SlideShare a Scribd company logo

Unit 2 arithmetics

Download as PPT, PDF
C
chidabdu

The result is incorrect as sign extension was not considered. For signed multiplication, sign extension of operands is required before performing multiplication as for unsigned numbers. Signed Multiplication (contd..)  For signed multiplication:  Extend the sign bit of both operands to fill the register.  Perform multiplication as for unsigned numbers.  The final product now represents a signed number in 2's complement form. Example: (-13) × (+11) -13 in 2's complement is 1 0 0 1 1 +11 in 2's complement is 0 1 0 1 1 After sign extension: -13 = 1 1 1 1 1

EducationBusinessTechnology
Related topics:
Computer Organization
1 of 79
Downloaded 378 times
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
UNIT 2
xi yi Carry­in ci Sum s i Carry­out c i+1 At the ith stage: 0 0 0 0 0 Input: 0 0 1 1 0 ci is the carry-in 0 1 0 1 0 0 1 1 0 1 Output: 1 0 0 1 0 si is the sum 1 0 1 0 1 1 1 0 0 1 ci+1 carry-out to (i+1)st 1 1 1 1 1 state si   = xi yi ci + xi yi ci + xi yi ci + xi yi ci = x i ⊕ yi ⊕ ci ci +1 = yi c i + x i ci + x i y i Example: X 7 0 1 1 1 Carry­out xi Carry­in + Y = + 6 = +00 1 1 1 1 0 0 0 yi ci+1 ci Z 13 1 1 0 1 si Legend for stage i
Sum Carry yi c i xi xi yi si c c i +1 i ci x xi yi i yi ci + 1 Full adder ci (FA) s i Full Adder (FA): Symbol for the complete circuit for a single stage of addition.
•Cascade n full adder (FA) blocks to form a n-bit adder. •Carries propagate or ripple through this cascade, n-bit ripple carry adder. xn ­ 1 yn­ 1 x1 y1 x0 y0 cn ­ 1 c1 cn FA FA FA c0 sn ­ 1 s1 s0 Most significant bit Least significant bit (MSB) position (LSB) position Carry-in c0 into the LSB position provides a convenient way to perform subtraction.
K n-bit numbers can be added by cascading k n-bit adders. xk n ­ 1 yk n ­ 1 x2n ­ 1 y2n ­ 1 xn y n xn ­ 1 y n ­ 1 x 0 y 0 cn n­bit n­bit n­bit c c kn 0 adder adder adder s s( s s s s kn ­ 1 k ­ 1) n 2n ­ 1 n n­ 1 0 Each n-bit adder forms a block, so this is cascading of blocks. Carries ripple or propagate through blocks, Blocked Ripple Carry Adder
•Recall X – Y is equivalent to adding 2’s complement of Y to X. •2’s complement is equivalent to 1’s complement + 1. •X – Y = X + Y + 1 •2’s complement of positive and negative numbers is computed similarly. xn ­ 1 yn­ 1 x1 y1 x0 y0 cn ­ 1 c1 cn FA FA FA 1 sn ­ 1 s1 s0 Most significant bit Least significant bit (MSB) position (LSB) position
y y y n­ 1 1 0 Add/Sub control x x x n­ 1 1 0 c n­bit adder n c 0 s s s n­ 1 1 0 •Add/sub control = 0, addition. •Add/sub control = 1, subtraction.
Detecting overflows  Overflows can only occur when the sign of the two operands is the same.  Overflow occurs if the sign of the result is different from the sign of the operands.  Recall that the MSB represents the sign.  xn-1, yn-1, sn-1 represent the sign of operand x, operand y and result s respectively.  Circuit to detect overflow can be implemented by the following logic expressions: Overflow = xn −1 yn −1sn −1 + xn −1 yn −1sn −1 Overflow = cn ⊕ cn −1
x0 y0 Consider 0th stage: •c1 is available after 2 gate delays. •s1 is available after 1 gate delay. c1 FA c0 s0 Sum Carry yi c i xi xi yi si c c i +1 i ci x i yi
Cascade of 4 Full Adders, or a 4-bit adder x0 y0 x0 y0 x0 y0 x0 y0 FA FA FA FA c0 c4 c3 c2 c1 s3 s2 s1 s0 •s0 available after 1 gate delays, c1 available after 2 gate delays. •s1 available after 3 gate delays, c2 available after 4 gate delays. •s2 available after 5 gate delays, c3 available after 6 gate delays. •s3 available after 7 gate delays, c4 available after 8 gate delays. For an n-bit adder, sn-1 is available after 2n-1 gate delays cn is available after 2n gate delays.
Recall the equations: si = xi ⊕ yi ⊕ ci ci +1 = xi yi + xi ci + yi ci Second equation can be written as: ci +1 = xi yi + ( xi + yi )ci We can write: ci +1 = Gi + Pi ci where Gi = xi yi and Pi = xi + yi •Gi is called generate function and Pi is called propagate function •Gi and Pi are computed only from xi and yi and not ci, thus they can be computed in one gate delay after X and Y are applied to the inputs of an n-bit adder.
ci +1 = Gi + Pi ci ci = Gi −1 + Pi −1ci −1 ⇒ ci+1 = Gi + Pi (Gi −1 + Pi −1ci −1 ) continuing ⇒ ci+1 = Gi + Pi (Gi −1 + Pi −1 (Gi − 2 + Pi− 2 ci −2 )) until ci+1 = Gi + PiGi −1 + Pi Pi−1 Gi −2 + .. + Pi Pi −1 ..P1G0 + Pi Pi −1 ...P0 c 0 •All carries can be obtained 3 gate delays after X, Y and c0 are applied. -One gate delay for Pi and Gi -Two gate delays in the AND-OR circuit for ci+1 •All sums can be obtained 1 gate delay after the carries are computed. •Independent of n, n-bit addition requires only 4 gate delays. •This is called Carry Lookahead adder.
x y x y x y x y 3 3 2 2 1 1 0 0 c4 c 3 c 2 c 1 . c 4-bit carry-lookahead B cell B cell B cell B cell 0 adder s s s s 3 2 1 0 G3 P3 G2 P2 G P G P 1 1 0 0 Carry­lookahead logic xi yi . . . c i B-cell for a single stage B cell Gi P i si
Carry lookahead adder (contd..)  Performing n-bit addition in 4 gate delays independent of n is good only theoretically because of fan-in constraints. ci+1 = Gi + PiGi −1 + Pi Pi−1 Gi −2 + .. + Pi Pi −1 ..P1G0 + P Pi −1 ...P0 c0 i  Last AND gate and OR gate require a fan-in of (n+1) for a n- bit adder.  For a 4-bit adder (n=4) fan-in of 5 is required.  Practical limit for most gates.  In order to add operands longer than 4 bits, we can cascade 4-bit Carry-Lookahead adders. Cascade of Carry-Lookahead adders is called Blocked Carry-Lookahead adder.
Unit 2 arithmetics
Carry-out from a 4-bit block can be given as: c4 = G3 + P G2 + P3 P2 G1 + P3 P2 P G0 + P3 P2 P1P0 c0 3 1 Rewrite this as: P0I = P3 P2 P1 P0 G0I = G3 + P3 G2 + P3 P2 G1 + P3 P2 P1G0 Subscript I denotes the blocked carry lookahead and identifies the block. Cascade 4 4-bit adders, c16 can be expressed as: c16 = G3 + P3I G2I + P3I P2I G1I + P3I P2I P10 G0 + P3I P2I P10 P00 c0 I I
x15­12 y15­12 x11­8 y11­8 x7­4 y7­4 x3­0 y3­0 c16 4­bit adder c12 4­bit adder c8 4­bit adder c4 4­bit adder . c0 s15­12 s11­8 s7­4 s3­0 G3 I P3 I G2 I P 2I G 1I P 1I G 0I P0 I Carry­lookahead logic After xi, yi and c0 are applied as inputs: - Gi and Pi for each stage are available after 1 gate delay. - PI is available after 2 and GI after 3 gate delays. - All carries are available after 5 gate delays. - c16 is available after 5 gate delays. - s15 which depends on c12 is available after 8 (5+3)gate delays (Recall that for a 4-bit carry lookahead adder, the last sum bit is available 3 gate delays after all inputs are available)
Unit 2 arithmetics
Product of 2 n-bit numbers is at most a 2n-bit number. Unsigned multiplication can be viewed as addition of shifted versions of the multiplicand.
Multiplication of unsigned numbers (contd..)  We added the partial products at end.  Alternative would be to add the partial products at each stage.  Rules to implement multiplication are:  If the ith bit of the multiplier is 1, shift the multiplicand and add the shifted multiplicand to the current value of the partial product.  Hand over the partial product to the next stage  Value of the partial product at the start stage is 0.
Typical multiplication cell Bit of incoming partial product (PPi) jth multiplicand bit ith multiplier bit ith multiplier bit carry out FA carry in Bit of outgoing partial product (PP(i+1))
Combinatorial array multiplier Multiplicand 0 m3 0 m2 0 m1 0 m0 (PP0) q0 0 PP1 p0 q1 0 r lie PP2 tip p1 ul q2 M 0 PP3 p2 q3 0 , p7 p6 p5 p4 p3 Product is: p7,p6,..p0 Multiplicand is shifted by displacing it through an array of adders.
Combinatorial array multiplier (contd..)  Combinatorial array multipliers are:  Extremely inefficient.  Have a high gate count for multiplying numbers of practical size such as 32- bit or 64-bit numbers.  Perform only one function, namely, unsigned integer product.  Improve gate efficiency by using a mixture of combinatorial array techniques and sequential techniques requiring less combinational logic.
Sequential multiplication  Recall the rule for generating partial products:  If the ith bit of the multiplier is 1, add the appropriately shifted multiplicand to the current partial product.  Multiplicand has been shifted left when added to the partial product.  However, adding a left-shifted multiplicand to an unshifted partial product is equivalent to adding an unshifted multiplicand to a right-shifted partial product.
Register A  (initially 0) Shift right C a a q q n ­ 1 0 n­ 1 0 Multiplier Q Add/Noadd control n-bit Adder MUX Control sequencer 0 0 m m0 n ­ 1 Multiplicand M
M 1   1   0   1 Initial configuration 0 0   0   0   0 1   0   1   1 C A Q 0 1   1   0   1 1   0   1   1 Add Shift First cycle 0 0   1   1   0 1   1   0   1 1 0   0   1   1 1   1   0   1 Add Shift Second cycle 0 1   0   0   1 1   1   1   0 0 1   0   0   1 1   1   1   0 No add Shift Third cycle 0 0   1   0   0 1   1   1   1 1 0   0   0   1 1   1   1   1 Add Shift Fourth cycle 0 1   0   0   0 1   1   1   1 Product
Unit 2 arithmetics
Signed Multiplication  Considering 2’s-complement signed operands, what will happen to (-13)×(+11) if following the same method of unsigned multiplication? 1 0 0 1 1 ( ­ 13) 0 1 0 1 1 ( + 11) 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1 0 0 1 1 Sign extension is shown in blue 0 0 0 0 0 0 0 0 1 1 1 0 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 0 0 1 ( ­ 143) Sign extension of negative multiplicand.
Signed Multiplication  For a negative multiplier, a straightforward solution is to form the 2’s-complement of both the multiplier and the multiplicand and proceed as in the case of a positive multiplier.  This is possible because complementation of both operands does not change the value or the sign of the product.  A technique that works equally well for both negative and positive multipliers – Booth algorithm.
Signed Multiplication Hardware Implementation
Booth Algorithm Consider in a multiplication, the multiplier is positive 0011110, how many appropriately shifted versions of the multiplicand are added in a standard procedure? 0 1 0 1 1 0 1 0 0 +1 +1 + 1+1 0 0 0 0 0 0 0 0 0 1 0 1 1 0 1 0 1 0 1 1 0 1 0 1 0 1 1 0 1 0 1 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 1 0 0 0 1 1 0
Booth Algorithm  Since 0011110 = 0100000 – 0000010, if we use the expression to the right, what will happen? 0 1 0 1 1 0 1 0 +1 0 0 0 ­ 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2's complement of 1 1 1 1 1 1 1 0 1 0 0 1 1 the multiplicand 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 1 0 0 0 1 1 0
Booth in the Booth scheme, -1 times the shifted multiplicand is In general, Algorithm selected when moving from 0 to 1, and +1 times the shifted multiplicand is selected when moving from 1 to 0, as the multiplier is scanned from right to left. 0 0 1 0 1 1 0 0 1 1 1 0 1 0 1 1 0 0 0 +1 ­1 +1 0 ­ 1 0 +1 0 0 ­ 1 +1 ­ 1 + 1 0 ­ 1 0 0 Booth recoding of a multiplier.
Booth Algorithm 0 1 1 0 1 ( + 13) 0 1 1 0 1 X1 1 0 1 0 (­ 6) 0 ­ 1 +1 ­ 1 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 0 0 1 1 0 0 0 0 1 1 0 1 1 1 1 0 0 1 1 0 0 0 0 0 0 1 1 1 0 1 1 0 0 1 0 ( ­ 78) Booth multiplication with a negative multiplier.
Booth Algorithm Multiplier V ersion of multiplicand selected by biti Bit i Bit i ­1 0 0 0 X M 0 1 + 1 X M 1 0 − 1 X M 1 1 0 X M Booth multiplier recoding table.
Booth Algorithm  Best case – a long string of 1’s (skipping over 1s)  Worst case – 0’s and 1’s are alternating 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 Worst­case multiplier +1 ­ 1 +1 ­ 1 +1 ­ 1 +1 ­ 1 +1 ­ 1 +1 ­ 1 +1 ­ 1 +1 ­ 1 1 1 0 0 0 1 0 1 1 0 1 1 1 1 0 0 Ordinary multiplier 0 ­1 0 0 + 1 ­ 1 +1 0 ­ 1 +1 0 0 0 ­1 0 0 0 0 0 0 1 1 1 1 1 0 0 0 0 1 1 1 Good multiplier 0 0 0 +1 0 0 0 0 ­1 0 0 0 +1 0 0 ­1
Booth Algorithm Flow Chart Table
Unit 2 arithmetics
Bit-Pair Recoding of Multipliers  Bit-pair recoding halves the maximum number of summands (versions of the multiplicand). Sign extension Implied 0 to right of LSB 1 1 1 0 1 0 0 0 0 − 1 +1 − 1 0 0 −1 −2 (a)  Example of bit­pair recoding derived from Booth recoding
Bit-Pair Recoding of Multipliers Multiplier bit­pair Multiplier bit on the right Multiplicand selected at position i i+1 i i−1 0 0 0 0 X  M 0 0 1 +1 X  M 0 1 0 +1 X  M 0 1 1 +2 X  M 1 0 0 −2 X  M 1 0 1 −1 X  M 1 1 0 −1 X  M 1 1 1 0 X  M (b) Table of multiplicand selection decisions
Bit-Pair Recoding of Multipliers 0 1 1 0 1 0 ­ 1 +1 ­ 1 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 0 0 1 1 0 0 0 0 1 1 0 1 1 1 1 0 0 1 1 0 1 1 0 1 ( + 13) 0 0 0 0 0 0 × 1 1 0 1 0 (­ 6 ) 1 1 1 0 1 1 0 0 1 0 ( ­ 78 ) 0 1 1 0 1 0 ­1 ­2 1 1 1 1 1 0 0 1 1 0 1 1 1 1 0 0 1 1 0 0 0 0 0 0 1 1 1 0 1 1 0 0 1 0 Figure 6.15.  Multiplication requiring only n/2 summands.41
Carry-Save Addition of Summands  CSA speeds up the addition process. P7 P6 P5 P4 P3 P2 P1 P0 42
Carry-Save Addition of Summands(Cont.,) P7 P6 P5 P4 P3 P2 P1 P0
Carry-Save Addition of Summands(Cont.,)  Consider the addition of many summands, we can:  Group the summands in threes and perform carry-save addition on each of these groups in parallel to generate a set of S and C vectors in one full-adder delay  Group all of the S and C vectors into threes, and perform carry-save addition on them, generating a further set of S and C vectors in one more full-adder delay  Continue with this process until there are only two vectors remaining  They can be added in a RCA or CLA to produce the desired product
Carry-Save Addition of Summands 1 0 1 1 0 1 (45) M X 1 1 1 1 1 1 (63) Q 1 0 1 1 0 1 A 1 0 1 1 0 1 B 1 0 1 1 0 1 C 1 0 1 1 0 1 D 1 0 1 1 0 1 E 1 0 1 1 0 1 F 1 0 1 1 0 0 0 1 0 0 1 1 (2,835) Product Figure 6.17.  A multiplication example used to illustrate carry­save addition as shown in Figure 6.18.
1 0 1 1 0 1 M x 1 1 1 1 1 1 Q 1 0 1 1 0 1 A 1 0 1 1 0 1 B 1 0 1 1 0 1 C 1 1 0 0 0 0 1 1 S 1 0 0 1 1 1 1 0 0 C 1 1 0 1 1 0 1 D 1 0 1 1 0 1 E 1 0 1 1 0 1 F 1 1 0 0 0 0 1 1 S 2 0 0 1 1 1 1 0 0 C 2 1 1 0 0 0 0 1 1 S1 0 0 1 1 1 1 0 0 C 1 1 1 0 0 0 0 1 1 S2 1 1 0 1 0 1 0 0 0 1 1 S 3 0 0 0 0 1 0 1 1 0 0 0 C3 0 0 1 1 1 1 0 0 C2 0 1 0 1 1 1 0 1 0 0 1 1 S4 + 0 1 0 1 0 1 0 0 0 0 0 C 4 1 0 1 1 0 0 0 1 0 0 1 1 Product Figure 6.18. The multiplication example from Figure 6.17 performed using carry­save addition.
Unit 2 arithmetics
Manual Division 21 10101 13 274   1101 100010010 26 1101 14 10000 13 1101 1 1110 1101 1  Longhand division examples.
Longhand Division Steps  Position the divisor appropriately with respect to the dividend and performs a subtraction.  If the remainder is zero or positive, a quotient bit of 1 is determined, the remainder is extended by another bit of the dividend, the divisor is repositioned, and another subtraction is performed.  If the remainder is negative, a quotient bit of 0 is determined, the dividend is restored by adding back the divisor, and the divisor is repositioned for another subtraction.
Circuit Arrangement Shift left an an­1 a0 qn­1 q0 Dividend Q A Quotient Setting N+1 bit  Add/Subtract adder Control Sequencer 0 mn­1 m0 Divisor M Figure 6.21. Circuit arrangement for binary division.
Binary Division Hardware Implementation
Restoring Division  Shift A and Q left one binary position  Subtract M from A, and place the answer back in A  If the sign of A is 1, set q0 to 0 and add M back to A (restore A); otherwise, set q0 to 1  Repeat these steps n times
Examples Initially 0 Shift 0 0 0 0 0 0 0 0 0 1 0 0 1 1 1 0 0 0 0 0 0 Subtract 1 1 1 0 1 First cycle Set q0 1 1 1 1 0 Restore 1 1 0 0 0 0 1 0 0 0 0 1 0 Shift 0 0 0 1 0 0 0 0 1 1 1 0 0 0 Subtract 1 1 1 0 1 1 1 Set q0 1 1 1 1 1 Second cycle Restore 1 1 1 0 0 0 0 1 0 0 0 0 0 Shift 0 0 1 0 0 0 0 0 Subtract 1 1 1 0 1 Set q0 0 0 0 0 1 Third cycle Shift 0 0 0 1 0 0 0 0 1 Subtract 1 1 1 0 1 0 0 1 Set q0 1 1 1 1 1 Fourth cycle Restore 1 1 0 0 0 1 0 0 0 1 0 Remainder Quotient Figure 6.22. A restoring­division example.
Restoring Division Flow Chart
Nonrestoring Division  Avoid the need for restoring A after an unsuccessful subtraction.  Any idea?  Step 1: (Repeat n times)  If the sign of A is 0, shift A and Q left one bit position and subtract M from A; otherwise, shift A and Q left and add M to A.  Now, if the sign of A is 0, set q0 to 1; otherwise, set q0 to 0.  Step2: If the sign of A is 1, add M to A
Examples Initially Shift 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 1 0 0 0 0 0 0 First cycle Subtract 1 1 1 0 1 Set q 0 1 1 1 1 0 0 0 0 0 Shift 1 1 1 0 0 0 0 0 Add 0 0 0 1 1 Second cycle Set q 1 1 1 1 1 0 0 0 0 0 Shift 1 1 1 1 0 0 0 0 1 1 1 1 1 Add 0 0 0 1 1 Third cycle Restore 0 0 0 1 1 Set q 0 0 0 0 1 0 0 0 1  remainder 0 Add 0 0 0 1 0 Remainder Shift 0 0 0 1 0 0 0 1 Subtract 1 1 1 0 1 Fourth cycle Set q 1 1 1 1 1 0 0 1 0 0 Quotient  A nonrestoring­division example.
Nonrestoring Division Flow Chart
Unit 2 arithmetics
If b is a binary vector, then we have seen that it can be interpreted as an unsigned integer by: V(b) = b31.231 + b30.230 + bn-3.229 + .... + b1.21 + b0.20 This vector has an implicit binary point to its immediate right: b31b30b29....................b1b0. implicit binary point Suppose if the binary vector is interpreted with the implicit binary point is just left of the sign bit: implicit binary point .b31b30b29....................b1b0 The value of b is then given by: V(b) = b31.2-1 + b30.2-2 + b29.2-3 + .... + b1.2-31 + b0.2-32
The value of the unsigned binary fraction is: V(b) = b31.2-1 + b30.2-2 + b29.2-3 + .... + b1.2-31 + b0.2-32 The range of the numbers represented in this format is: 0 ≤ V (b) ≤ 1 − 2 −32 ≈ 0.9999999998 In general for a n-bit binary fraction (a number with an assumed binary point at the immediate left of the vector), then the range of values is: 0 ≤ V (b) ≤ 1 − 2 − n
•Previous representations have a fixed point. Either the point is to the immediate right or it is to the immediate left. This is called Fixed point representation. •Fixed point representation suffers from a drawback that the representation can only represent a finite range (and quite small) range of numbers. A more convenient representation is the scientific representation, where the numbers are represented in the form: x = m1 .m2 m3 m4 × b ±e Components of these numbers are: Mantissa (m), implied base (b), and exponent (e)
A number such as the following is said to have 7 significant digits x = ±0.m1m2m3m4m5m6m7 × b ±e Fractions in the range 0.0 to 0.9999999 need about 24 bits of precision (in binary). For example the binary fraction with 24 1’s: 111111111111111111111111 = 0.9999999404 Not every real number between 0 and 0.9999999404 can be represented by a 24-bit fractional number. The smallest non-zero number that can be represented is: 000000000000000000000001 = 5.96046 x 10-8 Every other non-zero number is constructed in increments of this value.
•In a 32-bit number, suppose we allocate 24 bits to represent a fractional mantissa. •Assume that the mantissa is represented in sign and magnitude format, and we have allocated one bit to represent the sign. •We allocate 7 bits to represent the exponent, and assume that the exponent is represented as a 2’s complement integer. •There are no bits allocated to represent the base, we assume that the base is implied for now, that is the base is 2. •Since a 7-bit 2’s complement number can represent values in the range -64 to 63, the range of numbers that can be represented is: 0.0000001 x 2-64 < = | x | <= 0.9999999 x 263 •In decimal representation this range is: 0.5421 x 10-20 < = | x | <= 9.2237 x 1018
1 7 24 Sign Exponent Fractional mantissa bit •24-bit mantissa with an implied binary point to the immediate left •7-bit exponent in 2’s complement form, and implied base is 2.
Consider the number: x = 0.0004056781 x 1012 If the number is to be represented using only 7 significant mantissa digits, the representation ignoring rounding is: x = 0.0004056 x 1012 If the number is shifted so that as many significant digits are brought into 7 available slots: x = 0.4056781 x 109 = 0.0004056 x 1012 Exponent of x was decreased by 1 for every left shift of x. A number which is brought into a form so that all of the available mantissa digits are optimally used (this is different from all occupied which may not hold), is called a normalized number. Same methodology holds in the case of binary mantissas 0001101000(10110) x 28 = 1101000101(10) x 25
•A floating point number is in normalized form if the most significant 1 in the mantissa is in the most significant bit of the mantissa. •All normalized floating point numbers in this system will be of the form: 0.1xxxxx.......xx Range of numbers representable in this system, if every number must be normalized is: 0.5 x 2-64 <= | x | < 1 x 263
The procedure for normalizing a floating point number is: Do (until MSB of mantissa = = 1) Shift the mantissa left (or right) Decrement (increment) the exponent by 1 end do Applying the normalization procedure to: .000111001110....0010 x 2-62 gives: .111001110........ x 2-65 But we cannot represent an exponent of –65, in trying to normalize the number we have underflowed our representation. Applying the normalization procedure to: 1.00111000............x 263 gives: 0.100111..............x 264 This overflows the representation.
So far we have assumed an implied base of 2, that is our floating point numbers are of the form: x = m 2e If we choose an implied base of 16, then: x = m 16e Then: y = (m.16) .16e-1 (m.24) .16e-1 = m . 16e = x •Thus, every four left shifts of a binary mantissa results in a decrease of 1 in a base 16 exponent. •Normalization in this case means shifting the mantissa until there is a 1 in the first four bits of the mantissa.
•Rather than representing an exponent in 2’s complement form, it turns out to be more beneficial to represent the exponent in excess notation. •If 7 bits are allocated to the exponent, exponents can be represented in the range of -64 to +63, that is: -64 <= e <= 63 Exponent can also be represented using the following coding called as excess-64: E’ = Etrue + 64 In general, excess-p coding is represented as: E’ = Etrue + p True exponent of -64 is represented as 0 0 is represented as 64 63 is represented as 127 This enables efficient comparison of the relative sizes of two floating point numbers.
IEEE Floating Point notation is the standard representation in use. There are two representations: - Single precision. - Double precision. Both have an implied base of 2. Single precision: - 32 bits (23-bit mantissa, 8-bit exponent in excess-127 representation) Double precision: - 64 bits (52-bit mantissa, 11-bit exponent in excess-1023 representation) Fractional mantissa, with an implied binary point at immediate left. Sign Exponent Mantissa 1 8 or 11 23 or 52
•Floating point numbers have to be represented in a normalized form to maximize the use of available mantissa digits. •In a base-2 representation, this implies that the MSB of the mantissa is always equal to 1. •If every number is normalized, then the MSB of the mantissa is always 1. We can do away without storing the MSB. •IEEE notation assumes that all numbers are normalized so that the MSB of the mantissa is a 1 and does not store this bit. •So the real MSB of a number in the IEEE notation is either a 0 or a 1. •The values of the numbers represented in the IEEE single precision notation are of the form: (+,-) 1.M x 2(E - 127) •The hidden 1 forms the integer part of the mantissa. •Note that excess-127 and excess-1023 (not excess-128 or excess-1024) are used to represent the exponent.
In the IEEE representation, the exponent is in excess-127 (excess-1023) notation. The actual exponents represented are: -126 <= E <= 127 and -1022 <= E <= 1023 not -127 <= E <= 128 and -1023 <= E <= 1024 This is because the IEEE uses the exponents -127 and 128 (and -1023 and 1024), that is the actual values 0 and 255 to represent special conditions: - Exact zero - Infinity
Addition: 3.1415 x 108 + 1.19 x 106 = 3.1415 x 108 + 0.0119 x 108 = 3.1534 x 108 Multiplication: 3.1415 x 108 x 1.19 x 106 = (3.1415 x 1.19 ) x 10(8+6) Division: 3.1415 x 108 / 1.19 x 106 = (3.1415 / 1.19 ) x 10(8-6) Biased exponent problem: If a true exponent e is represented in excess-p notation, that is as e+p. Then consider what happens under multiplication: a. 10(x + p) * b. 10(y + p) = (a.b). 10(x + p + y +p) = (a.b). 10(x +y + 2p) Representing the result in excess-p notation implies that the exponent should be x+y+p. Instead it is x+y+2p. Biases should be handled in floating point arithmetic.
Floating point arithmetic: ADD/SUB rule  Choose the number with the smaller exponent.  Shift its mantissa right until the exponents of both the numbers are equal.  Add or subtract the mantissas.  Determine the sign of the result.  Normalize the result if necessary and truncate/round to the number of mantissa bits. Note: This does not consider the possibility of overflow/underflow.
Floating point arithmetic: MUL rule  Add the exponents.  Subtract the bias.  Multiply the mantissas and determine the sign of the result.  Normalize the result (if necessary).  Truncate/round the mantissa of the result.
Floating point arithmetic: DIV rule  Subtract the exponents  Add the bias.  Divide the mantissas and determine the sign of the result.  Normalize the result if necessary.  Truncate/round the mantissa of the result. Note: Multiplication and division does not require alignment of the mantissas the way addition and subtraction does.
While adding two floating point numbers with 24-bit mantissas, we shift the mantissa of the number with the smaller exponent to the right until the two exponents are equalized. This implies that mantissa bits may be lost during the right shift (that is, bits of precision may be shifted out of the mantissa being shifted). To prevent this, floating point operations are implemented by keeping guard bits, that is, extra bits of precision at the least significant end of the mantissa. The arithmetic on the mantissas is performed with these extra bits of precision. After an arithmetic operation, the guarded mantissas are: - Normalized (if necessary) - Converted back by a process called truncation/rounding to a 24-bit mantissa.
Truncation/rounding  Straight chopping:  The guard bits (excess bits of precision) are dropped.  Von Neumann rounding:  If the guard bits are all 0, they are dropped.  However, if any bit of the guard bit is a 1, then the LSB of the retained bit is set to 1.  Rounding:  If there is a 1 in the MSB of the guard bit then a 1 is added to the LSB of the retained bits.
Rounding  Rounding is evidently the most accurate truncation method.  However,  Rounding requires an addition operation.  Rounding may require a renormalization, if the addition operation de- normalizes the truncated number. 0.111111100000 rounds to 0.111111 + 0.000001 =1.000000 which must be renormalized to 0.100000  IEEE uses the rounding method.

More Related Content

PDF
Msp430 assembly language instructions &amp;addressing modes
Harsha herle
 
PPT
Parallel processing and pipelining
mahesh kumar prajapat
 
PPT
8086 microprocessor assembler directives.ppt
NaveenKumar5162
 
PPT
8259 a
Meghaditya Roy Chaudhury
 
PDF
Magnitude Comparator and types of MC
Easy n Inspire L
 
PPTX
DMA operation
Imran Khan
 
PPT
Cache coherence
Employee
 
DOCX
16 x 16 banyan switch
Ashu0711
 
Msp430 assembly language instructions &amp;addressing modes
Harsha herle
 
Parallel processing and pipelining
mahesh kumar prajapat
 
8086 microprocessor assembler directives.ppt
NaveenKumar5162
 
8259 a
Meghaditya Roy Chaudhury
 
Magnitude Comparator and types of MC
Easy n Inspire L
 
DMA operation
Imran Khan
 
Cache coherence
Employee
 
16 x 16 banyan switch
Ashu0711
 

What's hot (20)

PDF
ARM Architecture
Dwight Sabio
 
PDF
Project Report Of SRAM Design
Aalay Kapadia
 
PPTX
8086 Microprocessor Pipeline Architecture.pptx
Green University of Bangladesh
 
PPTX
Direct Memory Access(DMA)
Page Maker
 
PPTX
Stack in microprocessor 8085(presantation)
Safin Biswas
 
PPTX
TMSLF2407 DSP Controller
ANIRUDDHMAINI1
 
PPT
Micro programmed control
Syed Zaid Irshad
 
PPTX
Asynchronous Data Transfer.pptx
ArunaDevi63
 
PPTX
Carry look ahead adder
dragonpradeep
 
PPTX
Presentation1
Haya Saani
 
PPTX
SRAM
rohitladdu
 
PPT
carry look ahead adder
ASHISH MANI
 
PPTX
Linux Device Tree
艾鍗科技
 
PPTX
ARITHMETIC OPERATIONS IN 8085 MICROPROCESSOR
RamaPrabha24
 
PDF
Computer organization memory
Deepak John
 
PPS
Synchronous and-asynchronous-data-transfer
Anuj Modi
 
PDF
Microprocessor lab
kpaulraj
 
PPTX
IO Techniques in Computer Organization
Om Prakash
 
DOC
Chap 5
Tanuj Patel
 
PDF
8051 (microcontroller)class1
Nitin Ahire
 
ARM Architecture
Dwight Sabio
 
Project Report Of SRAM Design
Aalay Kapadia
 
8086 Microprocessor Pipeline Architecture.pptx
Green University of Bangladesh
 
Direct Memory Access(DMA)
Page Maker
 
Stack in microprocessor 8085(presantation)
Safin Biswas
 
TMSLF2407 DSP Controller
ANIRUDDHMAINI1
 
Micro programmed control
Syed Zaid Irshad
 
Asynchronous Data Transfer.pptx
ArunaDevi63
 
Carry look ahead adder
dragonpradeep
 
Presentation1
Haya Saani
 
SRAM
rohitladdu
 
carry look ahead adder
ASHISH MANI
 
Linux Device Tree
艾鍗科技
 
ARITHMETIC OPERATIONS IN 8085 MICROPROCESSOR
RamaPrabha24
 
Computer organization memory
Deepak John
 
Synchronous and-asynchronous-data-transfer
Anuj Modi
 
Microprocessor lab
kpaulraj
 
IO Techniques in Computer Organization
Om Prakash
 
Chap 5
Tanuj Patel
 
8051 (microcontroller)class1
Nitin Ahire
 
Ad

Similar to Unit 2 arithmetics (20)

PPT
Half adder and full adder
Kamil Hussain
 
PPT
Parallel Prefix Adders Presentation
Peeyush Pashine
 
PDF
Number system
mabroukamohammed
 
PPT
Ch1 2
Teksify
 
PPT
Dateof birthdesignproblem 1_
rofldork
 
PDF
Peta karnaugh
ti_stmik_nh_jambi
 
PPTX
Computer arthtmetic,,,
Ahsan Mehmood
 
PPTX
Computer arthtmetic,,,
Ahsan Mehmood
 
PPT
Digital 1
AJAL A J
 
PPT
comp.org Chapter 2
Rajat Sharma
 
PDF
Sienna 11 graphs
chidabdu
 
PPTX
Logicgates
Tarun Gehlot
 
PDF
Digital notes
Nosheb Danish
 
PDF
C02
tjohanna
 
PPT
Booth Multiplier
Sudhir Kumar
 
PDF
MCS 012 computer organisation and assembly language programming assignment…
Dr. Loganathan R
 
PDF
Csr2011 june14 17_00_pospelov
CSR2011
 
PDF
Binary addition
Martin Jacob
 
PDF
Computer science study material
Swarup Kumar Boro
 
PDF
16%20 lecture
Arif Siyal
 
Half adder and full adder
Kamil Hussain
 
Parallel Prefix Adders Presentation
Peeyush Pashine
 
Number system
mabroukamohammed
 
Ch1 2
Teksify
 
Dateof birthdesignproblem 1_
rofldork
 
Peta karnaugh
ti_stmik_nh_jambi
 
Computer arthtmetic,,,
Ahsan Mehmood
 
Computer arthtmetic,,,
Ahsan Mehmood
 
Digital 1
AJAL A J
 
comp.org Chapter 2
Rajat Sharma
 
Sienna 11 graphs
chidabdu
 
Logicgates
Tarun Gehlot
 
Digital notes
Nosheb Danish
 
C02
tjohanna
 
Booth Multiplier
Sudhir Kumar
 
MCS 012 computer organisation and assembly language programming assignment…
Dr. Loganathan R
 
Csr2011 june14 17_00_pospelov
CSR2011
 
Binary addition
Martin Jacob
 
Computer science study material
Swarup Kumar Boro
 
16%20 lecture
Arif Siyal
 
Ad

More from chidabdu (20)

PDF
Sienna 12 huffman
chidabdu
 
PDF
Sienna 10 dynamic
chidabdu
 
PDF
Sienna 9 hashing
chidabdu
 
PDF
Sienna 8 countingsorts
chidabdu
 
PDF
Sienna 7 heaps
chidabdu
 
PDF
Sienna 6 bst
chidabdu
 
PDF
Sienna 5 decreaseandconquer
chidabdu
 
PDF
Sienna 4 divideandconquer
chidabdu
 
PDF
Sienna 3 bruteforce
chidabdu
 
PDF
Sienna 2 analysis
chidabdu
 
PDF
Sienna 1 intro
chidabdu
 
PDF
Sienna 13 limitations
chidabdu
 
PPT
Unit 3 basic processing unit
chidabdu
 
PPT
Unit 5 I/O organization
chidabdu
 
PDF
Algorithm chapter 1
chidabdu
 
PDF
Algorithm chapter 11
chidabdu
 
PDF
Algorithm chapter 10
chidabdu
 
PDF
Algorithm chapter 9
chidabdu
 
PDF
Algorithm chapter 8
chidabdu
 
PDF
Algorithm chapter 7
chidabdu
 
Sienna 12 huffman
chidabdu
 
Sienna 10 dynamic
chidabdu
 
Sienna 9 hashing
chidabdu
 
Sienna 8 countingsorts
chidabdu
 
Sienna 7 heaps
chidabdu
 
Sienna 6 bst
chidabdu
 
Sienna 5 decreaseandconquer
chidabdu
 
Sienna 4 divideandconquer
chidabdu
 
Sienna 3 bruteforce
chidabdu
 
Sienna 2 analysis
chidabdu
 
Sienna 1 intro
chidabdu
 
Sienna 13 limitations
chidabdu
 
Unit 3 basic processing unit
chidabdu
 
Unit 5 I/O organization
chidabdu
 
Algorithm chapter 1
chidabdu
 
Algorithm chapter 11
chidabdu
 
Algorithm chapter 10
chidabdu
 
Algorithm chapter 9
chidabdu
 
Algorithm chapter 8
chidabdu
 
Algorithm chapter 7
chidabdu
 

Recently uploaded (20)

PDF
ANTIBIOTICS.pptx.pdf………………… xxxxxxxxxxxxx
HakHe
 
PDF
Abdominal Access Techniques with Prof. Dr. R K Mishra
mohitsuren827
 
PDF
VCE English Exam - Section C Student Revision Booklet
jpinnuck
 
PDF
Computing-Curriculum for Schools in Ghana
abigailanane203
 
PPTX
PPH.pptx obstetrics and gynecology in nursing
SrideviDevaraj5
 
PDF
Module 4: Burden of Disease Tutorial Slides S2 2025
Jonathan Hallett
 
PDF
3rd Neelam Sanjeevareddy Memorial Lecture.pdf
Academy of Grassroots Studies and Research of India (AGRASRI)
 
PDF
2.FourierTransform-ShortQuestionswithAnswers.pdf
Raji Murugan
 
PPTX
Institutional Correction lecture only . . .
limvtheghins
 
PDF
RMMM.pdf make it easy to upload and study
sajedul2
 
PDF
Insiders guide to clinical Medicine.pdf
AbhishekPatil315128
 
PPTX
Cell Types and Its function , kingdom of life
ClaireMangundayao1
 
PDF
FourierSeries-QuestionsWithAnswers(Part-A).pdf
Raji Murugan
 
PPTX
Pharmacology of Heart Failure /Pharmacotherapy of CHF
Rajshri Ghogare
 
PPTX
Pharma ospi slides which help in ospi learning
rameetkumar304
 
PPTX
IMMUNITY IMMUNITY refers to protection against infection, and the immune syst...
shilpa939953
 
PDF
Anesthesia in Laparoscopic Surgery in India
mohitsuren827
 
PDF
O7-L3 Supply Chain Operations - ICLT Program
labyankof
 
PDF
O5-L3 Freight Transport Ops (International) V1.pdf
labyankof
 
PPTX
human mycosis Human fungal infections are called human mycosis..pptx
shilpa939953
 
ANTIBIOTICS.pptx.pdf………………… xxxxxxxxxxxxx
HakHe
 
Abdominal Access Techniques with Prof. Dr. R K Mishra
mohitsuren827
 
VCE English Exam - Section C Student Revision Booklet
jpinnuck
 
Computing-Curriculum for Schools in Ghana
abigailanane203
 
PPH.pptx obstetrics and gynecology in nursing
SrideviDevaraj5
 
Module 4: Burden of Disease Tutorial Slides S2 2025
Jonathan Hallett
 
3rd Neelam Sanjeevareddy Memorial Lecture.pdf
Academy of Grassroots Studies and Research of India (AGRASRI)
 
2.FourierTransform-ShortQuestionswithAnswers.pdf
Raji Murugan
 
Institutional Correction lecture only . . .
limvtheghins
 
RMMM.pdf make it easy to upload and study
sajedul2
 
Insiders guide to clinical Medicine.pdf
AbhishekPatil315128
 
Cell Types and Its function , kingdom of life
ClaireMangundayao1
 
FourierSeries-QuestionsWithAnswers(Part-A).pdf
Raji Murugan
 
Pharmacology of Heart Failure /Pharmacotherapy of CHF
Rajshri Ghogare
 
Pharma ospi slides which help in ospi learning
rameetkumar304
 
IMMUNITY IMMUNITY refers to protection against infection, and the immune syst...
shilpa939953
 
Anesthesia in Laparoscopic Surgery in India
mohitsuren827
 
O7-L3 Supply Chain Operations - ICLT Program
labyankof
 
O5-L3 Freight Transport Ops (International) V1.pdf
labyankof
 
human mycosis Human fungal infections are called human mycosis..pptx
shilpa939953
 

Unit 2 arithmetics

  • 2. xi yi Carry­in ci Sum s i Carry­out c i+1 At the ith stage: 0 0 0 0 0 Input: 0 0 1 1 0 ci is the carry-in 0 1 0 1 0 0 1 1 0 1 Output: 1 0 0 1 0 si is the sum 1 0 1 0 1 1 1 0 0 1 ci+1 carry-out to (i+1)st 1 1 1 1 1 state si   = xi yi ci + xi yi ci + xi yi ci + xi yi ci = x i ⊕ yi ⊕ ci ci +1 = yi c i + x i ci + x i y i Example: X 7 0 1 1 1 Carry­out xi Carry­in + Y = + 6 = +00 1 1 1 1 0 0 0 yi ci+1 ci Z 13 1 1 0 1 si Legend for stage i
  • 3. Sum Carry yi c i xi xi yi si c c i +1 i ci x xi yi i yi ci + 1 Full adder ci (FA) s i Full Adder (FA): Symbol for the complete circuit for a single stage of addition.
  • 4. •Cascade n full adder (FA) blocks to form a n-bit adder. •Carries propagate or ripple through this cascade, n-bit ripple carry adder. xn ­ 1 yn­ 1 x1 y1 x0 y0 cn ­ 1 c1 cn FA FA FA c0 sn ­ 1 s1 s0 Most significant bit Least significant bit (MSB) position (LSB) position Carry-in c0 into the LSB position provides a convenient way to perform subtraction.
  • 5. K n-bit numbers can be added by cascading k n-bit adders. xk n ­ 1 yk n ­ 1 x2n ­ 1 y2n ­ 1 xn y n xn ­ 1 y n ­ 1 x 0 y 0 cn n­bit n­bit n­bit c c kn 0 adder adder adder s s( s s s s kn ­ 1 k ­ 1) n 2n ­ 1 n n­ 1 0 Each n-bit adder forms a block, so this is cascading of blocks. Carries ripple or propagate through blocks, Blocked Ripple Carry Adder
  • 6. •Recall X – Y is equivalent to adding 2’s complement of Y to X. •2’s complement is equivalent to 1’s complement + 1. •X – Y = X + Y + 1 •2’s complement of positive and negative numbers is computed similarly. xn ­ 1 yn­ 1 x1 y1 x0 y0 cn ­ 1 c1 cn FA FA FA 1 sn ­ 1 s1 s0 Most significant bit Least significant bit (MSB) position (LSB) position
  • 7. y y y n­ 1 1 0 Add/Sub control x x x n­ 1 1 0 c n­bit adder n c 0 s s s n­ 1 1 0 •Add/sub control = 0, addition. •Add/sub control = 1, subtraction.
  • 8. Detecting overflows  Overflows can only occur when the sign of the two operands is the same.  Overflow occurs if the sign of the result is different from the sign of the operands.  Recall that the MSB represents the sign.  xn-1, yn-1, sn-1 represent the sign of operand x, operand y and result s respectively.  Circuit to detect overflow can be implemented by the following logic expressions: Overflow = xn −1 yn −1sn −1 + xn −1 yn −1sn −1 Overflow = cn ⊕ cn −1
  • 9. x0 y0 Consider 0th stage: •c1 is available after 2 gate delays. •s1 is available after 1 gate delay. c1 FA c0 s0 Sum Carry yi c i xi xi yi si c c i +1 i ci x i yi
  • 10. Cascade of 4 Full Adders, or a 4-bit adder x0 y0 x0 y0 x0 y0 x0 y0 FA FA FA FA c0 c4 c3 c2 c1 s3 s2 s1 s0 •s0 available after 1 gate delays, c1 available after 2 gate delays. •s1 available after 3 gate delays, c2 available after 4 gate delays. •s2 available after 5 gate delays, c3 available after 6 gate delays. •s3 available after 7 gate delays, c4 available after 8 gate delays. For an n-bit adder, sn-1 is available after 2n-1 gate delays cn is available after 2n gate delays.
  • 11. Recall the equations: si = xi ⊕ yi ⊕ ci ci +1 = xi yi + xi ci + yi ci Second equation can be written as: ci +1 = xi yi + ( xi + yi )ci We can write: ci +1 = Gi + Pi ci where Gi = xi yi and Pi = xi + yi •Gi is called generate function and Pi is called propagate function •Gi and Pi are computed only from xi and yi and not ci, thus they can be computed in one gate delay after X and Y are applied to the inputs of an n-bit adder.
  • 12. ci +1 = Gi + Pi ci ci = Gi −1 + Pi −1ci −1 ⇒ ci+1 = Gi + Pi (Gi −1 + Pi −1ci −1 ) continuing ⇒ ci+1 = Gi + Pi (Gi −1 + Pi −1 (Gi − 2 + Pi− 2 ci −2 )) until ci+1 = Gi + PiGi −1 + Pi Pi−1 Gi −2 + .. + Pi Pi −1 ..P1G0 + Pi Pi −1 ...P0 c 0 •All carries can be obtained 3 gate delays after X, Y and c0 are applied. -One gate delay for Pi and Gi -Two gate delays in the AND-OR circuit for ci+1 •All sums can be obtained 1 gate delay after the carries are computed. •Independent of n, n-bit addition requires only 4 gate delays. •This is called Carry Lookahead adder.
  • 13. x y x y x y x y 3 3 2 2 1 1 0 0 c4 c 3 c 2 c 1 . c 4-bit carry-lookahead B cell B cell B cell B cell 0 adder s s s s 3 2 1 0 G3 P3 G2 P2 G P G P 1 1 0 0 Carry­lookahead logic xi yi . . . c i B-cell for a single stage B cell Gi P i si
  • 14. Carry lookahead adder (contd..)  Performing n-bit addition in 4 gate delays independent of n is good only theoretically because of fan-in constraints. ci+1 = Gi + PiGi −1 + Pi Pi−1 Gi −2 + .. + Pi Pi −1 ..P1G0 + P Pi −1 ...P0 c0 i  Last AND gate and OR gate require a fan-in of (n+1) for a n- bit adder.  For a 4-bit adder (n=4) fan-in of 5 is required.  Practical limit for most gates.  In order to add operands longer than 4 bits, we can cascade 4-bit Carry-Lookahead adders. Cascade of Carry-Lookahead adders is called Blocked Carry-Lookahead adder.
  • 16. Carry-out from a 4-bit block can be given as: c4 = G3 + P G2 + P3 P2 G1 + P3 P2 P G0 + P3 P2 P1P0 c0 3 1 Rewrite this as: P0I = P3 P2 P1 P0 G0I = G3 + P3 G2 + P3 P2 G1 + P3 P2 P1G0 Subscript I denotes the blocked carry lookahead and identifies the block. Cascade 4 4-bit adders, c16 can be expressed as: c16 = G3 + P3I G2I + P3I P2I G1I + P3I P2I P10 G0 + P3I P2I P10 P00 c0 I I
  • 17. x15­12 y15­12 x11­8 y11­8 x7­4 y7­4 x3­0 y3­0 c16 4­bit adder c12 4­bit adder c8 4­bit adder c4 4­bit adder . c0 s15­12 s11­8 s7­4 s3­0 G3 I P3 I G2 I P 2I G 1I P 1I G 0I P0 I Carry­lookahead logic After xi, yi and c0 are applied as inputs: - Gi and Pi for each stage are available after 1 gate delay. - PI is available after 2 and GI after 3 gate delays. - All carries are available after 5 gate delays. - c16 is available after 5 gate delays. - s15 which depends on c12 is available after 8 (5+3)gate delays (Recall that for a 4-bit carry lookahead adder, the last sum bit is available 3 gate delays after all inputs are available)
  • 19. Product of 2 n-bit numbers is at most a 2n-bit number. Unsigned multiplication can be viewed as addition of shifted versions of the multiplicand.
  • 20. Multiplication of unsigned numbers (contd..)  We added the partial products at end.  Alternative would be to add the partial products at each stage.  Rules to implement multiplication are:  If the ith bit of the multiplier is 1, shift the multiplicand and add the shifted multiplicand to the current value of the partial product.  Hand over the partial product to the next stage  Value of the partial product at the start stage is 0.
  • 21. Typical multiplication cell Bit of incoming partial product (PPi) jth multiplicand bit ith multiplier bit ith multiplier bit carry out FA carry in Bit of outgoing partial product (PP(i+1))
  • 22. Combinatorial array multiplier Multiplicand 0 m3 0 m2 0 m1 0 m0 (PP0) q0 0 PP1 p0 q1 0 r lie PP2 tip p1 ul q2 M 0 PP3 p2 q3 0 , p7 p6 p5 p4 p3 Product is: p7,p6,..p0 Multiplicand is shifted by displacing it through an array of adders.
  • 23. Combinatorial array multiplier (contd..)  Combinatorial array multipliers are:  Extremely inefficient.  Have a high gate count for multiplying numbers of practical size such as 32- bit or 64-bit numbers.  Perform only one function, namely, unsigned integer product.  Improve gate efficiency by using a mixture of combinatorial array techniques and sequential techniques requiring less combinational logic.
  • 24. Sequential multiplication  Recall the rule for generating partial products:  If the ith bit of the multiplier is 1, add the appropriately shifted multiplicand to the current partial product.  Multiplicand has been shifted left when added to the partial product.  However, adding a left-shifted multiplicand to an unshifted partial product is equivalent to adding an unshifted multiplicand to a right-shifted partial product.
  • 25. Register A  (initially 0) Shift right C a a q q n ­ 1 0 n­ 1 0 Multiplier Q Add/Noadd control n-bit Adder MUX Control sequencer 0 0 m m0 n ­ 1 Multiplicand M
  • 26. M 1   1   0   1 Initial configuration 0 0   0   0   0 1   0   1   1 C A Q 0 1   1   0   1 1   0   1   1 Add Shift First cycle 0 0   1   1   0 1   1   0   1 1 0   0   1   1 1   1   0   1 Add Shift Second cycle 0 1   0   0   1 1   1   1   0 0 1   0   0   1 1   1   1   0 No add Shift Third cycle 0 0   1   0   0 1   1   1   1 1 0   0   0   1 1   1   1   1 Add Shift Fourth cycle 0 1   0   0   0 1   1   1   1 Product
  • 28. Signed Multiplication  Considering 2’s-complement signed operands, what will happen to (-13)×(+11) if following the same method of unsigned multiplication? 1 0 0 1 1 ( ­ 13) 0 1 0 1 1 ( + 11) 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1 0 0 1 1 Sign extension is shown in blue 0 0 0 0 0 0 0 0 1 1 1 0 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 0 0 1 ( ­ 143) Sign extension of negative multiplicand.
  • 29. Signed Multiplication  For a negative multiplier, a straightforward solution is to form the 2’s-complement of both the multiplier and the multiplicand and proceed as in the case of a positive multiplier.  This is possible because complementation of both operands does not change the value or the sign of the product.  A technique that works equally well for both negative and positive multipliers – Booth algorithm.
  • 31. Booth Algorithm Consider in a multiplication, the multiplier is positive 0011110, how many appropriately shifted versions of the multiplicand are added in a standard procedure? 0 1 0 1 1 0 1 0 0 +1 +1 + 1+1 0 0 0 0 0 0 0 0 0 1 0 1 1 0 1 0 1 0 1 1 0 1 0 1 0 1 1 0 1 0 1 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 1 0 0 0 1 1 0
  • 32. Booth Algorithm  Since 0011110 = 0100000 – 0000010, if we use the expression to the right, what will happen? 0 1 0 1 1 0 1 0 +1 0 0 0 ­ 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2's complement of 1 1 1 1 1 1 1 0 1 0 0 1 1 the multiplicand 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 1 0 0 0 1 1 0
  • 33. Booth in the Booth scheme, -1 times the shifted multiplicand is In general, Algorithm selected when moving from 0 to 1, and +1 times the shifted multiplicand is selected when moving from 1 to 0, as the multiplier is scanned from right to left. 0 0 1 0 1 1 0 0 1 1 1 0 1 0 1 1 0 0 0 +1 ­1 +1 0 ­ 1 0 +1 0 0 ­ 1 +1 ­ 1 + 1 0 ­ 1 0 0 Booth recoding of a multiplier.
  • 34. Booth Algorithm 0 1 1 0 1 ( + 13) 0 1 1 0 1 X1 1 0 1 0 (­ 6) 0 ­ 1 +1 ­ 1 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 0 0 1 1 0 0 0 0 1 1 0 1 1 1 1 0 0 1 1 0 0 0 0 0 0 1 1 1 0 1 1 0 0 1 0 ( ­ 78) Booth multiplication with a negative multiplier.
  • 35. Booth Algorithm Multiplier V ersion of multiplicand selected by biti Bit i Bit i ­1 0 0 0 X M 0 1 + 1 X M 1 0 − 1 X M 1 1 0 X M Booth multiplier recoding table.
  • 36. Booth Algorithm  Best case – a long string of 1’s (skipping over 1s)  Worst case – 0’s and 1’s are alternating 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 Worst­case multiplier +1 ­ 1 +1 ­ 1 +1 ­ 1 +1 ­ 1 +1 ­ 1 +1 ­ 1 +1 ­ 1 +1 ­ 1 1 1 0 0 0 1 0 1 1 0 1 1 1 1 0 0 Ordinary multiplier 0 ­1 0 0 + 1 ­ 1 +1 0 ­ 1 +1 0 0 0 ­1 0 0 0 0 0 0 1 1 1 1 1 0 0 0 0 1 1 1 Good multiplier 0 0 0 +1 0 0 0 0 ­1 0 0 0 +1 0 0 ­1
  • 39. Bit-Pair Recoding of Multipliers  Bit-pair recoding halves the maximum number of summands (versions of the multiplicand). Sign extension Implied 0 to right of LSB 1 1 1 0 1 0 0 0 0 − 1 +1 − 1 0 0 −1 −2 (a)  Example of bit­pair recoding derived from Booth recoding
  • 40. Bit-Pair Recoding of Multipliers Multiplier bit­pair Multiplier bit on the right Multiplicand selected at position i i+1 i i−1 0 0 0 0 X  M 0 0 1 +1 X  M 0 1 0 +1 X  M 0 1 1 +2 X  M 1 0 0 −2 X  M 1 0 1 −1 X  M 1 1 0 −1 X  M 1 1 1 0 X  M (b) Table of multiplicand selection decisions
  • 41. Bit-Pair Recoding of Multipliers 0 1 1 0 1 0 ­ 1 +1 ­ 1 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 0 0 1 1 0 0 0 0 1 1 0 1 1 1 1 0 0 1 1 0 1 1 0 1 ( + 13) 0 0 0 0 0 0 × 1 1 0 1 0 (­ 6 ) 1 1 1 0 1 1 0 0 1 0 ( ­ 78 ) 0 1 1 0 1 0 ­1 ­2 1 1 1 1 1 0 0 1 1 0 1 1 1 1 0 0 1 1 0 0 0 0 0 0 1 1 1 0 1 1 0 0 1 0 Figure 6.15.  Multiplication requiring only n/2 summands.41
  • 42. Carry-Save Addition of Summands  CSA speeds up the addition process. P7 P6 P5 P4 P3 P2 P1 P0 42
  • 43. Carry-Save Addition of Summands(Cont.,) P7 P6 P5 P4 P3 P2 P1 P0
  • 44. Carry-Save Addition of Summands(Cont.,)  Consider the addition of many summands, we can:  Group the summands in threes and perform carry-save addition on each of these groups in parallel to generate a set of S and C vectors in one full-adder delay  Group all of the S and C vectors into threes, and perform carry-save addition on them, generating a further set of S and C vectors in one more full-adder delay  Continue with this process until there are only two vectors remaining  They can be added in a RCA or CLA to produce the desired product
  • 45. Carry-Save Addition of Summands 1 0 1 1 0 1 (45) M X 1 1 1 1 1 1 (63) Q 1 0 1 1 0 1 A 1 0 1 1 0 1 B 1 0 1 1 0 1 C 1 0 1 1 0 1 D 1 0 1 1 0 1 E 1 0 1 1 0 1 F 1 0 1 1 0 0 0 1 0 0 1 1 (2,835) Product Figure 6.17.  A multiplication example used to illustrate carry­save addition as shown in Figure 6.18.
  • 46. 1 0 1 1 0 1 M x 1 1 1 1 1 1 Q 1 0 1 1 0 1 A 1 0 1 1 0 1 B 1 0 1 1 0 1 C 1 1 0 0 0 0 1 1 S 1 0 0 1 1 1 1 0 0 C 1 1 0 1 1 0 1 D 1 0 1 1 0 1 E 1 0 1 1 0 1 F 1 1 0 0 0 0 1 1 S 2 0 0 1 1 1 1 0 0 C 2 1 1 0 0 0 0 1 1 S1 0 0 1 1 1 1 0 0 C 1 1 1 0 0 0 0 1 1 S2 1 1 0 1 0 1 0 0 0 1 1 S 3 0 0 0 0 1 0 1 1 0 0 0 C3 0 0 1 1 1 1 0 0 C2 0 1 0 1 1 1 0 1 0 0 1 1 S4 + 0 1 0 1 0 1 0 0 0 0 0 C 4 1 0 1 1 0 0 0 1 0 0 1 1 Product Figure 6.18. The multiplication example from Figure 6.17 performed using carry­save addition.
  • 48. Manual Division 21 10101 13 274   1101 100010010 26 1101 14 10000 13 1101 1 1110 1101 1  Longhand division examples.
  • 49. Longhand Division Steps  Position the divisor appropriately with respect to the dividend and performs a subtraction.  If the remainder is zero or positive, a quotient bit of 1 is determined, the remainder is extended by another bit of the dividend, the divisor is repositioned, and another subtraction is performed.  If the remainder is negative, a quotient bit of 0 is determined, the dividend is restored by adding back the divisor, and the divisor is repositioned for another subtraction.
  • 50. Circuit Arrangement Shift left an an­1 a0 qn­1 q0 Dividend Q A Quotient Setting N+1 bit  Add/Subtract adder Control Sequencer 0 mn­1 m0 Divisor M Figure 6.21. Circuit arrangement for binary division.
  • 51. Binary Division Hardware Implementation
  • 52. Restoring Division  Shift A and Q left one binary position  Subtract M from A, and place the answer back in A  If the sign of A is 1, set q0 to 0 and add M back to A (restore A); otherwise, set q0 to 1  Repeat these steps n times
  • 53. Examples Initially 0 Shift 0 0 0 0 0 0 0 0 0 1 0 0 1 1 1 0 0 0 0 0 0 Subtract 1 1 1 0 1 First cycle Set q0 1 1 1 1 0 Restore 1 1 0 0 0 0 1 0 0 0 0 1 0 Shift 0 0 0 1 0 0 0 0 1 1 1 0 0 0 Subtract 1 1 1 0 1 1 1 Set q0 1 1 1 1 1 Second cycle Restore 1 1 1 0 0 0 0 1 0 0 0 0 0 Shift 0 0 1 0 0 0 0 0 Subtract 1 1 1 0 1 Set q0 0 0 0 0 1 Third cycle Shift 0 0 0 1 0 0 0 0 1 Subtract 1 1 1 0 1 0 0 1 Set q0 1 1 1 1 1 Fourth cycle Restore 1 1 0 0 0 1 0 0 0 1 0 Remainder Quotient Figure 6.22. A restoring­division example.
  • 55. Nonrestoring Division  Avoid the need for restoring A after an unsuccessful subtraction.  Any idea?  Step 1: (Repeat n times)  If the sign of A is 0, shift A and Q left one bit position and subtract M from A; otherwise, shift A and Q left and add M to A.  Now, if the sign of A is 0, set q0 to 1; otherwise, set q0 to 0.  Step2: If the sign of A is 1, add M to A
  • 56. Examples Initially Shift 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 1 0 0 0 0 0 0 First cycle Subtract 1 1 1 0 1 Set q 0 1 1 1 1 0 0 0 0 0 Shift 1 1 1 0 0 0 0 0 Add 0 0 0 1 1 Second cycle Set q 1 1 1 1 1 0 0 0 0 0 Shift 1 1 1 1 0 0 0 0 1 1 1 1 1 Add 0 0 0 1 1 Third cycle Restore 0 0 0 1 1 Set q 0 0 0 0 1 0 0 0 1  remainder 0 Add 0 0 0 1 0 Remainder Shift 0 0 0 1 0 0 0 1 Subtract 1 1 1 0 1 Fourth cycle Set q 1 1 1 1 1 0 0 1 0 0 Quotient  A nonrestoring­division example.
  • 59. If b is a binary vector, then we have seen that it can be interpreted as an unsigned integer by: V(b) = b31.231 + b30.230 + bn-3.229 + .... + b1.21 + b0.20 This vector has an implicit binary point to its immediate right: b31b30b29....................b1b0. implicit binary point Suppose if the binary vector is interpreted with the implicit binary point is just left of the sign bit: implicit binary point .b31b30b29....................b1b0 The value of b is then given by: V(b) = b31.2-1 + b30.2-2 + b29.2-3 + .... + b1.2-31 + b0.2-32
  • 60. The value of the unsigned binary fraction is: V(b) = b31.2-1 + b30.2-2 + b29.2-3 + .... + b1.2-31 + b0.2-32 The range of the numbers represented in this format is: 0 ≤ V (b) ≤ 1 − 2 −32 ≈ 0.9999999998 In general for a n-bit binary fraction (a number with an assumed binary point at the immediate left of the vector), then the range of values is: 0 ≤ V (b) ≤ 1 − 2 − n
  • 61. •Previous representations have a fixed point. Either the point is to the immediate right or it is to the immediate left. This is called Fixed point representation. •Fixed point representation suffers from a drawback that the representation can only represent a finite range (and quite small) range of numbers. A more convenient representation is the scientific representation, where the numbers are represented in the form: x = m1 .m2 m3 m4 × b ±e Components of these numbers are: Mantissa (m), implied base (b), and exponent (e)
  • 62. A number such as the following is said to have 7 significant digits x = ±0.m1m2m3m4m5m6m7 × b ±e Fractions in the range 0.0 to 0.9999999 need about 24 bits of precision (in binary). For example the binary fraction with 24 1’s: 111111111111111111111111 = 0.9999999404 Not every real number between 0 and 0.9999999404 can be represented by a 24-bit fractional number. The smallest non-zero number that can be represented is: 000000000000000000000001 = 5.96046 x 10-8 Every other non-zero number is constructed in increments of this value.
  • 63. •In a 32-bit number, suppose we allocate 24 bits to represent a fractional mantissa. •Assume that the mantissa is represented in sign and magnitude format, and we have allocated one bit to represent the sign. •We allocate 7 bits to represent the exponent, and assume that the exponent is represented as a 2’s complement integer. •There are no bits allocated to represent the base, we assume that the base is implied for now, that is the base is 2. •Since a 7-bit 2’s complement number can represent values in the range -64 to 63, the range of numbers that can be represented is: 0.0000001 x 2-64 < = | x | <= 0.9999999 x 263 •In decimal representation this range is: 0.5421 x 10-20 < = | x | <= 9.2237 x 1018
  • 64. 1 7 24 Sign Exponent Fractional mantissa bit •24-bit mantissa with an implied binary point to the immediate left •7-bit exponent in 2’s complement form, and implied base is 2.
  • 65. Consider the number: x = 0.0004056781 x 1012 If the number is to be represented using only 7 significant mantissa digits, the representation ignoring rounding is: x = 0.0004056 x 1012 If the number is shifted so that as many significant digits are brought into 7 available slots: x = 0.4056781 x 109 = 0.0004056 x 1012 Exponent of x was decreased by 1 for every left shift of x. A number which is brought into a form so that all of the available mantissa digits are optimally used (this is different from all occupied which may not hold), is called a normalized number. Same methodology holds in the case of binary mantissas 0001101000(10110) x 28 = 1101000101(10) x 25
  • 66. •A floating point number is in normalized form if the most significant 1 in the mantissa is in the most significant bit of the mantissa. •All normalized floating point numbers in this system will be of the form: 0.1xxxxx.......xx Range of numbers representable in this system, if every number must be normalized is: 0.5 x 2-64 <= | x | < 1 x 263
  • 67. The procedure for normalizing a floating point number is: Do (until MSB of mantissa = = 1) Shift the mantissa left (or right) Decrement (increment) the exponent by 1 end do Applying the normalization procedure to: .000111001110....0010 x 2-62 gives: .111001110........ x 2-65 But we cannot represent an exponent of –65, in trying to normalize the number we have underflowed our representation. Applying the normalization procedure to: 1.00111000............x 263 gives: 0.100111..............x 264 This overflows the representation.
  • 68. So far we have assumed an implied base of 2, that is our floating point numbers are of the form: x = m 2e If we choose an implied base of 16, then: x = m 16e Then: y = (m.16) .16e-1 (m.24) .16e-1 = m . 16e = x •Thus, every four left shifts of a binary mantissa results in a decrease of 1 in a base 16 exponent. •Normalization in this case means shifting the mantissa until there is a 1 in the first four bits of the mantissa.
  • 69. •Rather than representing an exponent in 2’s complement form, it turns out to be more beneficial to represent the exponent in excess notation. •If 7 bits are allocated to the exponent, exponents can be represented in the range of -64 to +63, that is: -64 <= e <= 63 Exponent can also be represented using the following coding called as excess-64: E’ = Etrue + 64 In general, excess-p coding is represented as: E’ = Etrue + p True exponent of -64 is represented as 0 0 is represented as 64 63 is represented as 127 This enables efficient comparison of the relative sizes of two floating point numbers.
  • 70. IEEE Floating Point notation is the standard representation in use. There are two representations: - Single precision. - Double precision. Both have an implied base of 2. Single precision: - 32 bits (23-bit mantissa, 8-bit exponent in excess-127 representation) Double precision: - 64 bits (52-bit mantissa, 11-bit exponent in excess-1023 representation) Fractional mantissa, with an implied binary point at immediate left. Sign Exponent Mantissa 1 8 or 11 23 or 52
  • 71. •Floating point numbers have to be represented in a normalized form to maximize the use of available mantissa digits. •In a base-2 representation, this implies that the MSB of the mantissa is always equal to 1. •If every number is normalized, then the MSB of the mantissa is always 1. We can do away without storing the MSB. •IEEE notation assumes that all numbers are normalized so that the MSB of the mantissa is a 1 and does not store this bit. •So the real MSB of a number in the IEEE notation is either a 0 or a 1. •The values of the numbers represented in the IEEE single precision notation are of the form: (+,-) 1.M x 2(E - 127) •The hidden 1 forms the integer part of the mantissa. •Note that excess-127 and excess-1023 (not excess-128 or excess-1024) are used to represent the exponent.
  • 72. In the IEEE representation, the exponent is in excess-127 (excess-1023) notation. The actual exponents represented are: -126 <= E <= 127 and -1022 <= E <= 1023 not -127 <= E <= 128 and -1023 <= E <= 1024 This is because the IEEE uses the exponents -127 and 128 (and -1023 and 1024), that is the actual values 0 and 255 to represent special conditions: - Exact zero - Infinity
  • 73. Addition: 3.1415 x 108 + 1.19 x 106 = 3.1415 x 108 + 0.0119 x 108 = 3.1534 x 108 Multiplication: 3.1415 x 108 x 1.19 x 106 = (3.1415 x 1.19 ) x 10(8+6) Division: 3.1415 x 108 / 1.19 x 106 = (3.1415 / 1.19 ) x 10(8-6) Biased exponent problem: If a true exponent e is represented in excess-p notation, that is as e+p. Then consider what happens under multiplication: a. 10(x + p) * b. 10(y + p) = (a.b). 10(x + p + y +p) = (a.b). 10(x +y + 2p) Representing the result in excess-p notation implies that the exponent should be x+y+p. Instead it is x+y+2p. Biases should be handled in floating point arithmetic.
  • 74. Floating point arithmetic: ADD/SUB rule  Choose the number with the smaller exponent.  Shift its mantissa right until the exponents of both the numbers are equal.  Add or subtract the mantissas.  Determine the sign of the result.  Normalize the result if necessary and truncate/round to the number of mantissa bits. Note: This does not consider the possibility of overflow/underflow.
  • 75. Floating point arithmetic: MUL rule  Add the exponents.  Subtract the bias.  Multiply the mantissas and determine the sign of the result.  Normalize the result (if necessary).  Truncate/round the mantissa of the result.
  • 76. Floating point arithmetic: DIV rule  Subtract the exponents  Add the bias.  Divide the mantissas and determine the sign of the result.  Normalize the result if necessary.  Truncate/round the mantissa of the result. Note: Multiplication and division does not require alignment of the mantissas the way addition and subtraction does.
  • 77. While adding two floating point numbers with 24-bit mantissas, we shift the mantissa of the number with the smaller exponent to the right until the two exponents are equalized. This implies that mantissa bits may be lost during the right shift (that is, bits of precision may be shifted out of the mantissa being shifted). To prevent this, floating point operations are implemented by keeping guard bits, that is, extra bits of precision at the least significant end of the mantissa. The arithmetic on the mantissas is performed with these extra bits of precision. After an arithmetic operation, the guarded mantissas are: - Normalized (if necessary) - Converted back by a process called truncation/rounding to a 24-bit mantissa.
  • 78. Truncation/rounding  Straight chopping:  The guard bits (excess bits of precision) are dropped.  Von Neumann rounding:  If the guard bits are all 0, they are dropped.  However, if any bit of the guard bit is a 1, then the LSB of the retained bit is set to 1.  Rounding:  If there is a 1 in the MSB of the guard bit then a 1 is added to the LSB of the retained bits.
  • 79. Rounding  Rounding is evidently the most accurate truncation method.  However,  Rounding requires an addition operation.  Rounding may require a renormalization, if the addition operation de- normalizes the truncated number. 0.111111100000 rounds to 0.111111 + 0.000001 =1.000000 which must be renormalized to 0.100000  IEEE uses the rounding method.