 42 M 6800 1974 8 / 16 1 MHz 6000 G4 1999 64 / 32 800 MHz 10.5 M CISC Complex Instruction Set Computers (6800) RISC Reduced Instruction Set Computers (G4) CISC + RISC ( Pentium 4 ) EPIC Explicitly Parallel Instruction Computing Intel + HP Itanium (www.dell.com)

  7. Moore’s Law The number of transistors on a chip doubles every 18 months Example: From 1992 ‑1995 memory chip size quadrupled from 16M to 64 M

  8. All about microprocessors • ROM is where the program that initializes the PC is stored. • RAM access is fast • ADDRESS BUS carries the address of RAM, ROM, and other peripherals • 6800 has 16 address lines, A0 through A15 • DATA BUS carries data back and forth between the microprocessor and RAM, ROM, and other peripherals. The 6800 has a data bus that is 8 bits wide.

  9. What is a von Neumann Machine? • von Neumann was a world‑renowned mathematician. In 1952 he pointed out the primitive architecture of the ENIAC which had to be programmed by moving jumpers and cables and used 10 switches to represent the digits 0‑9. von Neumann’s machine used binary digits 0‑1 and stored program and data in memory. The ALU had an accumulator. Today’s computers are essentially von Neumann machines.

  10. What is a Virtual Machine? A Virtual Machine is a software implementation for which no direct hardware implementation exists. This is really a buzz word. All high‑level languages and assembly languages are “ virtual machines” ! The term “Java Virtual Machine” is being bandied about a lot these days. It is simply a Java Program. Ultimately, all “virtual machines” have to be implemented in hardware to get “real results”!

  11. Memory • Big Endian: 4 bytes arranged as:0123 (32‑bit word) • Little Endian: 4 bytes arranged 3210 (32‑bit word) • DRAM Organization • nxn matrix structure is popular for large memory sizes • RAS ‑ Row Address Strobe • CAS ‑ Column Address Strobe • RAS ‑ CAS strategy reduces the number of pins but slows down access because two addressing cycles are required • SRAM Fast for cache

  12. ROM • ROM • PROM • EPROM ‑ UV erasable • EEPROM ‑ Electrically erasable byte at a time • FLASH ‑ electrical block erasable ‑ digital film ‑ more like RAM!

  13. Hamming Code • Detecting errors is easy (attach a parity bit) • Correcting errors is complex and requires Hamming code with several parity bits • Hamming distance between two code words is the number of single bit errors required to convert one code word to another. Example: 1101 and 1110 are a distance 2 apart • n = m + r where r = parity bits, m = bits in the original data and n = total number of bits transmitted

  14. 2r >= m+r+1 • Example: For 8‑bit code one needs 4 parity bits for a total of 12 bits • Parity bits are bits 1, 2, 4, 8 etc (P1, P2, P4, P8 ) • 3 = 1+2 • 5 = 1+4 • 6 = 2+4 P1 checks bits 3, 5, 7, 9, 11 • 7 = 1+2+4 P2 checks bits 3,6,7,10,11 • 9 = 1+8 P4 checks bits 5,6,7,12 • 10 =2+8 P8 checks bits 9,10,11,12 • 11 = 1+2+8 • 12 = 4+8

  15. Problemo du Jour ... • Example We need to send 11101001. What 12-bit Hamming word should we send?

  16. Solution de Problemo du Jour … 101111001001

  17. IEEE – 754 Floating Point Format (32-bit) • 1 sign bit 8 exponent bits 23 significand bits • sign bit is 0 for positive numbers and 1 for negative numbers • exponent is in “excess-127” format so an exponent of +4 will be written as +131 and an exponent of –9 will be written as +118. Hence exponents can range from -127 to +128 (written as 0 through 255) • The “normalized” significand must be in the form 1.xxxxx…. Hence only the xxxxx part is actually written in the 23 bits and “1.” is understood

  18. Convert 10.125 to IEEE-754 format • 10.125 (decimal) = +1010 .001000 • = +1 .010001000 x 23 (normalized) • Our sign bit = 0 • Our exponent = 3 or 130 (excess 127) = 10000010 • Our significand = 0100010000 . . . ( 13 more zeros!) • Putting it all together we get . . . • 10.125 = 0100000100100010 . . . (16 more zeros) • = 41220000 (hex)

More Related