Sunday, May 6, 2012

Let's Be Futuristic for a Moment

I am fascinated by the artificial intelligence (AI) technology. It excites me but at the same time it scares me. I worry about becoming an obsolete species on the earth. Even though I tend to believe that it will be human being's ultimate fate like dinosaurs, but this time it will be AIs who may replace us! Nevertheless, AI technology promises to take the civilization much further. Does cloud computing share any link with AI technology? Can AI affect cloud computing or vice versa? Please remember that the current AI technologies are not even close to what scientists and researchers predict it to be. Watch the video to find out the current state of AIs:


Its functionality is extremely limited even though it still affects the everyday world especially the world wide web in a big way. I wanted the last post of this blogging cycle to be looking towards the future so that you will have some thoughts for the week. Here is a video that provides research findings on how AI can perform surveillance.:

 

Let us start with a recent announcement from Google. Machine Learning is considered to be a form of AI. It can learn while processing data and has the ability to be proficient in a particular application over time. Google announced a new machine learning engine (http://econfuture.wordpress.com/2010/08/31/google-cloud-computing-machine-learning-and-self-destruction/) that will be offered as a part of Google's cloud computing strategy. It promises to perform customer sentiment analysis, message routing decisions, recommendations, suspicious activity detection etc. In other word it can make some ambiguous decisions which computers so far has not been able to do. Thus it can take over some business activities which have been performed by human beings so far. This new application will make an organization flatter by eliminating middle managers and will enhance the capabilities of low-wage offshore workers. Essentially, this new engine will motivate companies to move its IT needs to Google cloud. Here is a video that explains the importance of AI in businesses:



We have been concentrating on how cloud computing can enable small and medium businesses and start ups to compete with large enterprises by transforming the IT infrastructure cost from a fixed cost to a variable cost. But this new Google offering may shake the ecosystem again. It will empower the large and complex organizations to streamline the workforce hence will make it easier to run. The large corporations will essentially have similar organizational structure as the SMBs but with other assets such as more capital, machines and intellectual properties.

Cloud computing technology emphasizes on storing the applications, processes and data in the same cloud. Thus AIs can access an extremely large volume of data and can use unlimited computing power to make its decision. Traditional IT infrastructures, though extremely efficient, still cannot compete with clouds in this aspect. Thus cloud computing will help AIs to make better predictions (http://omarsbrain.wordpress.com/2010/09/17/artificial-intelligence-raining-from-%E2%80%9Cthe-cloud%E2%80%9D-on-ubiquitous-computers/).

It seems like cloud computing and AI can help each other to reach some of their promised potential in the future and a lot of researchers in both fields are working towards it. I will finish this post by letting you watch a video on how IBM's Watson, which is capable of answering ambiguous questions to some extent, functions:



Enterprise Resource Planning and Cloud Computing

Enterprise Resource Planning (ERP), an important source of competitive advantage for companies, essentially integrates management information systems of all the business functions across the organizations, assimilates the organization with its business partners around the globe and automates the whole information flow process. You all know the importance and functionality of ERP. If you want to start somewhere here is a video for you on ERP:

If you are a student of business management like me, you must have come across numerous case studies on how successful construction and implementation of ERP has made many organization successful. Now some people are suggesting that moving your own ERP system to cloud  can be beneficial. I was as confused as you are feeling right now. Given that every ERP system is extremely customized according to the need of the user organization, your ERP is very unique. Depending on how creative and strategy oriented you had been during the development process of your own ERP system, you can actually differentiate yourself either by adding value or by reducing cost. Most of the time ERP assumes the central position in the core business. Then why are companies such as SAP, Netsuite and others are developing cloud ERPs for medium and small businesses and why would companies avail a standardized ERP system? Let us solve that puzzle. 

Cloud EPR market is ruled by NetSuite at this point followed by SAP. The true ERP providers are listed below:

IMA conducted a survey that reveals how ready companies are to get into cloud ERP and here is the result (http://www.netsuiteblogs.com/blog/2011/11/the-ima-survey-issues-facing-erp.html):

Versus 12 months ago

The same survey also revealed the following data on evaluation criteria for choosing cloud ERP:

Areas to Look for

But why would any company want to move its existing ERP system to a cloud? The main reason are the scalability of cloud ERP and believe it or not better security prospect of using cloud ERP. (http://cloudcomputingtopics.com/2012/04/why-you-should-move-your-erp-into-the-cloud/).
By deploying the existing ERP to the cloud, a company can make it available to all the authenticated users inside or outside of the company and connect them through internet. In the traditional ERP model, most of the big organizations had to essentially construct and integrate the IT infrastructure of business partners such as suppliers and distributors from around the globe. In this extremely competitive global economy, organizations look for vendors and business partners in all corners of the world to bring down the cost of production. Though IT has been a great enabler in this effort by facilitating the flow of information among partners, it has become a significant cost for companies. Cloud ERP promises to break that barrier. 

Another obstacle of traditional ERP system has been its rigidity. Constructing an ERP system in costly and once an ERP system is in operation, its hard to change it. Since cloud ERPs enable interoperability and access, an organization can add and integrate additional and emergent functions to the system. In the same way, the business intelligent system can be connected to the cloud ERP now. 

SAP offers the following classes of cloud ERP currently (http://www.oracle.com/us/solutions/sap/database/sap-erp-cloud-352626.pdf)


SAP Business by design is for medium and small businesses and SAP on-demand-services cater to the large enterprises representing add-on features. 


SAP business suite is a construction toolkit for for developing an in-house SaaS ERP cloud stack.

Constructing ERP system is a capital intensive process. Small and medium businesses (SMBs) mostly lacked the necessary capital to avail an in-house traditional ERP system and thus has been unable to reduce the barriers of entry in many sectors. Cloud ERP essentially removes that barrier by proving on-demand services and by eliminating the capital expense for building an ERP infrastructure. Here is a video that would explain this phenomenon:


Cloud ERP market is in the extreme nascent state at this point. But experts believe that Microsoft, Oracle and other usual suspects with enter this market soon. The providers need to enhance the reliability, connectivity issues and interoperability among platforms from different countries to motivate users to avail cloud ERPs. Below is the Gartner hype cycle for cloud ERP (http://www.cloudcomputing-news.net/blog-hub/2012/jan/10/how-cloud-computing-and-erp-mobility-are-reordering-gartners-hype-cycle-for-erp/):

The market prediction of cloud ERP is promising but as you can see the enterprises are not quite ready to move to this particular cloud immediately (http://www.cloudcomputing-news.net/blog-hub/2012/apr/16/roundup-of-saas-erp-forecasts-and-market-estimates-2012/)


Let me finish this post by an interesting development in the cloud ERP sector. Groupon recently partnered with Netsuite to accommodate its global hyper growth. Watch the following video to find out the reasons behind Groupon's move:


Lets move on to Fin Cloud Now

This year IDC reported that the securities and investment will be the earliest adopters of cloud computing while banking industry may lag behind because of various regulations that shapes the industry.  (http://www.microsoft.com/en-us/news/download/features/2012/IDC_Cloud_jobs_White_Paper.pdf)
But most experts agree that financial services industry will be benefiting immensely from cloud computing.

Is such phenomenon evident in the market? Believe it or not, there exists a 'FinCloud Institute' (http://www.fincloudinstitute.org/) whose mission is " to provide research-based insights into how Cloud Computing strategies are being implemented by financial institutions.". Below is the picture of the institute itself:




Yes, you guessed it right! Cloud Computing is a welcomed term in the financial services world. Don't get me wrong! The industry is in the early stage of adoption and not all firms are moving their IT need to cloud yet. But the industry is intrigued by the possibilities that cloud promises and is evaluation the technology with great interest. 
The following diagram depicts the current breadth of cloud computing usage in the financial services industry:
 (http://www.wallstreetandtech.com/it-infrastructure/224200533?pgno=2)

The above diagram reveals an interesting fact to me. The current usage shows that Financial Services uses cloud computing for a wide range of applications including risk management, price modelling and valuation and wealth management applications. If you think about it, cloud computing is not so different compared to outsourcing from the perspective of risk. Companies have been reluctant to outsource processes that are strategic for the company and needs subjective decision making because they are hard to codify
("Getting Offsoring Right", Aron R. and Singh, J. V., Harvard Business Review, 2005, 83 (12), 135-143) While back-office operations and client report is easily codifiable, risk management and wealth management have traditionally not been outsourced with a few exceptions. The above diagram shows that while 12% of the companies uses cloud for risk management, 17% of the companies use cloud for back office operations. The numbers are not that different and may make cloud computing more relevant in the future because users are willing to move functions with more operational and structural risks to cloud.

The biggest enabler for financial services industry seems to be the scalability aspect followed by cost saving. Take a look at the following diagram: (http://www.wallstreetandtech.com/it-infrastructure/224200533?pgno=2)


In 2011, NYSE Technology revealed a plan to build a custom cloud for financial markets (http://meship.com/Blog/2011/06/01/nyse-builds-a-custom-cloud-for-financial-markets/) with EMC Corp. and VMware INC. This cloud aims to provide a platform that will aggregate exchanges, broker dealer and buy-side participants. The stakeholders will have access to historical market data and analysis. Such a strategy will look into reducing information assymetry among buyers, sellers, small firms and large firms and in the process will make the market more efficient and friction free. Below is a video that explains NYSE technologies focus:

Cloud computing promises to be a great enabler for trading services since it can make the trading process faster. Currently, trading technology operates on millisecond scale (http://www.bis.gov.uk/assets/foresight/docs/computer-trading/11-1222-dr3-technology-trends-in-financial-markets) . In other words, it takes milliseconds to a price data packet or r volume data packet from exchange to the traders' desk. By using proximity server and real time computing power, cloud computing promises to bring down the trading time to microsecond in the near future.

After the financial meltdown of 2008, the relevance of cloud computing has increased manifold for financial services industry. Especially the governance and compliance issues of financial services industry can be addressed by using cloud computing since the technology is a great aggregator and can make the idustry more transparent. There is ongoing research (http://www.iseing.org/emcis/emcis2010/Proceedings/Accepted%20Refereed%20Papers/C70.pdf) that uses service software engineering to lower the political, social and technical barriers and to build a global infrastructure for financial services.

Here is a video that explains how cloud computing can address specific needs for finance industry:


Some recent adoption stories of cloud computing in financial services industry can be found in the fincloud website (http://www.fincloudinstitute.org/). Recently  WePay hired the PCI-complaint cloud services of Layered Tech as an online payment platform. A consortium of eight regional banks chose Axletree to avail SaaS based SWIFT connectivity to increase security and reliability while decreasing cost of ownership. Confirmation.com developed a cloud based solution to prevent false information reporting an collusion in the financial services industry.

For me, the most interesting proposal so far has been how a cloud based iPad App can enable real time trading. Essentially cloud can provide fully interactive research tools and can stream charts news, data and real time data to mobile phones and iPads to increase the devices' trading capabilities. The cloud based app makes the interface of iPad and mobile phones much more simplistic and easy to use. Here is how it looks like (http://media.wiley.com/Lux/55/235955.image1.jpg)

image1.jpg  

Let me finish this post by discussing the security issue that prevents the financial industry to adopt cloud computing at a greater rate. Because of various regulations that are characteristics to the industry, the companies are extremely cautious about cloud computing. Since the technology is new and executives still lack the necessary knowledge to understand the technology, they are comprehensive about it. But with time, executives are converging towards the consensus that they have been using third parties for a long time for many of their operations and business functions and cloud computing actually is less risky than a third patry who is located offshore. Such a realization will bring down the resistance in the financial services industry towards cloud computing. With this comment, I will finish this post and let you watch the following video that explains the current state of readiness of financial industry towards cloud computing:


Helix Nebula: Its a Science Cloud!

By now, you must have understood that in this post I will be covering the applicability of cloud computing phenomenon in basic scientific research and development. We are talking about connecting the dots between a technology that offers unlimited computing power and the science community who generate, analyze and integrate an unfathomable amount of data. Its a natural fit! But there exist some other benefits of cloud computing that are extremely appealing to the researchers in various fields of science. Lets start with the Helix Nebula, the latest initiative to build an European cloud for scientific research, then move on to some more specific science clouds such as cloud for material science, cloud for life sciences etc.

In this blog, you have seen several characteristics of cloud computing and the most attractive one for various industries is the on-demand-service (http://csrc.nist.gov/publications/nistpubs/800-145/SP800-145.pdf) that essentially enables businesses to reduce their IT infrastructure cost. But for scientific community, the resource pooling and rapid elasticity (http://csrc.nist.gov/publications/nistpubs/800-145/SP800-145.pdf) aspects of cloud computing are the biggest enablers. These two aspects together render unlimited computing power and an extremely efficient and transparent collaboration platform.  Watch the following video that explains why such aspects are attractive to the researchers and scientists:


The European Organization for Nuclear Research, famously know as CERN, is the home of the large Hadron collider that empower high energy physics research. On March 12, 2012, CERN  announced (http://press.web.cern.ch/press/pressreleases/releases2012/PR03.12E.html) that  it will collaborate with European Molecular Biology Laboratory (EMBL), European Space Agency (ESA) and a consortium of leading IT providers to launch the "Helix Nebula-the Science Cloud" that will host the massive IT requirement of European Science community and later will be available to industries and various governments institutions. You have to remember, that CERN's grid computing facility is considered to be able to handle extremely high volume data. Here is a video that explains CERN's current computing capabilities:


So why do they need Helix Nebula? Lets look at some numbers. CERN  produces 6 GB of data per second, keeps 150,000 CPUs continuously busy and stores 15 petabytes of data per year. (http://gigaom.com/cloud/super-science-cloud-coming-to-europe/). The scientific community is extremely data intensive and has an insatiable hunger for computer power and storage. This enormous volume of data needs to be analyzed in different ways by different data mining processes to extract valuable an logical information. Helix Nebula will be used by EMBL to simplify the analysis of large genomes and the European Space Agency will use Helix Nebula to study earthquakes and volcanoes. Companies such as Atos, Capegemini, Cloud Sigma, Logica, SAP etc will be the commercial partners in the Helix Nebula Project. The official report on the strategic planning process is available here: (http://cdsweb.cern.ch/record/1374172/files/CERN-OPEN-2011-036.pdf)

Researchers in University of Washington (http://arxiv.org/ftp/arxiv/papers/1110/1110.0543.pdf) are trying to develop a Science Cloud Computing (SCC) platform that will provide simulation capabilities specific for Material Science, Condensed Matter Physics and Quantum Mechanics. This platform can be used in Amazon EC2 and consists of a virtual machine with UNIX operating system, material science codes and relevant interfacing tools. Thus this cloud will be able to generate simulations and solve the structure of novel materials just like local computer clusters that are being used currently. But the cloud will have more computing power, it will reduce the time to solve the structure and will be able to increase collaborative efforts among scientists. Here is sample simulation picture  that provides structural details of a superconductor(http://arxiv.org/ftp/arxiv/papers/1110/1110.0543.pdf).


Similar efforts can be observed to facilitate life sciences projects which also in extremely data intensive. Solving the sequence of a specific human genome and to facilitate virtualized and networked environment of R&D, Oracle will be offering a health sciences cloud in the near future. This cloud will be a automated continuum of practices and process (http://www.oracle.com/us/industries/life-sciences/oracle-health-sciences-cloud-wp-367168.pdf) and would look like the following:

Oracle recently released the following video as well:


IBM's life sciences cloud aims to facilitate the bacteria research:


It is evident that cloud computing is extremely relevant for scientific community because it has the scalability to accommodate unlimited volume of data, can provide faster and efficient data mining and analyzing capabilities and hence brings down the time to finish an experiment and facilitate collaboration on a standardized platform that can be accessed and used by researchers from all over the world in real time. The possibility that the cloud platform will enable integration of various fields of science to solve a problem in a virtual world where distance is not an obstacle anymore, is extremely attractive to the science community. I believe in the future we will see a lot of customized clouds for various fields of scientific research. Some of the pressing issues of current era such as food shortage, search for renewable alternative energy and finding novel materials for semiconductor production will find cloud computing to be a great enabler.

Sunday, April 22, 2012

Cloud Computing Promises to Improve Healthcare Services

As a student of Johns Hopkins University, I get exposed to various aspects of healthcare almost everyday in all of my classes. Additionally, healthcare has been a subject of interest for all relevant magazines, newspapers and blogs because of the introduction of President Obama's Universal Healthcare law. Thus whenever I encounter a new technological innovation, I tend to be curious about its impact on healthcare industry. In this post, I will try to reveal how healthcare industry can leverage benefits of cloud computing to improve patient care and to make healthcare more affordable.

In my second post titled " Why Should MBA Students Pay Attention to Cloud Computing", I have introduced the concept of readiness of various industries towards adopting cloud computing. ICD estimates that (http://www.microsoft.com/en-us/news/download/features/2012/IDC_Cloud_jobs_White_Paper.pdf)
healthcare industry would be slow in adopting cloud computing because of regulations and security measures. But healthcare reform is expected to push the industry to look for ways to reduce the cost of healthcare. Thus experts believe that in the near future, acceptance of cloud computing in the healthcare industry will see a positive growth.

In fact, Accenture reports that (http://newsroom.accenture.com/images/20020/HealthcareCloud.pdf) 32% healthcare respondents are currently using cloud computing in some form.
The same report mentions that 73% of cloud users in healthcare industry plan to move more applications to cloud in the near future.
Before I get into the particular benefits that cloud computing promises to bring into healthcare industry, let me show you some videos where experts discuss impacts of cloud computing on healthcare industry. the first video is a cartoon that explains some of the benefits of cloud computing.


Here is another one where experts from both the ICT field and healthcare field explains the benefits and usage:

So as you can see, people in healthcare sector are not only aware of cloud computing, they are interested in the technology and are exploring various aspects of the technology before they can decide to adopt it.

Potential Benefits for Healthcare Industry

  • Healthcare databases are characterized by the presence of fragmented, redundant and inconsistent data. Cloud computing can provide better database management that would lead to higher quality of patient information and better update processes.(http://newsroom.accenture.com/images/20020/HealthcareCloud.pdf )
  • Effectiveness of Electronic Medical Records (EMR) has been low (http://www.johnseelybrown.com/cloudcomputingdisruption.pdf ) because it was primarily driven by demand of the healthcare providers, had a high cost of implementation, generated low ROI, demanded high level of change management and was characterized by technological challenges such as integration difficulties and security and privatization. Cloud computing can change the scenario by introducing collaboration among participants, high performance, ease of usage and open standards into the system. These phenomena will accommodate demand  of proactive consumers, attend to wellness and chronic illness management, facilitate share management of care for aging parents/relatives and would improve patient safety. 
  • Cloud computing will also be able to facilitate collaboration among different systems of different healthcare organizations that in turn would lead to efficient information sharing among organizations. This would lead to better care for patients. 
  • The three most important benefits of cloud computing that we have already discussed in the first post of this blog are cost reduction, enhancement of the speed of the operating process and flexibility or on-demand scalability that are generic to all industries and healthcare is not an exception.
Accenture (http://newsroom.accenture.com/images/20020/HealthcareCloud.pdf) provides the example of Pathwork Diagnostics, a developer of diagnostic tests. Pathwork helps oncologists identify hard-to-find tumors. It uses a machine learning algorithm to match the specimen of a tumor with the specimens of other tumors, specifications of which  are stored in a very large library. This process is extremely time consuming and can take months to complete using a mid-level server. Thus Pathwork opted for cloud computing that enabled on-demand scalability for the company. 
  • Small hospitals and nursing homes usually do not have an extensive IT infrastructure and cannot afford enough IT stuff to maintain their existing IT needs or to integrate new IT innovations. For such organizations, cloud computing will prove to be a great enabler by providing all such facilities that won'e need to be maintained in-house. 
  • In the article titled " A cloud computing solution for patient's data collection in health care institutions". Rolim et al. introduced a cloud based system (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3222190/ ) that enables automation of the process that collects vital patient data via a network of sensors which are connected to legacy medical devices. The main benefit of the system is that it eliminates manual data collection and enables real-time data collection under all circumstances. 
  • Another obvious benefit of cloud computing is that it enables physicians to access patient data anytime and anywhere through mobile devices such as smartphones and tablets. 
  • Cloud computing potentially can provide unlimited computing power. Thus diagnosis of various illnesses, human genome sequencing, image analysis, drawing analogies among different genomes, identifying pathogens etc. using cloud computing is cost effective, reliable and fast. The bioinformatic research sector has been the early adopter of cloud computing to leverage such benefits. (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3222190/)
Thus it is evident that healthcare industry can benefit tremendously by adopting cloud computing. So why don't we see widespread adoption of cloud computing among healthcare providers? The providers are still concerned about data security in the cloud. They also do not trust the cloud service providers to protect the privacy of the patients. Government regulations in the healthcare arena are extremely strict and sometimes providers do not possess enough knowledge about the minute details of such regulations. Thus healthcare providers tends to be extremely cautious so that they do not break any law unintentionally. Healthcare sector is also characterized by loss of governance and organizational inertia (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3222190/) that keep them away from introducing technological change in their systems. 

But its not all bad news. The scenario is changing over time as healthcare industry has been witnessing other sectors adopting cloud computing. Such strategies of other sectors encourage the healthcare providers to explore benefits and relevance of cloud computing. The following video discusses the change in attitude  of healthcare providers towards cloud computing:

Let me finish this post by saying that the penetration of cloud computing in healthcare sector is inevitable. Because of the traditional characteristic of the sector, the adoption rate might be slow. But the benefits of cloud computing are too obvious for healthcare providers to ignore. 

Impact of Cloud Computing on Semiconductor Industry and Nanotechnology

We all know what Semiconductors are. Semiconductor materials have intermediate conductivity between conductors and insulators. Silicon and Germanium are examples of semiconductor materials.  If you want to know the basics of semiconductors, this site should be very helpful to you. http://www.wsi.tum.de/Institute/Scientificbackground/AnIntroductiontoSemiconductors/tabid/65/Default.aspx
There are three most commonly used semiconductor devices, namely diodes, transistors and integrated circuits.These devices together are essentially the building blocks and brains of almost all electronic devices such as computers, mobile phone, electronic appliances etc. The CPU in your computer and smartphone uses microchips that are made of silicone semiconductor. Below is a picture of such a microchip (http://www.topnews.in/chip-vendors-track-losses-299220)


Here is a video that would give you a comprehensive view of this very important sector that has been a growth factor for many industries and accounts for 10% of global GDP.


The biggest name in semiconductor industry is Intel. Some of the other important firms are AMD in PC chips market and ARM that holds 95% market share in the global mobile phone chip market. The diagram below summarizes the 2011 global MPU market share:


Cloud computing, as we have seen previously, will transfer end users' ICT need to the clouds and thus will lower the demand of client storage facilities. Number of servers will be consolidated in a cloud and usage of desktops will be reduced. Will such phenomenon affect the semiconductor industry negatively? Let's find out.

The server market (http://www.gsaglobal.org/events/2011/0908/docs/GSA_Octane_PPT_web.pdf)
is projected to see a medium growth because of the emergence of cloud computing. While the need of building data centers to facilitate cloud capabilities and initial duplication will drive the demand of servers, virtualization effect will lead lowered demand of servers. Thus the net effect would be a medium growth of server market that in turn would add to the growth of semiconductor industry.

Demand for desktops will continue to go down because of the availability of mobile models. Even though this shrinkage is not entirely due to the cloud computing, the growth in demand of cloud-driven mini notebooks does facilitate this shrinkage. Improved LAN/WAN connectivity in low end PCs facilitates cloud computing and thus the demand for low end PCs will go up. Thus the net effect will be very high demand of cloud driven mini notebooks and low end PCs, that in turn will spur the further growth of semiconductor industry.

3G and 4G technology increased the global availability of mobile cloud services (http://www.gsaglobal.org/events/2011/0908/docs/GSA_Octane_PPT_web.pdf). Here is a video that talks about this phenomenon:

Smartphones offer better interfaces than that of PCs and will represent 60% of mobile chip market in 2013. Such projections indicate high projected growth in the mobile chip industry.

With the widespread usage of cloud, communication infrastructure will increase manifold that in turn would lead to high demand in semiconductor industry.

Need of high security measures will facilitate growth in secure network and storage systems that in turn would lead to building better network paths. This phenomenon would also increase the demand in semiconductor industry. Thus positive growth in storage infrastructure and consumer devices would lead to higher demand for semiconductor chips. As I have discussed before, only client server market will see a negative demand curve. Thus the overall effect of cloud computing on semiconductor industry is positive and cloud computing would be one of the major drivers for innovations in electronic industry hence in semiconductor industry.
Sergis Mushell, research analyst for Gartner research, summarizes (http://www.gsaglobal.org/events/2011/0908/docs/GSA_Octane_PPT_web.pdf) these finding in the following table:

Nanotechnology sector is very closely related to semiconductor industry, alternative energy industry and biotechnology industry. Since its inception, nanotechnology industry has worked towards increasing the efficiency of semiconductor chips by introducing new materials such as carbon nanotubes and graphene. Below is a picture of such carbon based materials.

I have worked with single walled carbon nanotubes extensively as a researcher. In my opinion, needs of cloud computing sector would drive innovations in the nanotechnology sector as well. Let me introduce my opinions before I finish this post.

As evident from our discussion on cloud computing so far, to be attractive, cloud computing would need to be fast and reliable. Thus cloud servers which are used by many users at the same time, would need to be more efficient. They would need to use chips that are not only lightweight and small, but also are energy efficient in the sense that they would need to generate less heat and consume less energy. Silicon based semiconductor industry has followed Moore's law successfully so far that essentially says number of transistors that can be placed on ICs can be doubled in every two years without increasing the cost but increasing the efficiency of the ICs. Today, the transistors are at the 32 nm level (2.3 billion transistors placed 32 nm apart). After it reaches 16 nm level (where electrons will travel between 200 atoms), it can not be made smaller anymore because then it will approach a single atomic layer of silicon. Here is a video that will explain the end of Moore's Law:


Graphene which is one-atom-thick layer of carbon, conduct electricity 30 times faster (http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2011/08/20/BUV81KPF77.DTL) than silicon.Semiconductors made of this material would not only speed up the computer but also would consume very little energy. Scientists predict a very new type of transistors can be produced using this materials. Similar phenomenon are expected from single walled carbon nanotubes as well (http://www.eet.bme.hu/~mizsei/Nanoelektronika/nanotubesemi.pdf).

Thus I believe demand of efficiency, reliability and availability of cloud computing would encourage semiconductor companies to experiment with these new materials that in turn would lead to achieving the full potential of nanotechnology. Firms such as Intel and IBM with history of futuristic research are deploying substantial resources towards finding new materials in their purest form to facilitate the growth of nanotechnology. Such efforts will lead to new innovations in material science and nanotechnology which in turn will change the current landscape of semiconductor industry and future cloud computing sector.

Current Business Model of Cloud Computing Providers: Financial Aspects

In this post, I want to tell you how the cloud computing providers generate profit and how they can improve their profitability in the future.

Current State

Let's start with the financial aspects of the current business models of cloud computing providers. There are three components that are important in the business model.

  • Value Proposition
  • Pricing Mechanism (Revenue Model)
  • Cost Structure
The diagram below shows how business model of cloud computing is described by Osterwalder, A. in 2004. (http://www.hec.unil.ch/aosterwa/PhD/Osterwalder_PhD_BM_Ontology.pdf
As I understand it, the value proposition of a particular provider would determine what the provider's cost structure would be and what kind of price the provider can demand from its customers. Thus a short discussion of the current value chain is necessary to understand the cost structure and pricing mechanism. 

Value Chain

The diagram below( http://www.happiestminds.com/services/it-services/cloud-business-solutions) describes the cloud computing value chain and was developed based on the work of Jaekel and Luhn (2009) titled "Cloud Computing – Business Models, Value Creation Dynamics and Advantages for Customers", Leimeister et al. (2010) titled "The business perspective of cloud computing: actors, roles, and value networks", and Zhang et al. (2010) titled "Cloud computing: state-of-the-art and research challenges". 
By now, you probably understood that firms choose to serve any number of components of this value chain according to their capabilities and long term strategies. Moreover, firms form strategic relationships, technical alliances and consortium memberships to integrate functions so that they can deliver a value added service to the consumers. 

Pricing Mechanism

Osterwalder, A. in 2004 (http://www.hec.unil.ch/aosterwa/PhD/Osterwalder_PhD_BM_Ontology.pdf) developed the probable pricing strategies for cloud computing providers which I have summarized in the diagram below. If you want to read the detailed discussion on each type of pricing, you can consult this article: http://www.intertic.org/Policy%20Papers/Jaatmaa.pdf, page 27-28.

 Most of the current cloud providers offer fixed pricing and volume dependent pricing. Some of the industry giants provide some customization as most of the consumers are SMEs and look for little to no customization. Thus you can find some differential pricing in the market mainly because of volume dependent pricing. The market pricing category, in my opinion, would be relevant mainly for big enterprises which are yet to move their entire ICT need to Clouds. As I discussed in the previous post, providers need to innovate more to offer additional value added services to attract big enterprises. 

Cost Structure

 In the Service Delivery Book, Office of Government Commerce (2001) lists hardware, software, people, accommodation, external service and transfer as traditional ICT cost types. Because of the presence of elastic resource utilization and virtualization, cost structure in the cloud computing paradigm differs significantly from that of the traditional ICT. In the article titled "The Method and Tool of Cost Analysis for Cloud Computing (2009)", Li et al. listed two cost types that are to be included while calculating the accounting cost of cloud computing for the providers. 
  • Total cost of ownership (TCO): Its the fixed cost of providers that they incur to build the cloud infrastructure and to operate the cloud. 
  • Cloud utilization cost: Its the variable cost that is incurred by the users depending on extent of usage. 
Li et al. developed the following diagram to depict the cost structure of cloud computing:


Greenberg et al. have come up with four cost elements in their article titled "The cost of a cloud: research problems in data center networks". These elements are summarized below (http://www.intertic.org/Policy%20Papers/Jaatmaa.pdf):


Because of automation, operational staff cost is below 5% are thus is not included in the cost calculation.
Depending on the existing capabilities and long term strategies, providers design their value propositions. Cost structure and pricing strategy or revenue model are direct dervatives of the value propositions. Thus ROI of the providers depends how attractive their value propositions are. 

Recommendations

I believe that in the near future, cloud service providers would have to come up with more customization and value added services either by themselves or by collaborating with other firms that possess different capability sets to convince big enterprises to move their ICT capabilities and need to clouds. Here is a video that emphasizes the important of product differentiation in cloud computing paradigm:

An analogy can be drawn from Citibank's e-business strategy (Citibank's E-business Strategy for Global Corporate Banking, HKU797, October 24, 2008). I have came across this case study in my Digital Marketplace class at Carey Business School. The case describes how Citibank developed and consolidated its e-business strategy according to the customers need. Citibank targeted big corporations for whom it not only offered a truly global e-banking service that provided transaction services, corporate finance services and treasury market services, but also offered to perform back office tasks for the corporations. Thus Citibank offered a complete solution package along with consultancy to integrate customers' need while developing a consolidated global e-business strategy. 

The cloud service providers should follow a similar strategy while formulating their service offerings. The companies that are already in the cloud computing sector or are contemplating entering this sector, have already been in the ICT field offering other ICT related services. Firms such as IBM and Cisco have established themselves in the IT consulting arena and can use this capability to work with both SMEs and big enterprises closely to provide a complete solution package. Such solution package should include a road map for the clients complete ICT need including infrastructure, platform,  softwares and applications and storage, agility consideration, data monitoring, risk assessment, measures against failures, data monitoring and security. To enable such offerings, the providers should built capabilities to accommodate customization and develop their strategies accordingly. 

New entrants should have some advantages because they do not have an existing cloud and thus can built one that is scalable and agile and can accommodate customers's need. Last but not the list, every cloud provider should address the issue of security which seems to be the main concern of the consumers and some experts in the field. Many companies hesitate to move their ICT needs to cloud because of the security issue. As I described in my post titled "Critical Notes on Cloud Computing", scientists and engineers are producing extremely promising research in this arema. Cloud computing providers should pay close attention to new finding in this area and should facilitate such research to come up with innovative ways to ensure security in the cloud. Here is a video that discusses efforts of Professor Ken Birman in this area:


Thus to ensure economy of scale and economy of scope, cloud computing providers should ensure security in the cloud, innovate more to offer customization and should provide a complete solution so that enterprises can move their entire ICT needs to cloud. These are the ways by which cloud can reach its full potential and can prove to be a disruptive innovation to change the face of ICT industry.

In my next post I am going to introduce cloud's effect on semiconductor industry which has been one of the most important facilitators of traditional ICT industry. 

Industry Analysis and Current Competitive Landscape of Cloud Computing: A Providers' Perspective

In my last three posts, I have introduced the concepts and technology behind cloud computing, how cloud computing can potentially change business models of companies in various sectors, some critical issues with cloud computing and how the future is projected for cloud computing. In this series, I am going to talk about relevance and benefits of cloud computing for particular industry sectors.

As has been projected by several research firm such as Forrester and Gartner,the potential future demand of cloud computing exemplifies the pervasiveness of cloud computing. Such a characteristic of this technology is bound to attract more firms to enter the cloud computing industry as providers. Currently, the landscape of this sector is characterized by the presence of big firms . In this post, I am going to point out some of the key characteristics of the sector and the competitive landscape of the sector from a provider's point of view.

As can be seen from the list below (http://cloudcomputingcompaniesnow.com/), there are many cloud computing providers and some of them has been trusted brand names in the information and communication technology sector and e-business sector for quite some time now.

These organizations differ in their service offering. Below is a view of the larger cloud market that I found in this website: http://cloudtimes.org/2011/11/30/top-paas-saas-and-iaas-cloud-companies-by-cloudtimes/



While Amazon is the biggest name in the IaaS market, Salesforce emerged as the best performer in the SaaS market. Though it is hard to quantify such measurements, the best approximations are as follows (http://cloud-computing-today.com/2011/06/30/quantifying-cloud-computing-market-share-2/):

Amazon's revenue from IaaS was $500-700 million followed by Rackspace which generated a revenue of $100 million approximately.

In the SaaS space, Salesforce has emerged to be the best performer with a revenue of $1.3 billion followed by NetSuite, Rightnow, SuccessFactor etc. with revenues around $200 million each.


Industry Analysis of Cloud Computing

The most comprehensive way to analyze an industry for me is to apply Porter's Five Forces. Thus I am going to begin the cloud computing sector analysis in the light of five forces. I have consulted the book titled "Cloud Computing: Principles and Paradigms" by Buyya, R., Broberg, J. and Goscinski, A. for this analysis.

The diagram below is the representation ((http://www.lidi.info.unlp.edu.ar/WorldComp2011-Mirror/SER8106.pdf) of cloud market using Porter's Five Forces (Market Model)  



Cloud Market
Presence of many firms in the sector leads to a less concentrated market. Such low concentration indicates a highly competitive sector. Presence of almost all ICT industry giants (https://www.deloitte.com/assets/Dcom-Global/Local%20Assets/Documents/TMT/cloud_-_market_overview_and_perspective.pdf) such as IBM, Apple and Google also is an indication of high existing rivalry among competitors. Though their service offerings differ depending on IaaS, PaaS, SaaS or combination of any two, even in the same service area the competition tends to stay high because of low product differentiation.

New Entrants
In SaaS market, the threat of new entrants is high because of the requirement of low initial investment, low time to market and low exit barrier. In PaaS market (https://www.deloitte.com/assets/Dcom-Global/Local%20Assets/Documents/TMT/cloud_-_market_overview_and_perspective.pdf), presence of industry giants such as Google and Amazon creates a high barrier to entry which can be reduced by product differentiation and innovation. In IaaS market, threat of new entrants has been low because of the requirement of high initial investment and presence of high risk.

Buyers
Bargaining power of buyers tend to stay high in the SaaS market because of low switching cost and availability of many choices (http://www.lidi.info.unlp.edu.ar/WorldComp2011-Mirror/SER8106.pdf).
In PaaS market, lack of interoperability standards and usage of proprietary development language lead to a vendor lock-in which in turn leads to high switching cost for buyers. Thus in PaaS market, bargaining power of buyers tend to stay low because of high switching cost. Below is a video from Microsoft Azure which reveals how customers of this cloud would not be able to switch to a different cloud easily because their applications were developed on Azure. 


IaaS is considered to be a commodity. Thus product differentiation is low and bargaining power of buyers is high. Amazon EC2 has been the early mover in the sector and is preferred by buyers mainly because of its brand name. Thus creating brand loyalty is only way to stay competitive in this market.

Suppliers
Suppliers in the cloud computing market are the cloud service providers. In SaaS market, suppliers have low bargaining power because of the low switching cost of buyers and presence of many players. In PaaS market, as I have discussed before, vendor lock-in leads to high bargaining power of suppliers. In IaaS, brand loyalty is the only way providers can gain some bargaining power.

Technology Development
Threat of substitute for all cloud computing services has been low because the cloud computing technology itself is in the growth phase. Open Source computing can be the only substitute for cloud computing. But SMEs tend to prefer cloud computing because of availability of consulting services from the providers and non existent switching cost. 

Thus the overall cloud market looks very attractive from the providers' perspective. The sector has enough demand and supporting technology to reach economy of scale. Furthermore, substantial research conducted by academia and industry in this area and enriched collaboration between academic and corporate R&D will lead to new innovations that have the potential to enable product and service differentiation for new entrants in the future. Even though economy of scope is not yet possible in this sector, it can achieved in the future with new innovations.

Let me finish this discussion by showing you where cloud computing stands on the Rogers S curve. Don't curse me yet. I am going to tell you what Rogers S curve is. As the name suggests, this curve is shaped like S (mathematically, it is known as logistic function) and was introduced by Everett Rogers to describe the technology adoption life cycle. It depicts how the adoption of a new innovation progresses over time depending on the demographic and psychological characteristics of the targeted adopters. According to many experts, the S curve for cloud computing is as follows: (http://www.informationweek.com/news/software/info_management/228900680)


What does this mean? The early adopters of cloud computing are SMEs because in its current state cloud computing meets the demand pattern and IT need criteria of the same. But to attract big enterprises that already have extensive IT infrastructures,  cloud computing providers need to add more value to their services and would need to introduce more service innovations. Some of the big corporations use Clouds mainly as a storage space but are yet to exploit cloud's full potential. To switch from their existing IT infrastructure to Cloud, the big enterprises would demand more functionality, additional value added services and customization.

I hope the this post was able to present a broad view of the cloud computing sector from the providers' perspective. In my next post, I will discuss the current cost structure, value chain and profitability of cloud computing providers. I will try to present my ideas on what kind of product differentiation and customization would attract big enterprises to move to cloud and how a new entrant can compete successfully with the existing brand names. 

Sunday, April 8, 2012

Critical Notes on Cloud Computing

This would be my third and final post on the introduction of cloud computing. In previous two posts, I introduced the basic concepts and technologies associated with cloud computing, various advantages of cloud computing and micro and macro economic impacts that cloud computing can potentially generate. Now let us analyze this technology critically to find out some of the disadvantages and other critical issues associated with cloud computing and the whether providers can overcome such issues in the future.

Recently Amazon's EC2 faced outage in its Virginia data center that affected more than 70 clients including Reddit, Quora and New York Times ((http://money.cnn.com/2011/04/22/technology/amazon_ec2_cloud_outage/index.htm). Amazon links many of its data centers together to protect customers from an isolated failures. Such a strategy should ensure a downtime of 4.4 hours. But the recent outage has lasted for 36 hours. Amazon explained that a 'networking event' caused many of Amazon's storage volumes to create self back-ups. Amazon's storage capacity was reduced in the process, which in turn, prevented clients from accessing the cloud. The details of Amazon's explanation can be found in the website http://aws.amazon.com/message/65648/.
Here is a picture that summarizes the basic infrastructure of Amazon's cloud services that I found  here: http://www.computescotland.com/network-online-power-outages-4231.php
This outage problem gives rise to a critical issue associated with cloud computing- Reliability. After reading my first two posts, you must have wondered with so much potential, why usage of cloud computing is not so wide spread yet. Here are some of the expert's opinion against cloud computing:

  • Richard Stallman, founder of Free Software Foundation, thinks that users should not give away their information to a third party.
  • Larry Ellison, Founder of Oracle Corporation, thinks adoption of cloud computing is 'fashion driven'. You can listen to him in the following video:


The main criticisms against cloud computing have to do with privacy and ownership. Three most critical issues with cloud computing are:

  • Reliability: When an application runs on a cloud server, the same server is being accessed by other customers at the same time. Because the interaction between client terminal and cloud is achieved over internet, depending on internet connectivity such a situation increases the probability of failure. The following charts from the site http://www.marketingautomationsoftware.com/blog/reliability-in-the-cloud-1060711/ summarizes the monthly downtime of email providers:


  • Control: The users or customers of cloud computing services lose physical control over their own data and would have to trust the provider completely since the data is stored on providers' servers (http://minutiaeetal.wordpress.com/2008/10/25/three-critiques-of-cloud-computing/ ). Data is alterable, can be deleted with ease and can be locked out as we saw in case of Amazon's EC2. Such control issues are making some of the companies worrisome about the technology. 
  • Security: The biggest criticism against cloud computing is security issues (http://www.eecs.berkeley.edu/~elaines/docs/ccsw.pdf ). If a hacker breaks through the firewall of the cloud, it would gain access to all of client data compared to a limited amount of data that is stored in a PC. Thus a hacker finds cloud more attractive than a PC. 
While such issues with cloud computing are legitimate, in my opinion they would be overcome in future. The industry is still in its nascent state and just like other emerging technologies, it is still discovering the problems that would affect the users. Such problems can be revealed only through usage over time and solutions would be developed accordingly. Some of most intelligent engineers and researchers all over the world are working to improve the technology to realize its full potential. With emerging problems will evolve innovative solution. An article by Chow, R. et al (http://www.eecs.berkeley.edu/~elaines/docs/ccsw.pdf) describes how information-centric security can eliminate many of the risks. Such an approach requires the data to be" self-describing and defending, regardless of its environment" and thus would not depend on third party or providers to ensures its integrity and security. 

Some of the issues have their origin in the attitude of users. Users need to trust the providers completely to adopt a particular cloud service. Many experts think, such trust issues will be resolved over time simply because providers would strive to ensure customer satisfaction. As the cloud industry matures, the intentions and service quality of the providers would be transparent enough for users to trust the providers' integrity. Take a look at the video below to see how experts feel about the trust issue:

As Forbes points out (http://www.forbes.com/sites/tomgillis/2011/05/24/criticism-abounds-but-cloud-computing-is-here-to-stay/), the benefits of cloud computing is overwhelmingly more than the cost associated with the risk of cloud computing. Some isolated incidents such as the one with Amazon's EC2 is not intentional and inherent. Emerging technologies always face problem in the beginning because its new and untested. But over time and with increased usage such problems are usually detected and solved. 

Let me conclude this post by discussing why cloud computing could emerge as a marketable technology now, even though the concept of cloud computing existed since the 1960s. It has to do with emergence of supporting technologies and customer readiness. First of all, companies in all industry sectors uses information and communication technologies (ICT) either as their main business or to support their main lines of businesses. Growing dependence on ICT makes ICT expenditure a significant cost driver. Now companies are interested in cloud computing because of its potential as an ICT cost reducer. Increasing usage of smartphones and tablets increases demand of cloud computing among individual users as well. Thus there are enough demand in the market for cloud computing. Secondly, the success of cloud computing depends on several supporting technologies. Realization of high storage capacity with smaller sized chips, availability of high speed internet all over the world and power of virtualization enable cloud computing to be a sustainable and successful technology in this current era. 

My next series of posts will be about the particular benefits of cloud computing relevant to a specific industry. I will start with the business model of cloud computing service providers and later will move on to various user industries such as Financial Services, Communication, Internet Content Providers, Social Media and Healthcare.