Reference no: EM131958748

Project -

Project Description -

(A) Revise the floating point addition unit presented in the textbook (Figure 3.15) to reflect the following requirements/restrictions:

(1) add two numbers of same sign only (i.e. both positive or both negative),

(2) replace the rounding unit with a truncation unit, i.e. modify Figure 3.15 by simplifying the normalization stage and control unit so there is no iteration back),

(3) explicitly show how hidden one is handled. No need to trace data.

(B) Design a floating point multiplication unit with overflow/underflow detection feature for multiplying two 32-bit floating point numbers in IEEE 754 format. An overflow detection will output 00 if no overflow/underflow; output 10 if overflow; 01 if underflow (note: you decide what is the input to this unit.) No need to show details of the overflow detection unit, i.e. make it a black box). When underflow occurs , i.e. the exponent of the result is too (negatively) smaller to fit the field, set both exponent and fraction parts with all 0s (i.e. the result is +0.0 or -0.0 in this case.) When overflow occurs, i.e. the exponent of the result is too (positively) large to fit the field, set the exponent to all 1s, and fraction to all 0s. In both cases the sign bit should be calculated as usual. Also use a truncation unit instead of the rounding method so that only one time/round of normalization needed. Show how the hidden 1 is handled. Trace your design with 3 sets of given data. Assume an integer multiplication unit available to use, treat it as a black box, i.e. no need to show details.

Expected Results - The following products are expected:

(1) A neatly (computer or hand) drawn hardware diagrams (similar to Figure 3.15) for a simplified FP addition units as described in (A) above.

(2) A neatly (computer or hand) drawn hardware diagram similar (in format) to the one in Figure 3.15 but performs the multiplication with overflow/underflow detection unit added. For each functional unit, write a meaningful name on it, i.e. ALU, integer_multiplier (or i-mult), AND/OR/XOR/NOR gate, shift left/right, ...

(3) For the FP multiplication unit, trace the test data (a), (b), and (c). Here tracing a test data set means that you walk through the diagram and show the input/output of each major functional unit in the diagram. For each test case, you may make a copy of your diagram, and put tracing data along each functional unit of your diagram. The following sample tracing gives you a rough idea of the input/output data for each functional unit (note: functional units may vary.)

Explain the principle of the overflow detection in your design of FP multiplication unit.

(1) How to determine there is an underflow?

(2) How to determine there is an overflow?

Answer yes or no and provide a brief justification to each answer.

(1) Will your FP addition unit work if one of the numbers is 0.0?

(2) Will your FP multiplication unit work if one of the numbers is 0.0?

Attachment:- Assignment Files.rar