Proceedings of the SMART–2020, IEEE Conference ID: 50582

9th International Conference on System Modeling & Advancement in Research Trends, 4th–5th, December, 2020
Faculty of Engineering & Computing Sciences, Teerthanker Mahaveer University, Moradabad, India

Cloud Computing: Comparison and Analysis of

Cloud Service Providers—AWS, Microsoft
2020 9th International Conference System Modeling and Advancement in Research Trends (SMART) | 978-1-7281-8908-6/20/$31.00 ©2020 IEEE | DOI: 10.1109/SMART50582.2020.9337100

and Google

Dr. Manish Saraswat1 and Dr. R.C. Tripathi2

Department of MCA, Geetanjali Institute of Technical Studies, Udaipur, Rajasthan-India
1

2
Facutly of Engineering & Computing Sciences, Teerthanker Mahaveer University (TMU), Moradabad, UP-India
E-mail: [email protected], [email protected]

Abstract—The cloud computing refers to network During the last few years’ tremendous growth have
that enables to distribute processing, application, storage been seen in field of cloud computing, according to
capabilities among many remote located computer systems. In gartner[12] public cloud is expected $411 million till end
cloud computing platform the IT resources are utilized and of 2020. As more and more organizations are eternalized
released as per the requirement by using internet. It is a better
towards cloud services, it is becoming difficult to select
option to organizations and ordinary users to utilize services
(IaaS,PaaS ,SaaS,DaaS etc) as provided by cloud service appropriate service provider among a large pool of CSPs for
providers and need pay as use. Currently a large number of long terms of suitability. Now a day a large number of cloud
service providers in market and due to diversity of features providers have been emerges, but there no universal standard
and services, it very difficult to find suitable provider for s for and development of these providers in parallel directions
long term needs. As per market share top three providers are many providers focusing on computational capability and
Amazon, Microsoft and Google. In this paper we will analysis provide CPU, storage, Database and networking services
some of the tools such as compute, storage space management to end users. Some of the services providers focus in
and performance offered by AWS, Azure and GCP which are reducing the cost of the service, while others focus on
the top three market leaders in cloud computing technology.
un- interrupted service delivery and scalability factor for
In this paper, we will summarize and compare the features
of AWS, Azure & GCP to provide help to organizations and service delivery. These diversified dimensions have made
users to choose the suitable features which will fulfill the long it very difficult to choose a suitable service provider based
term requirements of the users. on the requirements of the individual user or organization.
Keywords: Cloud Computing, CSP, AWS, MS Azure, GCP, This paper presents a comparative feature study of top three
cloud computing platform, IaaS,PaaS, SaaS and, DaaS cloud service providers-AWS, MS Azure & Google’s cloud
computing platform to resolve the above dimensions. The
I. Introduction
next section in this paper will explain the fundamentals
In current era almost all organizations are need of cloud computing and cloud computing architecture
computerization, it is not possible to make large investment [4, 9, 13].
in infrastructure, IT resources (powerful network, server,
software and memory) and IT staff to manage them. Cloud II. Fundamentals of Clou Computing
computing is provides an Internet based platform to collect, Architecture
utilize and manage computer resources automatically [1]. The cloud computing is gaining attention and popularity
The cloud computing is providing a large number of services within computing industry. The Cloud computing is a
to users and organizations through it’s equipped data centres model for enabling ubiquitous, convenient, on-demand
with tightly coupled resources and available resources can network access to a shared pool of configurable computing
be dynamically provisioned to end users to fulfill their needs. resources which is rapidly provisioned and released with
The users or organization can consume all the available minimal effort’s. This is the definition of cloud computing
computation resources from large pool without thinking as per the National Institute of Standards and Technology
the resource limitations, source and scalability of cloud (NIST) taxonomy. There are five essential characteristics,
resources. Thus the users and organization has complete three service models, and four deployment models in
relief from initial investment to use cloud resources, and the cloud computing platform provided by AWS, MS
they have to payonly for the services to which they have Azure and GCP. All these are mentioned in the following
used [1, 2, 3]. table no-1:

Copyright © IEEE–2020 ISBN: 978-1-7281-8908-6 281

Authorized licensed use limited to: University of Prince Edward Island. Downloaded on May 31,2021 at 14:30:17 UTC from IEEE Xplore. Restrictions apply.
9th International Conference on System Modeling & Advancement in Research Trends, 4th–5th December, 2020
Faculty of Engineering & Computing Sciences, Teerthanker Mahaveer University, Moradabad, India
Table 1: Nist Based Cloud Models and Characteristics [1,2,3,9] •• Utility Computing: It is the fundamental concept
Service Models where computing resources like storage, network,
SaaS PaaS Build with: IaaS CPU etc. are provided to end users as utility
Use: It is a licensed This software can be Use: It is service.
Software on Demand develop and deployed, collection •• Service Oriented Architecture: Here data centers
Used by: End Users Users: Application Computer in cloud provide resources as web services
Services: E-Mail, Developers Services: Resources &
Google Execution, Messaging, Network’s (software, infra, platform) and heterogeneous
Doc, Finance, Games, App Development, Users: Infra and system incorporate each other.
Virtual Desktop, CRM, Integration, Network architect •• Scalability: The term scalability means ability to
Communication-Face Database-MySQL, Services: VMs, expand without limitations. This feature provides
book Oracle, Web Storage, Servers,
Server- Tomcat Networking, Load large scale of solutions in cloud computing; for
development tools, Balancers, System example Amazon, Google, and Microsoft etc.
RedHat Management have hundred thousands of servers in worldwide.
Deployment Models
The service providers can extend the nodes and
servers to meet the growing demands of software
Public Private Hybrid Commun
& infrastructure for the world- wide users.
Open for General Open for Open for ity Open
•• Reliability: The cloud uses multiple redundant
Public and Large licensed users, general for
Industrial group : single public & communit computational nodes which ensure high guaranteed
system and services organization: licensed y users availability of services this make the cloud more
are easily accessible services & users that have reliable than local computers.
to general public on accessibility of shared
•• Versatility: This characteristic ensures that
demand cloud concerns
infrastructure is (mission, cloud computing is not aim for certain specific
exclusively security, applications. A lot of applications can be run and
available within policy supported by cloud in parallel.
organization or etc.)
•• Virtualization This enable’s execution of multiple
owned persons
user’s requests on single physical platform. The
ESSENTIAL CHARACTERISTICS OF CLOUD COMPUTING
cloud uses hypervisor software which virtually
isolates the physical resources from the users by
a) ON DEMAND SELF SERVICE: This characteristic can be
using virtual machine instance.
explained as providing computer related services, such as E-mail,
Applications software, Network for Server etc. It is provided by AWS, •• Autonomic computing This ensure self
IBM, Saleforce.com, MS Azure etc. on demand without interaction management of cloud systems by allowing
with each service provider. automation in various tasks such as resource
(b) BROAD NETWORK ACCESS: In this all cloud capabilities provisioning, capacity management, service
are provided available over the network. It can be accessed by using measurement and disaster management etc.
dedicated broad band network and by using the heterogeneous client’s •• Maintainability The maintenance of resources is
devices such as mobile phones, laptop’s and desk top computers.
responsibility of service providers so it is easy to
(c) RESOURCE POOLING: In this method the resources of all the maintain in minimum cost [3, 10,11].
service providers are pooled together to serve the multiple customers
having different physical and virtual resources. In this setup the III. Literature Review
customer generally has no control or knowledge over the exact location
Zhong et al.[5] studied features of major cloud service
of the provided resources. The customer may be able to specify location
at a higher level of abstraction; It includes resources like storage, providers’ i.e AWS, Google, and Microsoft and concluded
processing, memory, and network bandwidth. that existing service providers offering diversified features
(D) RAPID ELASTICITY: In this capabilities are provided to users therefore end users force to use more than one service
in elastic manner and with provision of grab and released. The dynamic provider for business. Noman Islam et al.[8] presented
allocation helps to scale rapidly outward and inward bound data with a comparative study on various service providers and
demand. concluded that there are two directions of providers, one
(E) MEASURED SERVICE: In this method the service provider and focusing on reliability of service and another focusing
user can monitor, control, and measure all the resources to which they on cost reduction. Privacy and security issues are totally
are sharing. For example: storage, processing, bandwidth, and active
ignored. Buggya et al.[6] compared the features of AWS,
user account at a particular time can be monitored.
GCP, Microsoft live mesh, sunnetwork.com and found
There are several concepts, technologies, characteristics service providers features are not comply with market
which combine together to make the cloud computing needs and suggested more research is required to introduce
flexible and convenient for the user’s. We will discuss some new models. Hofer et al.[7] conducted a study on AWS,
of the important computing technologies in the section GCP, and Microsoft and concluded that interoperability is
below: major issue it must be address.
282 Copyright © IEEE–2020 ISBN: 978-1-7281-8908-6

Authorized licensed use limited to: University of Prince Edward Island. Downloaded on May 31,2021 at 14:30:17 UTC from IEEE Xplore. Restrictions apply.
 Market Share AWS AZURE GCP Microsof
tabulated
Growth Rate GCP AZURE AWS

Number of services AWS AZURE GCP Table:

Cloud Computing: Comparison andNumbers ofCloud
Analysis of Data Service Providers—AWS,
AWS AZURE GCP
Microsoft and Google
Centers Para
IV. Major Cloud Service Providers Number of services
Integration tools on- AWS
AZURE AZURE
AWS GCP
GCP
In present time’s a lot of cloud vendors are there in the Numbers of Data AWS AZURE GCP Infra
premise service
market. Some of popular vendors are mentioned below: Centers (IaaS
a. Amazon Web Service (AWS): It was launched in Integration tools on- AZURE AWS GCP
collec
hardw
year 2006. premise service
softw
b. Microsoft Azure: It was launched in year 2010.
eleme
c. Google Compute Engine: It was launched in year 6 ace to
2008. enabl
5
d. Sale force: It was launched in year 2012. cloud
e. Alibaba: It was launched in year 2009. comp
f. Rackspace: It was launched in year 2008. 3 Series 1 servic
g. IBM It was launched in year 2011 Series 2 virtua
2 resou
h. Oracle: It was launched in year 2012, & many
more… 1 Series 3 Comp
Servi
In this paper our focus is only on the products and (PaaS
characteristics of top three vendors; such as, Amazon, CategoryCategoryCategoryCategory Provi
Google and Microsoft that are proving extensive cloud 1 2 3 4 as ser
services. The information about products, features of these the in
vendors have been collected from provider’s website. Fig. 1: Performance Chart of all the Cloud Vendors
A. Amazon Web Service (AWS) The Fig 1 explains the comparative performance chart
The Amazon is one of the pioneer and oldest cloud of all the cloud technology vendors. The above analysis
service providing company, which has started its cloud shows that MS Azure is the best among all the cloud
based services in year 2006. It has allowed cloud users to computing service providers in present time.
perform and manage the server’s instances on data centers C. Google Cloud Platform (GCP)
from its inception. It is the one of top CSPs in IaaS (market The Google is one of the fastest growing company in IaaS
share: 47.8%) and PaaS (market share: 34%) till the end of and PaaS with growth rate of 60.2% and 57.9% respectively
2019. It is offering cloud based storage, computation and till end of 2019. It joined the cloud market in year 2008 by
others functionalities which can be accessed by cloud users simply providing E-mail, Calendars, online documentation.
to set up applications and services on demand through the GCP is the goggle’s hosting service for web application, it
REST and SOAP protocol over the HTTP [4,5,19]. uses pre-defined runtime environment for development and
B. Microsoft Azure deployment of cloud based applications by applying java
The Microsoft is a pioneer company to provide cloud and Python. The primary drivers behind the GCP popularity
services and holding top place in SaaS (17% of market Share) are its scalability, user’s based infrastructure and payment.
and second in PaaS and IaaS where it sharing approximate The comparison of various criteria’s among AWS, Azure
15.5% market till the end of 2019. The Microsoft Azure, and GCP is tabulated as [10,17]:
allows clients to use the cloud services purely on the cloud, V. Methodology
or on combination of any existing applications, data centre
Before moving on cloud the customer must analyze
or infrastructure. It is an ideal vendor for hybrid cloud
and review several points like objective to moving on cloud,
combination model. The MS Azure uses following three
ideal deployment and service model because each model
principles to provide the cloud services
has different level of controls, offerings, need of application
•• Windows based environment that is used to run
migration on cloud, security and cost management. There
applications and store at data-centers
are a lot of parameters which must be understood by
•• SQL-Azure offers data services based on SQL
customers to choose right service providers to fulfill their
server & (iii).Net services which provides the
needs. The parameters are:
distributed services as infrastructure for local as
•• infrastructure & computing services
well as cloud based applications. [7,8,15]
•• network technologies services
Table 2: Rank of Major Cloud Vendors
•• storage technologies
Criteria Rank-1 Rank-2 Rank-3 •• data base support
Entered in Market AWS GCP AZURE •• back-up services
Availability Zones AWS AZURE GCP •• key tools. On the basis of stated parameters,
Market Share AWS AZURE GCP The cloud service providing companies such as
Amazon, Microsoft and Goggle are providing various
Growth Rate GCP AZURE AWS
products which are tabulated below [8,10],:
Copyright © IEEE–2020 ISBN: 978-1-7281-8908-6 283

Authorized licensed use limited to: University of Prince Edward Island. Downloaded on May 31,2021 at 14:30:17 UTC from IEEE Xplore. Restrictions apply.
9th International Conference on System Modeling & Advancement in Research Trends, 4th–5th December, 2020
Faculty of Engineering & Computing Sciences, Teerthanker Mahaveer University, Moradabad, India
Table: 3 Summery of technical parameters of AWS, Azure and GCP[6,812,14]
Parameter Amazon Web Service Microsoft Azure Google Cloud Platform
Infrastructure (IaaS) collection Amazon- EC2(Xen Microsoft common language run Google Cloud Platform (GCP)
of hardware/ software elements/ virtualization Engine time(CLR) (Hypervisor-V)
interf ace to enable/access cloud
computing services/
virtualized resources.
Computing Services (PaaS) : AWS Platform-as-a- service(PaaS) Google App Engine; Graphic
Provide the IT as service over Beanstalk, Amazon Lightsail, Function-as-a- service(FaaS) processing unit; Docker
the internet or AWS Batch, AWS Service Fabric Container;
Azure Batch
network with delivery on demand Fargate, Compute Engine; &
and price as –use.
Network Technologies: It allow to Amazon allows VPC services; It provides branch cache, direct Cloud Interconnect, Cloud VPN,
access the network infra- structure Amazon Route 53 services, The access for services, It helps in core Carrier Peering, and Direct
directly and securely. Cloud Front service, & AWS- Network Guide framing, DNS Peering
Direct Service, and load balancing
Connection.
Storage technologies : maintain Following services of Amazon can Azure It allows scalable object Data Store, Blob store Google
and manage the data and makes be availed such as simple storage store It can store text and binary cloud, SQL
the accessible on network on service, Amazon Glacier, Amazon data. Azure Files: It manages
demand) Elastic Block Store, & Amazon file’s for cloud for on- premises
Elastic file System. deployments. Azure Queues: A
messaging
Relational Data Base Support: Amazon- RDS, SQL, Azure SQL, Azure-SQL- DW Cloud-SQL, Cloud- Spanner
allow to users to access the MySQL and Oracle
database without installing
software/setting- up
environment
Back-Up Services: restore/backup Glacier Archival storage, Recovery Near-line, Cold-line are frequently
of data or application backups, Site recovery used to access data
remotely or online.
Key Tools: for AI/IOT/ AI/ML:, AI/ML: AI/ML: The Cloud service helps in
serverless Machine Learning, Translate, Machine Learning, Azure Bot Machine Learning, Cloud Natural
Deep Lens, Deep Learning, Service & many more can be Language, Cloud Speech NLP,
Apache IoT: Core, Free RTOS, availed The
Device Defender, IoT Device Cloud Translation API, For Cloud
Management. Serverless: AWS- IoT: IoT Hub Video, Intelligence, & Cloud Job
Examples are ,IoT Edge, Stream Analytics, Time Discovery
Series Insights Serverless: This IoT: Cloud
Functions
can also be
Lambda, Serverless Application availed here, IoT Core can be managed
Repository. Serverless: Cloud Functions are
available
Thus we analyzed several parameters of AWS, Azure and GCP which may be useful for cloud users to choose
appropriate service provider for their business or organization. Studied parameters are listed as:
Table 4: Summery of General Parameters of AWS, Azure and GCP[13,17,18,19]
Parameter Amazon Web Service (AWS) Microsoft Azure Google Cloud Platform (GCP)
Type of Service Model All computing models- IaaS, All three computing models- All three computing models-
PaaS,SaaS but major in IaaS IaaS,PaaS,SaaS but major in SaaS IaaS,PaaS, SaaS but
a,PaaS major in PaaS
Focus Enterprise Friendly offering s and Microsoft tools Integration, Open Open Source & Portability and
Public Cloud (off Premise) Source and Hybrid Cloud(on/off Hybrid Cloud
premise) (off premise)
Availability Regions & Range of 66 It is extremely wide 54 It is wide 20 It is just Limited but is
services expandable
Unique Selling Point of cloud This provides Tight security This Integrated with Microsoft It helps in Top-notch data &
computing: Ability to quickly system tools analytics tools
respond

284 Copyright © IEEE–2020 ISBN: 978-1-7281-8908-6

Authorized licensed use limited to: University of Prince Edward Island. Downloaded on May 31,2021 at 14:30:17 UTC from IEEE Xplore. Restrictions apply.
 Cloud Computing: Comparison and Analysis of Cloud Service Providers—AWS, Microsoft and Google

Security Moderate Moderate Moderate

Nature of services It provides Enterprise- centric It is Not totally ‘enterprise- ready’ It is Designed for cloud- based
services businesses
Reach It is an Established global presence It is Expanding gradually It is New to the market
Major drawback It is Difficult to use, options may The Management tooling is It has Limited data centers across
seem confusing inefficient the world
Price and Payment Free, Commercial (Pay-as-you- Commercial (Pay-as-you- Free, Commercia l (Pay-as- you-
go) ($0.11/Hour- ( roundup) go),($0.02/Min utes) (roundup- go) ($0.02/Min utes) (Minimum 10
Subscription: Hourly/ Monthly commitment Subscription: minutes) Subscriptio n:
Monthly Hourly/Mo nthly.

VI. Conclusions Referecnes

The selections of vender depend on business needs and [1] Cloud Computing by A. Srinivasan, Pearson India ISBN:
technical requirements of particular company. During the 9789332537439 Cloud Computing: Concepts, Technology &
Architecture, Prentice Hall Service Technology Series 1st Edition,
study it is observed certain organizations need a specific and ISBN-10: 9780133387520
cloud service provider while others need the guidance to [2] Architecting the Cloud: Design Decisions for Cloud Computing
choose most appropriate vender for business or to provide Service Models (SaaS, PaaS, and IaaS),ist Edition, Wiley by J.Kavis,
the services to customers. The conclusions for AWS, Azure ISBN-10: 1118617614
and GCP are as given below: [3] Eissa Alreshidi, “Comparative Review of Well Known Cloud Service
Providers”, Sci.Int(Lahore).31(1)B.65- 170,2019,ISSN-1013-5316
A. AWS [4] Q Zhing,L Cheng,” Cloud Computing: State of Art and research
•• More Global reach: it has a lot of Data Centers. challenges”,journal of International Services and Applications,Vol-
1,issue-1,2010.
•• Wider. Stable, Reliable lot of Services: It has large
[5] R Buyyaa,Yeoa, ”Cloud Computing and Emerging IT Platform”,
number of products. Future Generation Computer Systems, vol 25,2009,pp-599-616.
•• Need a Public solutions [6] C.N Hofer and Karagiannis,”cloud computing services: Taxonomy
•• You are a larger company: Its pricing policies and comparison”, Journal of Internet services and Applications, Vol-
move towards enterprise level companies where 2, 2011, pp-81-94
[7] Noman Islam and Aqueel ur Rahaman,”A Comparative Study of
need versatile solutions even expensive.
Major Cloud Service providers for Cloud Computing”, IEEE 4th
B. AZURE International Conference on Computer and Communication Systems
•• Migrating on cloud first time (ICCCS); 2019
[8] Cloud Computing Vendor & Service Provider Comparison, Crisp
•• Most of Applications/solutions are Microsoft/ Vender Universe, Research Report-2108
Windows based. [9] Rajendran and Swamynathan, “Parameters for Comparing cloud
•• Need a Hybrid deployment Mode/solutions service Providers: A Comparative Analysis”, DOI: 10.1109/
•• You are a Developer CESYS.2016.7889826, IEEE Xplore: 30 March 201711.Dutta and
DUtta, ”Comparative Study of Cloud Services offered by Amazon,
C. GCP Microsoft,and Google, International Journal of Trends in scientific
•• Focusing on container based model(containers Research and Development(IJTSRD),Vol-3,issue-3, April-2019
are types of software that can share OS Kernel [10] Sajay and Suvanam, “A Study of Cloud Computing Environments
for High Performance Applications”, Proceedings of 2016
virtually without
International Conference on Data Mining and Advanced Computing
•• need of virtual machines) (SAPIENCE)-2016.
•• Focusing on web based applications [11] Almishal and Youssef, ”Cloud Service Providers: A Comparative
•• Flexible pricing You are already digital Study, International Journal of Computer Applications & Information
organization and now moving on to develop & Technology Vol. 5, Issue II April May 2014 (ISSN: 2278-7720)
[12] https://www.gartner.com/, Accessed 31 January, 2020
deploy cloud based apps
[13] http://cloudcomputing.bestfind.com/, Accessed 11 February,
•• and soft wares 2020 15.https://www.azure.microsoft.com/Solutions/, accessed 11
Therefore AWS is best when you are looking for a February, 2020
platform with broadest reliable and stable services and [14] 16.https://www.gartner.com/reviews/customers- choice/,accessed 18
pricing is not a major challenge when the if business servers February, 2020 [3]
[15] .https://cloud.google.com/gcp/, Accessed 02 March, 2020 18.https://
are running on windows then Azure is best because it
azure.microsoft.com/en-in/free/cloud- services/,Accessed 06 March,
offering windows platform integrated services. If you are a 2020
start-up and medium-sized company and rapidly scaling up [16] 19.https://aws.amazon.com/, Accessed 10 March, 2020
with large volume of user data, innovations and resources
then GCP is ideal.

Copyright © IEEE–2020 ISBN: 978-1-7281-8908-6 285

Authorized licensed use limited to: University of Prince Edward Island. Downloaded on May 31,2021 at 14:30:17 UTC from IEEE Xplore. Restrictions apply.