Cloud Computing And China: Two Sides Of A Digital Bangladesh

What can the current digital Bangladesh policy, an attempt to improve connectivity, information access and literacy, achieve for Bangladesh? While the policy is well intended, in Bangladesh, like anywhere else in the world, a government can merely provide an environment for economic development. Individuals and businesses subsequently have to take advantage of emerging opportunities, some explicit and others incidental, hopefully effecting a (economic) benefit beyond themselves and for the whole of the country. Will this be the case for the digital Bangladesh? Imagine, if you will, this policy succeeds and opportunities arise from a better connected and educated population. What would be the obvious and the hidden opportunities and will Bangladeshis exploit them?

China and Apple

You may think, a digital Bangladesh will naturally continue on its trajectory, competing in the global labor market. The change in industry would probably be from garment factories to electronics assembly lines. Initially this may sound like a desirable development and improvement. We can infer the implications of this change for Bangladesh by examining the relationship between the most valuable company in the world, Apple, and the most populous country in the world, China. They occupy the opposite ends of the value chain. Apple designs, markets, and sells products while China produces and assembles them.  (China in the following comparison is used as a synonym for the numerous companies and workers in China involved in the production of goods)

So who benefits from the often hundreds or thousands of dollars a consumer spends on an Apple product? Apple keeps a large part while the other significant part is spend on manufacturing the product, most of it usually in China (each party receiving somewhere between 20-60% depending on the device). Unsurprisingly, Apple retains more of its share as an economic benefit than China. The asymmetry of the relationship is stunning nevertheless. China spends nearly all of its share on materials, a small part on labor, and retains a tiny fraction as profit. Apple on the other hand turns a significant part of its share into profits. Recent documents from a patent trial against Samsung revealed Apple’s gross margins on the iPhone and iPad to be 49% to 58% and 23% to 32% respectively.

The economic inequality continues with the effect it has on labor. Apple, as we would expect, pays its US workers, highly skilled engineers, designers, and so forth, more than the Chinese the manufacturing labourers. The imbalance is so great that the comparatively small number of Apple employees together earn more than the vast armies of assembly line labourers. You may postulate that the chinese workers in spite of this are benefiting by earning a living. Unfortunately, a never ending stream of news of workers’ perils employed by companies like Foxconn, incidentally a manufacturer of Apple products, has highlighted these workers’ struggle. They face poor working conditions, exhausting work loads, while earning meagre salaries. China is undertaking great efforts on climbing the value chain with a lot of wooing from foreign investors and governments. It is indeed advancing but not for all Chinese (partly because of its currency policy) and it is a slow progress despite its global economic significance. Bangladesh, a comparative lightweight, can not hope to get the same international attention and will have to rely more on its own strengths.

Inevitable or improbable

Do we want Bangladesh to solely continue on the seemingly inevitable cheap labor path, aspiring to be a tardy copy of the Chinese model, increasing national revenue but hardly improving its workers’ lives? Should it be contempt to dream of dull factories and kilometres of conveyor belts manned with Bangladeshis hardly making a living? Should Bangladeshis, or at least some of them, not attempt the improbable, bypass years of labour, and capture a share of the high end of the value chain? Should Bangladesh not grow a competitive, wealthy middle class fuelling local services and the employment it entails? Such a development would greatly accelerate the economic development of the country as a whole.

I am not proposing to Bangladeshis to copy the Apple business model and design tablets and phones in Dhaka. This is a hotly contested market which is crowded by the most potent technology companies in the world. There are, however, opportunities at the high end of the value chain for ingenious and industrious individuals and businesses, Bangladesh’s best and brightest. Many of the opportunities arouse from a technological revolution of the last years. It makes it more important for an individual or organisation to be creative, fast and adaptable. Most importantly to get a business off the ground it removed the need for large capital investments, favourable trading agreements with developed countries, to be located in a developed country, or even to have natural resources. How is that possible?

Cloud Computing

The answer lies with cloud computing, a technological paradigm shift emerging from the internet’s rapid growth over the last two decades. Cloud computing combined with a digital Bangladesh, connectivity and computer literacy, bridges the digital gap enough to empower Bangladeshi. They can now compete in the digital space and build the next hot IT startups potentially growing them to billion dollar companies. To understand how cloud computing empowers third world entrepreneurs we have to understand what it is.

The internet was the third revolution in information distribution in mankind’s history after the printing press and the alphabet. Information and subsequent electronic services are becoming ubiquitous, accessible anywhere to anyone at any time. This development affected incredible changes of how we process and store information. In the beginning of the internet, for example, only a few organisations exchanged emails, served web pages, hosted discussion groups and so forth. They were able to provide these services with comparatively simple software and a few computers. After all, only a fraction of the thousands of initial internet users were active at a time sending around mostly text based information. Compare this to today. Approximately 2 billion people are online, nearly all the time, with popular services providing realtime video streaming of high definition video and audio content, dwarfing the initial text content the internet was designed for.

The technology to provide these services affordably and at such scale were not even imagined a few years ago. A key invention and economic development that enabled this is cloud computing. It is the commoditisation and globalisation of computing resources and electronic services. Until very recently, if an organisation wanted to store and process large amounts of data it had to buy computers and either build or rent space in a nearby data centre, a warehouse like complex with industrial sized air-conditioning, generators, redundant fibre network connections, security services, and so forth. The capital expenditure for these and the fast depreciation of the computers housed plus the complexity and expenditure of staff monitoring and managing the computers quickly ran into the millions of dollars. Not something affordable for startups, individuals or for business cases where these resources are utilised only intermittently. However, in the last years cloud computing changed this radically.

Cloud computing allows anyone to rent inexpensive, virtual computing resource of various sizes and shapes anywhere in the world on short (as little as an hour) or long term (years). All you need is a computer, internet connection, and credit card to access storage and computing power greater than what organisations had access to just a few years ago and in many cases still have not. You do not actually rent a physical machine or storage. The cloud service provider standardised these into abstract commodities. For example, if you rent a computing resource it likely will not reflect a physical machine in a data centre. It may be one of many virtual computers hosted on a larger more powerful and cost-effective physical computer. What you rent is the virtual equivalent of a small physical machine. For all of your purposes the virtual machine will be indistinguishable from a physical one of the computing capacity you requested.

Google data center

Google data center

The data centres of your cloud provider are likely geographically distributed so you can rent computing resources close to your customers to be responsive due to low network delays. Additional resource can be acquired in another distant location to make your service reliable, i.e. if a natural disaster, network or power outage hits the East coast of the USA you can fall back to your rented resources on the West coast, Europe or Asia. With some services like data storage you don’t even have to worry about locality and your cloud computing provider will guarantee you certain properties like redundancy and speed, and manage the global distribution and duplication of your data for the level of speed and availability you order. While employing cloud computing requires great skill, the manpower to manage a virtual computer centre in the cloud is significantly less than a physical one which reduces costs and investments further. As a result of this companies shift workloads from traditional data centres to cloud computing services. The recent Global Cloud Index by Cisco predicts that by 2016 two-thirds of data centre traffic will originated from cloud computing and 60% of the data centre work load will be handled by cloud computing.

What are some examples of this development? Say you are a student who wants to churn some time series data for a research project making some predictions. You could run one prediction on your computer for a few days and hope to get a good result. Alternatively you could rent a dozen virtual computers for a few hours costing you as little as a dollar and get the result faster. Or how about computing variations of the prediction on a few dozen virtual computers in parallel for a few dollars? Another example, you are a bio-engineering company which needs large scale computing power of thousands of computers to calculate protein folding for a new product. You can rent the computers needed for potentially less than a hundred dollars an hour and once you have your result shut them down without any further financial commitments. A final and common example are startups which regularly utilise cloud computing. They need IT infrastructure, i.e. databases, storage, computing resource, web servers, and so on. They have little capital and do not know what their business model will be in six months or how many customers they may have. Until recently a very impossible situation to resolve. Today, it has become common and successful, even best practice, to rent resources according to the current requirements and add, remove, up- and downgrade resources ad-hoc as needed.

Amazon sells more than you think

The biggest cloud computing provider today is Amazon. If you thought Amazon is an online retailer that understands technology think again. Amazon, in reality, is a technology if not cloud computing company that just happens to do retail too. Amazon offers what it calls Amazon Web Services which is an ever increasing range of rentable computing services covering nearly all possible IT needs of an organisation around computing, storage, databases, applications, networking, and much more. Make no mistake, this is a business with a profit margin for Amazon and if you have an exact, significant, predictable and steady need of a computing or storage service you likely are cheaper off providing the resource yourself. What most organisations have recognised though is that this is an exceptional scenario. Most organisations have peak times, seasons, and frequently uncertain future requirements. They benefit from the ability to adjust computing resources with their needs.

Netflix, the popular online TV and movie streaming service, exemplifies how cloud computing can shape a company and give it a competitive edge. Their service has 31.8 million subscribers in 51 countries with approximately 25 millions in the US. Netflix’s computing need is highly volatile by time and geography, i.e. not many people watch TV at 3 am Monday morning but nearly every household watches TV on Sunday night. Netflix is big, they accounted for 32.7% of downstream traffic in the USA in peak times in 2011. A tremendous amount of data traffic. Amazingly, they provide nearly all their services through cloud computing and of that nearly all through Amazon. I had the pleasure of attending a talk by Adrian Cockcroft, Cloud Architect at Netflix, and he described their ability to raise, shrink, grow and destroy large clusters of computing resources anywhere in the world (in one of the many Amazon data centres) depending on demand. This would have been considered science fiction not long ago. Yet, there they are, running a successful business with tens of millions of subscribers without owning the infrastructure on which everything depends. It enabled them to generated a year on year turnover of $905 million and a gross margin of 26.8% with a market capitalisation of around $3.8 billion. It is hard to imagine a service like Netflix being globally present, fast growing, affordable, and profitable if the computing resources needed at peak time could not be relinquished back to the cloud service provider at low demand time to save money.

Digital Bangladesh should be a Silicon Bangladesh

So how does cloud computing impact Bangladesh’s chances for a better future? The lesson for Bangladesh is that the digital gap just got a lot smaller. The barrier of large capital expenditure for businesses providing services utilising computing resources has vanished. Startups around the world are taking advantage of this fact and thoroughly rely on cloud computing. Most importantly, all the resources needed are available to anyone with an internet connection and minimal funds.

As a result, an entrepreneur in the digital space is purely limited by her imagination, knowledge, and experience with capital and locality becoming insignificant. As a result less is differentiating a digital entrepreneur sitting in a coffee shop in Dhaka from one in Silicon Valley today than it was yesterday. Cloud computing levels the playing field enough to pose an opportunity for a Digital Bangladesh or better a Silicon Bangladesh.

One last hurdle, the knowledge gap, remains to be overcome. An internet entrepreneur in the developed world will have the benefit of a better education and will have been submersed in a life with technology. Bangladesh, even with a successful digital Bangladesh policy, will not bridge this gap completely any time soon. Hopefully it will give its citizens tools and abilities to close the gap themselves though. What it will take to realise the opportunity and overcome this hurdle is a new breed of Bangladeshi’s, the next generation. Young people have a passion to challenge the status quo, to overthrow a paradigm, and they believe in their ability to shape a better tomorrow. Their ignorance of the probable and inevitable, to be a factory worker or an outsourced web designer, allows them to be more, the improbable, digital entrepreneurs. They can bridge the knowledge gap by educating themselves utilising the online education revolution and unleash their ideas leveraging opportunities like cloud computing. I hope digital Bangladesh will succeed and I hope it aspires to more than assembly lines.

This article was written as part of a regular column by Christian Prokopp for the daily online newspaper.

Leave a Reply




This site uses Akismet to reduce spam. Learn how your comment data is processed.