COMSATS University, Islamabad

Department of Computer Science

Maximum Marks: 100

Name: _____________ Reg #:_______________

Part-1 :
Course Title: CSC604 - Advanced Computer Architecture
Q1 [Marks=20]
Let suppose a computer has a cache, main memory (RAM) and a hard disk used for virtual
memory. An access to the cache takes is 1ns, access to main memory takes 10ns and access to
the disk takes 1000 ns. Suppose hit ratio of cache is 0.9 and the main memory is 0.8. The
effective access time required to access a referenced word on the system is the case when
1. Simultaneous access memory organization is used.
2. Hierarchical access memory organization is used.

Q2 [Marks 15=5*3]

1. State and define the hazards presented by instruction level parallelism. For each one,
indicate how it can be resolved.
2. Why do conditional branches impact the performance of a pipelined implementation?
3. What are advantages and disadvantages of the different types of computer architectures
that fall under Flynn’s taxonomy?
Part-2 :
Course Title: CSC604 - Advanced Computer Architecture Total marks=35

Q-1 marks=15
Consider that there are five universities Punjab University (PU),
University of Engineering & Technology, Lahore (UET), National
University of Science and Technology (NUST), Quid-e-Azam
University (QAU) and Lahore University of Management Sciences
(LUMS). These universities want to establish a video conferencing
network among them. With obvious notation, the matrix below
gives the cost (in thousands of dollars) of building a communication
link between any two universities.

Draw graph from given matrix and then find MST using
Prims Algorithm.
Q2. It is suggested that merge sort is not a good strategy for arrays
of size less than about 8 to 16, for which “insertion sort” may be
better. Realizing that in our “merge sort” algorithm, it is the arrays
of small sizes which are handled very often, it may be good idea to
modify our “merge sort” algorithm so that when the size of sub-
array is less than, say 8, it is not split further, but it is sorted by
insertion sort. Write Pseudo code for this modification in merge sort.
Question 2: marks=20
Find the time complexity of the following code. Use most suitable
asymptotic notation among O, Ω, and Θ to specify the time efficiency
class of the given code.
void fun(int n)
{
int k = n, j = 1, i = 0;
while (k > 0)
{
// Some O(1) task
while (i < n)
{
// Some O(1) task
i = i + j;
j++;
}
k = k/3
 }
}

Part-3 :
Course: Operating Systems Total marks= 30

Question No. 1: Inter-process Communication refers to the mechanisms and techniques used by
processes to communicate and synchronize with each other in a concurrent computing
environment, typically within the context of an operating system. Shared-memory and Message-
passing are the common methods used by OS to provide IPC service. Suppose a system consists
of multiple independent machines each having its own memory, connected by a network to work
in cooperation. Evaluate each of the aforementioned methods to provide IPC in this system.

Question No. 2: CPU scheduling plays a critical role in optimizing system performance and
resource utilization in modern operating systems. Evaluate the strengths and weaknesses of each
of the following algorithms, considering factors such as waiting time, response time, and
possibility of starvation.
First-Come, First-Served (FCFS), Shortest-job-first (SJF), Priority Scheduling and Round Robin
(RR)

Question No. 3: Consider a large application developed to provide financial services to its users.
Developers of this application realized that its performance gets inferior when the number of
users gets higher than a certain level. They analysed the application and found that there are
many services that can be implemented in parallel to improve the performance. The decided to
re-design the application to get the advantage of parallelism. There are two options to implement
the application either by implementing it using multiple processes or by using multiple threads.
Which approach is better? Give reasons.