[SOLVED] COMPSCI 4043 System and Networks 2021 R

System and Networks

COMPSCI 4043

Monday 26 April 2021

1.    (a)     Express the following in 32-bit two’s complement code, giving your answers in hexadecimal. Show your working.

i.        1023               ii.       -1023 [4]

(b)    If the calculation 100 + 30 – 20 is performed by an 8-bit CPU, using an 8-bit two’s complement

code, will an overflow be generated? Explain your answer and say what result you’d expect to be generated. [3]

(c)     For the mathematical set of integers, subtraction is always the same as adding the inverse of a  number. Thus 3 – 2 = 3 + (-2) and 3 – (-2) = 3 + 2. Is this also true for a two’s complement code? Justify your answer. [2]

(d)    Write a Sigma16 program that accesses an array X of n 16-bit two’s complement numbers and for

each element X[i] in X, stores a 0 or 1 in the corresponding element of a second array Y according to the following rule.

Y[i] = 0 if X[i] is odd; Y[i] = 1 if Y[i] is even [7]

(e)    Write a segment of Sigma16 code (not a complete program) which sets R2 to 0 if the second most significant bit in R1 is a 0, and sets R2 to 1,  otherwise. Estimate how many Sigma16 cycles your segment will take to run as you’ve written it. [4]

For reference, here is part of the instruction set of the Sigma16 CPU.

2      (a)      How do cache memories speedup program execution? Many caches only cache read cycles. Why is this

and why is it usually not seen as a major limitation? Discuss whether there are circumstances where caching write cycles would provide some benefit. [6]

(b)     The following Sigma 16 code is intended to take a 10-element array of two’scomplement numbers (only first element is shown) and replace all the elements with their twos complement inverse. The number $8000 is not permitted. However, although the code will assemble, it contains several errors.

i.       Draw up aregister use table for the program (suitable for inclusion as comment).

ii.       Identify the errors and explain how you would correct them.

iii.       Write out the corrected program.

[6]

(c)      Estimate how many memory cycles the corrected program would take to run. [4]

(d)     In the corrected program estimate the advantage a system with a cache memory would gain if a primary memory cycle takes 10ns and a cache cycle 1ns. [4]

3.          (a)           In Internet communications, how does a router get an IP packet to its destination? Explain why the routing approach used by IP would not work with either a TCP segment or an Ethernet frame. [6]

(b)          An IP interface has the address 192.10.0.1. What binary value does this address represent? Also write the value in hex. [3]

(c)           “ IP is not reliable and IP carries TCP; therefore, TCP cannot be reliable either.”  Comment on the   truth of the premises and conclusion of this argument and discuss whether the reasoning is valid. [6]

(d)          Many TCP connections can be running simultaneously on a given host machine. Referring to what you know about the communication system and its relationship to the operating system, explain   how this is possible. [5]