Exercises

1:

Describe the architecture of a bicycle. What parts of the bicycle are not parts of the architecture? What are some implementation differences among different bicycles?

2:

Consider the car rental industry. What aspects of automobile architecture are essential to the ability of any driver to operate a randomly allocated rental car? What are some implementation differences among different automobiles that are not especially relevant to either the driver or the rental agency?

3:

Explain why harpsichords, organs, and accordions are not exemplars of "piano architecture."

4:

Why do you think that one computer manufacturer builds computers with a different architecture from those of another manufacturer? What does this imply about the importance of well-standardized high-level languages?

5:

What effect do you think the standardization of operating systems or command languages might have on the future development of computer architectures?

6:

Compute how much larger 1 EiB is than 1 EB, according to the IEC conventions for binary multiples.

7:

PA-RISC Architecture 1.1 defines extension registers that permit virtual addresses to be 16, 24, or 32 bits wider than the base level 32 bits. Express the three resulting maximum virtual address sizes using appropriate binary prefixes from Table 1-2.

8:

Adapt the program SQUARES in a high-level language to compute the cubes of the first five integers without using any explicit multiplication. Hint: An algorithm for N ³ can be discovered by writing down the series 1, 8, 27, 64, … and then inspecting the pattern of first, second, and third tabular differences. This is only a pencil-and-paper exercise, but save your work for future adaptation.

9:

Digital's first highly successful minicomputer, the PDP-8, was a 12-bit machine. What range of integers can be represented in a 12-bit unsigned binary number? In a 12-bit two's complement signed number?

10:

Convert 101010101₂ into hexadecimal and octal. Convert 10A34₁₆ into octal and decimal. Negate both values using 32-bit two's-complement hexadecimal form.

11:

Complete the following table by converting the given number in each row into the other bases.

	Decimal	Binary	Octal	Hexadecimal
a.	100
b.		100
c.			100
d.				100

12:

Perform the following hexadecimal arithmetic:

1. 205 6

2. AF9 + 9

3. 1A x B2 (use "long" multiplication, and think carefully about the "carries")

4. 1CFF + F2FF

13:

Write a high-level language program that inputs a decimal number and a radix and reformats that decimal number in the specified radix. For example, if the inputs are 10 and 16, the output should be A. Limit yourself to bases less than or equal to 16.