Unleashing the Full Economic Potential of Cloud Computing in China

Price of hosting in China is about 5 times higher than in Europe if we do not consider Internet connectivity, and 20 times higher if we take into account internet connectivity. 1,800,000 jobs could be created in China by bringing Cloud hosting price to levels similar to Europe.
  • Last Update:2022-05-04
  • Version:001
  • Language:en

Imagine a country in which two companies with a de facto monopoly on the market decide to sell smartphones and computers at a price that is 20 times higher than anywhere else in the world. How fast do you think IT would develop in this country, especially in small and medium businesses? And how many jobs would be created to provide computer related professional services? How many students would own a computer at home and train their IT autonomy beyond what they learn at school? How many companies could purchase very large computers to process big data? It is not unreasonable to state that with such a tremendous price difference, 10 times less jobs would be created compared to other countries and that student education would be much worse that in other countries.

The country we are describing is China. Of course, computers in China are not 20 times more expensive than in other parts of the world. But Cloud Computing services are 20 times more expensive than in France, which is probably the country in the world with the most cost efficient offering.

The Cloud Computing pricing situation in China prevents realising completely the full economic potential of Cloud Computing and prevents the creation of 1.800.000 jobs in the knowledge economy, as we will explain in this article.


Hosting is the most basic service which all other Cloud Computing services depend on. Let us compare hosting prices in France and China.

Hosting price in France and China
France China Price ratio Performance ratio
Cheapest Shared Hosting
11 GB space / 1 Gbps internet 5 GB space / 1 Mbps internet 1.08 tmes more 1000 times slower network
137 RMB / year 149 RMB / year
Personal Server (VM)
1 core / 2 GB RAM / 10 GB SSD / 100 Mbps internet 1 core / 2 GB RAM / 20 GB SAN 3.76 to 231 tmes more similar
30.8 RMB / month 116 RMB / month without Internet
7.136 RMB / month with 100 Mbps Internet
Small Business Server
6 core / 32 GB RAM / 2x250 GB SSD / 300 Mbps internet 4 core / 32 GB RAM / 250 GB SAN 5.5 to 96 times more similar
231 RMB / month 1.273 RMB / month without Internet
22.333 RMB / month with 300 Mbps internet
Big Bare Metal Server
16 core / 128 GB RAM / 1.5 TB SSD / 10 Gbps LAN / 500 Mbps Internet 16 core / 128 GB RAM / 100 GB SSD / 10 Gbps LAN 5 to 23 times more similar
1155 RMB / month 5.832 RMB / month without Internet
26.892 RMB / month with 300 Mbps internet

We are comparing in this table various offers that exist in China, based on bare metal server or virtual machines. From a performance point of view, bare metal is usually better, especially for storage. From a reliability point of view, there is not so much difference between bare metal and virtual machines if we look at actual statistics, as long as RAID or SSD is used for storage. The graph bellow shows the number of unexpected reboots of bare metal and virtual machines over a long period of time.

Unexpected reboots of various hosting infrastructures

This graph shows that the stability of virtual machines is same (SlapOS) or much worse (OpenStack) than the stability of bare metal hosting with RAID or SSD storage (OVH, Online, Soyoustart). Bare metal servers reboot in average once every 5 to 10 years, just as SlapOS virtual machines do under the same conditions. OpenStack virtual machines reboot unexpectedly 1 to 5 times per year, which is actually more than the number of reboots of a non redundant PC located in a home in France, Germany or Japan ("V" of ViFiB).

With this in mind, we could claim that current price of hosting in China is about 5 times higher if we do not consider Internet connectivity, and 20 times higher if we take into account internet connectivity. Internet access should not be underestimated in this benchmarking because it can have a high impact on performance of a Web site or on the kind of applications that can be implemented through Cloud Computing. High internet speed is required for resiliency, data archival, big dataset transfer, video applications, etc. All these applications are becoming essential for enterprise IT.

Macro-economic analysis

In a research paper published by IDC in 2012, the number of jobs created by Cloud Computing for China and India was estimated to be 6.75 million. Most of these jobs depend on cost reductions offered by Cloud Computing in enterprise IT and productivity increase. If we suppose that 3.5 million jobs are related to China out of the 6.75 million of both China and India, and if we consider that half of these jobs have not been created in China because of the high price of Cloud hosting, we can conclude that 1.75 million jobs could still be created through a price cut in Cloud Computing hosting in China.

This is a very conservative calculation. We are taking figures of 2015, not of 2018. We are applying a ratio of 50% economic development slowdown, not of 90% as in our introduction. In reality, it is much likely that even more jobs could be created than the estimated 1.75 million. Let us confirm this figure through a different calculation.

Micro-economic analysis

We are going to try to compute how many jobs could be created if the money that is currently "wasted" in expensive Chinese Cloud Computing companies (Aliyun, UCloud, Qingcloud) was instead invested into the knowledge economy by creating value added jobs which themselves accelerate economic growth and create further jobs through leverage effects. In other words, spending money on mere infrastructure (network, computer, hosting) does not bring per se any added value to a company. However, spending money on graphic designer, software developers, data scientists, professional translators, etc. produces knowledge which creates immediate value and beyond can produce innovation that leads to new products and new jobs. This is the difference between an infrastructure economy (as China has been for a few decades) and a knowledge economy (as Europe, USA and Japan have become and as China is now becoming).

Here is the table we use to compute jobs that could be created if Cloud Computing hosting was cheaper in China:

Jobs that could be created in China with lower Cloud hosting price
Item France China
Hosting Price including Internet
Personal server hosting monthly price (RMB) 30 200
Ecommerce server hosting monthly price (RMB) 77 500
ERP and e-business server hosting monthly price (RMB) 231 2,000
Big Data Server monthly hosting price (RMB) 1,155 10,000
Market segments
Number of students that could be trained to manage a hosted server 1,000,000 10,000,000
Companies that could control their own e-commerce site 1,000,000 10,000,000
Companies that could host their ERP and move to e-business 250,000 2,500,000
Companies that need Big Data 25,000 250,000
Competivity lost per segment due to high price of Cloud hosting in China
Lost competivity for education (RMB/year) N/A 20,400,000,000
Lost competivity for e-commerce (RMB/year) 50,760,000,000
Lost competivity for e-business (RMB/year) 53,070,000,000
Lost competivity for big data (RMB/year) 26,535,000,000
Lost competivity total 150,765,000,000
Job Impact
Typical employer cost of a job in the knowledge economy N/A 500,000
Number of direct jobs lost due to high price of Cloud hosting 301,530
Jobs created indirectly by knowledge economy jobs after 3 years 1,507,650
Total job creation impact 1,809,180

We first consider four cases:

  • students that could be trained to manage a small Cloud server online based on Linux;
  • e-commerce sites that could be managed by small companies without depending on a marketplace (Taobao, TMall, etc.);
  • ERP servers that could let small companies implement e-business and other innovative processes;
  • Big data servers that could be used by mid size companies for data science and AI.

For each of these cases, we compare the price in France and in China of the Cloud hosting infrastructure. The price we use in China contains some bandwidth, not equal to zero but not necessarily as much as in France. It is thus set to a value in between the two values that were benchmarked previously.

We then evaluate the market size for each case and consider a one to ten ratio between market size in France and in China. Current GDP ratio between France and China is about 5. Population ratio is about 20. We thus believe that a ratio of 10 is reasonable.

For each market segment, we compute how much competitiveness is lost in Chinese economy due to high price of Cloud hosting. For example, if small and medium size business in China were running their own e-commerce site as in France rather than depend on a marketplace, 10 million companies in China would be spending 500 RMB per month for Cloud hosting instead of 77 RMB in France. 433 RMB is being lost in terms of competitiveness. In other words, 433 RMB per month is paid to Cloud infrastructure that should never cost more than 77 RMB per month and that actually creates nearly no jobs due to automation, rather than being paid to graphic designers, software developers, data scientists, etc. This difference of 433 RMB multiplied by 12 months and by 10 million small businesses represents more than 50 billion RMB per year that are lost for the knowledge economy.

Let us go into some more details of our reasoning. Suppose that we tell a small business owner in China that for 500 RMB per month he gets an e-commerce site and nothing else, and that he needs to pay an extra 433 RMB per month for operating the e-commerce site, it is much likely that not a single small business owner will be interested. However, if we tell the same business owner that for 500 RMB per month he gets both the e-commerce site running and someone to operate it and customize it to his needs with possibly some innovations, there is a high chance that this offer will sound more interesting than losing more than 10% of its profits to a marketplace such as Taobao or TMall. Actually, monthly online sales of 5000 RMB are sufficient to justify the implementation of such dedicated e-commerce system.

By lowering the price on Cloud hosting, many creative and knowledge services suddenly become attractive to small business owners, which then leads to job creation in the knowledge economy.

If we sum up all the four segments, we reach a total of 150 billion RMB lost competitiveness every year in China due to high prices of Cloud Computing hosting. If we consider that the yearly employer cost of creative or knowledge related jobs is about 500,000 RMB - which translates into 250,000 RMB net salary after social expenses have been paid - the number of direct jobs that could be created by lower cloud hosting price in China is about 300,000. This also means that lowering cloud hosting prices in China creates 75 billion RMB additional income for local governments, consisting of social benefits and labor taxes.

Those 300,000 jobs that are the direct consequence of lowering Cloud hosting prices are not usual jobs: they are creative jobs. They lead to innovations, to new products, to translating a web site to other languages, to supply-chain optimization, etc. All this leads to additional income for each company, growth and further jobs. We consider that a ratio of 5 could apply to indirect effects of those direct 300.000 jobs, based on figures that have been calculated on the economic impact of Internet by McKinsey & Company (direct impact represents 15% to 25% of total job creation for an IT innovation).

About 1,500,000 additional traditional jobs should thus be created after three years in addition to the 300,000 creative jobs that could be created immediately by saving on Cloud hosting price. 1,800,000 jobs could thus be created in China by bringing Cloud hosting price to levels similar to Europe, consistently with previous evaluation.

Efficiency of subsidies for job creation

Sooner or later, China will experience drastic price cuts in Cloud hosting, initiated by new players. A price war will then follow, just like it happened in France with the introduction of the company called "Free" in mobile telecommunications. It will take about 5 years for market to stabilize and reach a point of reasonable prices (10 times less than today) and reasonable profits (15%).

The role of government in Cloud hosting in China is essential. Without its support, there is no way to change the current market. Why would then government in China support low cost cloud. Based on the previous analysis, reasons are quite clear:

  • for every million RMB of additional income in low cost cloud, 18 creative jobs are created immediately and 90 traditional jobs created indirectly within 3 years ;
  • for every million RMB invested in low cost cloud, 17 million income can be generated, 300 creative jobs created within 5 years and 1500 traditional jobs created indirectly within 8 years.

It is thus quite clear why government in China should urgently support the formation of low cost Cloud industry in order to stop the current waste of money in noncompetitive cloud offerings that prevent the development of the knowledge economy through creative jobs.


  • Photo Jean-Paul Smets
  • Logo Nexedi
  • Jean-Paul Smets
  • jp (at) rapid (dot) space
  • Jean-Paul Smets is the founder and CEO of Nexedi. After graduating in mathematics and computer science at ENS (Paris), he started his career as a civil servant at the French Ministry of Economy. He then left government to start a small company called “Nexedi” where he developed his first Free Software, an Enterprise Resource Planning (ERP) designed to manage the production of swimsuits in the not-so-warm but friendly north of France. ERP5 was born. In parallel, he led with Hartmut Pilch (FFII) the successful campaign to protect software innovation against the dangers of software patents. The campaign eventually succeeeded by rallying more than 100.000 supporters and thousands of CEOs of European software companies (both open source and proprietary). The Proposed directive on the patentability of computer-implemented inventions was rejected on 6 July 2005 by the European Parliament by an overwhelming majority of 648 to 14 votes, showing how small companies can together in Europe defeat the powerful lobbying of large corporations. Since then, he has helped Nexedi to grow either organically or by investing in new ventures led by bright entrepreneurs.
  • Photo Thomas Gambier
  • Logo Nexedi
  • Thomas Gambier
  • thomas (dot) gambier (at) nexedi (dot) com
  • Software Developer
  • Photo Sven Franck
  • Logo Nexedi
  • Sven Franck
  • sven (dot) franck (at) nexedi (dot) com
  • Photo Ni Yan
  • Logo Nexedi
  • Ni Yan
  • ni (dot) yan (at) nexedi (dot) com