Scribd
Upload
12 views4 pages

Chapter 10 Exercise 03

The document discusses various computing technologies, including multicore systems, near memory computation, machine learning, artificial neural networks, cloud computing, the Internet of Things (IoT), and Mobile Edge Computing (MEC). It outlines the merits and demerits of multicore systems, the advantages of near memory computation over traditional methods, and the significance of machine learning in various industries. Additionally, it explains the differences between deep learning and reinforcement learning, describes cloud computing and IoT's impact on human lives, and highlights how MEC improves upon Mobile Cloud Computing (MCC).

Uploaded by

brcwtwenencecznfzg
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
12 views4 pages

Chapter 10 Exercise 03

The document discusses various computing technologies, including multicore systems, near memory computation, machine learning, artificial neural networks, cloud computing, the Internet of Things (IoT), and Mobile Edge Computing (MEC). It outlines the merits and demerits of multicore systems, the advantages of near memory computation over traditional methods, and the significance of machine learning in various industries. Additionally, it explains the differences between deep learning and reinforcement learning, describes cloud computing and IoT's impact on human lives, and highlights how MEC improves upon Mobile Cloud Computing (MCC).

Uploaded by

brcwtwenencecznfzg
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 4

3.

Long answer questions:

a. Is multicore system a good choice? Explain with its merits,


demerits, and challenges.

Ans:- Yes, multicore system is a good choice.

Merits:
i. It is capable to perform more tasks compared to single core
processors.
ii. It can perform multiple tasks simultaneously with the help of
parallelism.
iii. It can process huge data compare to single core processor.
iv. A multicore system makes system faster.

Demerits:
i. Managing shared resources and scheduling application in many
cores becomes difficult.
ii. Multicore chips need larger area compared to single core.
iii. They are costly to manufacture.
iv. It needs high amounts of power than single core system.

Challenges:
i. Identifying how a task can be divided into separate,
concurrent sub-tasks.
ii. It is important to balance tasks between multiple cores to
make sure that each core is utilized effectively.
iii. Like tasks, data also needs to be divided to run on separate
cores.
iv. The data accessed by the tasks must be examined for
dependencies between two or more tasks.

b. What is near memory computation? How is it different from


traditional computations? Why are scientists doing research on
memory computation?

Ans:- The near or in-memory computing aims at processing close to


where the data resides. i.e, the processing element must be placed
close to the memory or storage.
In the traditional computational approach, the data is moved
from memory to the CPU for computation. When the data is huge or
when the data resides at a distance, it takes significant amount of
time to bring data which results in a performance bottleneck.
The paradigm of near memory computation has become prominent
with the introduction of 3D architecture. Active research is being
done in this field by both academic institutions and industries. It
is aimed at boosting the data processing capabilities of today’s
systems.

c. What is machine learning? Explain its importance in today’s era?


Ans:- Machine Learning is a subfield of AI(Artificial Intelligence).
It defines a technique where a machine can learn from its
experience. The machine learning algorithms are built based on some
models and the model is trained using a large dataset before using
it.
Machine learning has applications in all types of industries,
including manufacturing, financial services, healthcare, life
sciences etc. Suppose we want to build a handwriting recognition
system. No two person have same handwriting, but a machine(computer)
must be able to identify correctly all the handwritten characters.
To achieve this, we first feed the ML model with large set of sample
writings with the correct answers. Such trainings are expensive and
needs a lot of time. After the training the machine will be able to
extract some features or characteristics from the dataset, which
will help it to take correct decisions based on the given inputs.

d. What is an artificial neural network? Differentiate between deep


learning and reinforcement learning.

Ans:- AI (Artificial Intelligence) is the field of study of a device


that perceives its environment and takes actions to maximize its
chances of successfully achieving its goals. The field of study was
founded in 1955.

Deep Learning: A special class of machine learning (ML) is deep


learning (DL), that is based on an artificial neural network (ANN).
It uses multiple layers to extract higher-level features from the
input. An ANN is based on a collection of connected units or nodes
called artificial neurons. Each neuron can transmit signal to other
neurons (just like in a biological brain). An artificial neuron
which receives a signal then processes it and pass its output to
other connected neurons, the input in case of an ANN is a real
number, and the output of each neuron is calculated based on a non-
linear function of its input.

Reinforcement Learning: It is a paradigm of ML, where the model do


not have any labelled training data and the machine completely
learns from the environment is known as reinforcement learning (RL).
Based on the decision made, the machine receives some reward, and it
takes actions in an environment to maximize the cumulative reward.
Due to the availability of high-performance computers, this type of
ML is used in almost every field including operations research, game
theory, information theory, control theory, simulation-based
optimization, self-driving cars, and others.

e. Write a short note on cloud computing.


Ans:- Cloud computing is a resource sharing computing platform that
offers a large amount of space and computing capability to its users
via the Internet. The users are charged as a “pay-as-you-go” model
based on their resource usages. Computers with high compute and
storage capacities are placed in data centres. Users rents the
machines depending on their requirements and they are charged by the
service providers based on their use. Some popular cloud computing
providers are Amazon Web Services(AWS), Microsoft Azure, Google
Cloud etc.

f. What do you understand by the term IoT (Internet of Things)? How


can IoT impact human lives in the near future?

Ans:- The Internet of Things (IoT) describes the network of physical


objects or “things” that are embedded with sensors, computing and
communication elements for the purpose of connecting and exchanging
data with other devices and systems over the internet. These devices
range from ordinary household objects to sophisticated industrial
tools. The number of IoT devices connected to the internet is
expected to soon surpass 26 billion.
The internet of things is a prime technology that promises to
improve human lives. IoT has found its usage in every sector
including healthcare, home appliances, manufacturing, agriculture,
and others. IoT devices are also known as smart devices which can
take decision with minimal human intervention. The utmost utilities
achieved by IoT applications are decision-making and monitoring for
efficient and effective management.
g. Explain MEC framework with a diagram. How is it different from
MCC?

Ans:- MEC (Mobile Edge Computing) framework is introduced to bring


computation closer to the user. An example of MEC system is shown
below where few mobile devices are connected to some roadside units
(RSU). These RSUs along with some computation systems form the MEC
server. MEC is in turn connected to the Cloud system. The
applications running on mobile devices can send the compute-
intensive tasks to the nearby MEC servers at low overhead. A wide
range of mobile applications nowadays use MEC services to provide
quality of experience (QoE) to the users.

MEC helps in mitigating problems that are associated with MCC


such as high data latency, sometimes applications do not permit such
a long latency, MCC also requires a high Internet bandwidth. In case
of MEC since the computation system and the RSU are closer to the
users’ mobile devices the issue faced by MCC does not arise. The
basic principle of MEC is to bring computation power of MCC to the
edge of a network close to the users or the mobile devices.

You might also like