CMSC 100 From the Bottom Up: It's All Just Bits Professor Marie desJardins Tuesday, September 11, 2012 CMSC 100 -- Just Bits 1 Tue 9/11/12 Just Bits Inside the computer, all information is stored as bits A “bit” is a single unit of information Each “bit” is set to either zero or one How do we get complex systems like Google, Matlab, and our cell phone apps? 2 CMSC 100 -- Just Bits Tue 9/11/12 Today’s Topics Logic operations and truth tables Logic gates and simple circuits 3 Memory storage Binary numbers Conversions Hexadecimal Representing other data Integers (including negative numbers) Floating point numbers Characters (ASCII) CMSC 100 -- Just Bits Tue 9/11/12 Logic and Truth Tables CMSC 100 -- Just Bits 4 Tue 9/11/12 Manipulating Bits What does a bit value “mean”? 0 = FALSE [off, no] 1 = TRUE [on, yes] Just as in regular algebra, we can think of “variables” that represent a single bit If X is a Boolean variable, then the value of X is either: 0 [FALSE, off, no] or 5 1 [TRUE, on, yes] In algebra, we have operations that we can perform on numbers: Unary operations: Negate, square, square-root, … Binary operations: Add, subtract, multiply, divide, … What are the operations you can imagine performing on bits? Unary: ?? Binary: ?? Thought problem: How many unary operations are there? How many binary operations? CMSC 100 -- Just Bits Tue 9/11/12 Remember Function Tables? X X2 X Y X+Y 0 0 0 0 0 1 1 0 1 1 2 4 0 2 2 3 9 1 0 1 4 16 1 1 2 … … 1 2 3 2 0 2 … … … 6 CMSC 100 -- Just Bits Tue 9/11/12 Boolean Operations: “Truth Tables” 7 Or equivalently… Negation (NOT): X X X X 0 1 F T 1 0 T F Conjunction (AND): X Y XY X Y XY 0 0 0 F F F 0 1 0 F T F 1 0 0 T F F 1 1 1 T T T CMSC 100 -- Just Bits Disjunction (OR): X Y XY X Y XY 0 0 0 F F F 0 1 1 F T T 1 0 1 T F T 1 1 1 T T T Tue 9/11/12 Boolean Expressions EXERCISE: What would a truth table look like for the expression: (A B) (B C) A B C AB B B C (A B) (B C) 8 CMSC 100 -- Just Bits Tue 9/11/12 Boolean Expressions What would a truth table look like for the expression: (A B) (B C) 9 CMSC 100 -- Just Bits A B C 0 0 0 0 0 1 0 1 0 0 1 1 1 0 0 1 0 1 1 1 0 1 1 1 AB B B C (A B) (B C) Tue 9/11/12 Boolean Expressions What would a truth table look like for the expression: (A B) (B C) A B C AB B B C (A B) (B C) 0 0 0 0 1 1 0 0 0 1 0 1 1 0 0 1 0 1 0 0 0 0 1 1 1 0 1 1 1 0 0 1 1 1 1 1 0 1 1 1 1 1 1 1 0 1 0 0 0 1 1 1 1 0 1 1 10 CMSC 100 -- Just Bits Tue 9/11/12 (Boolean) Algebraic Laws DeMorgan’s Theorem Analogous to the distributive law in regular algebra (A B) = A B (A B) = A B A B A B (A B) A B A B F F F T T T T F T F T T F T T F F T F T T T T T F F F F 11 CMSC 100 -- Just Bits Tue 9/11/12 Logic Gates, Circuits, and Memory CMSC 100 -- Just Bits 12 Tue 9/11/12 Storing Bits How are these “bits” stored in the computer? A bit is just an electrical signal or voltage (by convention: “low voltage” = 0; “high voltage” = 1) A circuit called a “flip-flop” can store a single bit A flip-flop can be “set” (using an electrical signal) to either 0 or 1 The flip-flop will hold that value until it receives a new signal telling it to change Bits can be operated on using gates (which “compute” a function of two or more bits) 13 CMSC 100 -- Just Bits Tue 9/11/12 Logic Gates In hardware, Boolean operations are implemented using circuits called logic gates XOR: Exclusive or (one input is TRUE, but not both) 14 CMSC 100 -- Just Bits Tue 9/11/12 Logic Circuits We can implement any logical expression just by assembling the associated logical gates in the right order What would a logic circuit look like for the expression: (A B) (B C) 15 CMSC 100 -- Just Bits Tue 9/11/12 Storing Information One bit can’t tell you much… (just 2 possible values) Usually we group 8 bits together into one “byte” QUESTION:How many possible values (combinations) are there for one byte? A byte can just be thought of as an 8-digit binary (base 2) number Low-order or least significant bit == ones place Michael Littman's octupus counting video Next bit would be “10s place” in base 10 -- what about base 2? High-order bit or most significant bite in a byte == ?? place 16 CMSC 100 -- Just Bits Tue 9/11/12 Orders of Magnitude One 0/1 (“no/yes”) “bit” is the basic unit of memory Eight (23) bits = one byte 1,024 (210) bytes = one kilobyte (1K)* 1,024K = 1,048,576 (220 bytes) = one megabyte (1M) 1,024K (230 bytes) = one gigabyte (1G) 1,024 (240 bytes) = one terabyte (1T) 1,024 (250 bytes) = one petabyte (1P) ... 280 bytes = one yottabyte (1Y?) 17 CMSC 100 -- Just Bits Tue 9/11/12 Scaling Up Memory Computer chip: Each byte has an address (and we use binary numbers to represent those addresses…) 18 Many (millions) of circuits Etched onto a silicon wafer using VLSI (Very Large-Scale Integration) technology Lots of flip-flops or DRAM devices == memory chip An address is represented using a word, which is typically either; 2 bytes (16 bits) -- earliest PCs Only 64K combinations memory is limited to 64K (65,535) bytes! 4 bytes (32 bits) -- first Pentium chips This brings us up to 4G (4,294,967,295) bytes of memory! 8 bytes (64 bits) -- modern Pentium chips Up to 16.8 million terabytes (that’s 18,446,744,073,709,551,615 bytes!) CMSC 100 -- Just Bits Tue 9/11/12 Binary and Hexadecimal Numbers CMSC 100 -- Just Bits 19 Tue 9/11/12 Binary Numbers A binary (base-two) number is just like a decimal (base-ten) number, except that instead of ten possible digits (0...9), we only have two (0...1) 1510 47710 tens place ones place 1112 twos place fours place tens place hundreds place ones place 1110111012 QUESTION: What place value does this zero hold? ones place 20 CMSC 100 -- Just Bits Tue 9/11/12 Conversions Binary decimal: multiply each digit by its place value and add results Decimal binary: 21 Find the largest power of two that is less than or equal to the decimal number Put a one in that place column in the binary number Add spaces for the smaller binary digits E.g., if the largest multiple of two that fits is 256, you would write: 1 __ __ __ __ __ __ __ __ Find the next smallest multiple of two Put a one in that place column (and a zero in any columns you skipped) EXERCISE: You try it! Write 69710 in binary Write 110011102 in decimal CMSC 100 -- Just Bits2 Tue 9/11/12 Hexadecimal It would be very inconvenient to write out a 64-bit address in binary: 0010100111010110111110001001011000010001110011011110000011100000 Instead, we group each set of 4 bits together into a hexadecimal (base 16) digit: The digits are 0, 1, 2, …, 9, A (10), B (11), …, E (14), F (15) 0010 1001 1101 0110 1111 1000 1001 0110 0001 0001 1100 1101 1110 0000 1110 0000 2 9 D 6 F 8 9 6 1 1 C D E 0 E 0 …which we write, by convention, with a “0x” preceding the number to indicate it’s a heXadecimal number: 0x29D6F89611CDE0E0 22 CMSC 100 -- Just Bits Tue 9/11/12 Representing Information Positive integers: Just use the binary number system Negative integers, letters, images, … not so easy! There are many different ways to represent information Some are more efficient than others … but once we’ve solved the representation problem, we can use that information without considering how it’s represented… via 23 CMSC 100 -- Just Bits Tue 9/11/12 Representing Integers Simplest idea (“ones’ complement”): Use one bit for a “sign bit”: 1 means negative, 0 means positive The other bits are “complemented” (flipped) in a negative number So, for example, +23 (in a 16-bit word) is represented as: 0000000000010111 and -23 is represented as: 1111111111101000 But there are two different ways to say “zero” (0000… and 1111…) It’s tricky to do simple arithmetic operations like addition in the ones’ complement notation 24 CMSC 100 -- Just Bits This is solved by the twos’ complement representation, but we won’t go over that Tue 9/11/12 Floating Point Numbers Non-integers are a problem… Remember that any rational number can be represented as a fraction Irrational numbers can’t be written down at all, of course Notice that any representation we choose will by definition have limited precision, since we can only represent 232 different values in a 32-bit word 25 …but we probably don’t want to do this, since (a) we’d need to use two words for each number (i.e., the numerator and the denominator) (b) fractions are hard to manipulate (add, subtract, etc.) 1/3 isn’t exactly 1/3 (let’s try it on a calculator!) In general, we also lose precision (introduce errors) when we operate on floating point numbers You don’t need to know the details of how “floating point” numbers are represented CMSC 100 -- Just Bits Tue 9/11/12 Representing Characters ASCII representation: one byte [actually 7 bits…] == one letter == an integer from 0-128 26 CMSC 100 -- Just Bits [Chart borrowed from ha.ckers.org] Tue 9/11/12 Summary: Main Ideas It’s all just bits Abstraction Boolean algebra Representing numbers TRUE/FALSE Truth tables Logic gates Hexadecimal representation Ones’ complement Floating point numbers (main issues) ASCII representation (main idea) 27 CMSC 100 -- Just Bits Tue 9/11/12 ACTIVITY (if time) Design a one-bit adder (i.e., a logic circuit that adds two 1-bit numbers together) X + Y Z2 Z1 Z2 is needed since the result may be two binary digits long First let’s figure out the Boolean expression for each output… Create the truth table Then we’ll draw the logic circuit 28 CMSC 100 -- Just Bits Tue 9/11/12