Where’s enterprise cloud adoption heading in Greater China?

Cloud computing

It’s hard to believe cloud computing has been around for more than a decade during which it has matured into a mainstream technology, although enterprises in the region are still to fully leverage its potential.

The cloud has accelerated IT enablement and business innovation, reducing TCO, increasing agility and enabling IT to focus on generating business value. However, while going straight to the cloud is a no-brainer for start-ups, cloud migration is inevitably less than straightforward for well-established enterprises.

Short-term strategies

The issues enterprises face include: the uneven maturity of cloud solutions; cloud security; integration with legacy applications; evolving regulations in multiple jurisdictions; and a shortage of know-how. To date, leading enterprises in Greater China have pursued a couple of different strategies – which are by no means mutually exclusive – to leverage cloud opportunities.

One is to pursue simplification and “engineered” cost savings by migrating applications to the cloud. While levels of adoption vary, we typically see clients in Greater China looking at a “hybrid cloud architecture” end-state spanning private cloud and public cloud IaaS/Paas/SaaS solutions.

The other strategy has been to pursue business innovation via cloud-driven applications. Many enterprises in the region have created “systems of engagement” – typically including mobile apps – based on pure-cloud solutions. To avoid being slowed down by more stable in-house IT “systems of record” they have kept these applications separate where possible, resulting in the “multi-speed IT” phenomenon.

Legacy holdup

Both approaches harvest low-hanging fruit without tackling the transformation of legacy systems that support core business functions. Until this is addressed enterprises face diminishing returns on their cloud strategies and will be stymied from establishing truly differentiated new business models that combine valuable existing data, insights and functionality from legacy applications and exploit fast-moving technology developments such as APIs, IoT and Blockchain.

While mulling this challenge, enterprises must contend with the uncertainty of regulatory change and bewildering service provider choice and variation. For example, the monetary authorities in Hong Kong and Singapore are now more accepting of public cloud hosting while China, Indonesia and Korea insist on maintaining data sovereignty. On the service provider front the PaaS industry is at such an early stage that there are countless options and, while IaaS is more mature, there are still plenty of choices.

Containers and microservices

So how can enterprises extract more cloud value? One very promising avenue, which neatly de-risks regulatory and service provider uncertainty, is application containerization. To put this simply, containers do for applications what virtualization has done for infrastructure. Containers are lightweight and cloud-native by nature, so can easily be moved between on-premises systems and cloud platforms.

Application developers are using containerization to break down monolithic applications into discrete business-aligned “microservices”. For instance, an application with billing and credit-check functionality can be broken up into two microservices, which are more robust and resilient. During a large month-end run, the billing microservice can scale out independently from credit-check without restarting the entire application. This architecture also allows the billing and credit teams to develop software at their own pace.

While there’s been much recent work in this area – notably around standardization – the container concept is by no means new and has been proven at truly industrial scale by the likes of Google ad Facebook. Enterprise adoption – for instance AT&T’s microservices transformation in collaboration with IBM – is more recent but fast-moving.

In a six-month project with a Hong Kong public sector client we were able to completely re-architect one of its core application to run in containers. Among the benefits, the client has been able to reduce its upgrade window from seven hours to one and is now positioned to increase its pace of innovation, with the go-ahead being given to containerize more legacy apps.

Moving ahead

Particularly in China, we’re working with highly-competitive Internet, telecommunications and financial-services companies of such scale that building private clouds and innovating on microservices architecture is picking up momentum. Yet enterprise microservice deployments are still at an early stage and one of the key issues is that, by nature, microservices architecture breaks applications into business domains. If software development is not aligned to this then enterprises will find managing microservices applications increasingly difficult as they hit development pipeline bottlenecks.

In competitive and fast-changing industries, the enterprises that enjoy the greatest success in exploiting microservices to drive rapid innovation will also be the ones most adept at re-inventing their operations and culture to align to DevOps best practices. And, as always when sailing into new waters, it pays to have an experienced navigator on board. IBM has the rich experience to help customers on this journey.


Daniel Cheung

Daniel Cheung, Cloud Leader, Global Business Services, IBM Hong Kong


「談雲說 IT」第四回:雲與 IT 即服務

(文章亦同時被 Unwire Pro引用)


今次為之前的三回文章來個總結。當大家想到「IT 即服務」(IT-as-a-Service),說的不單是雲系統或個別服務;所有可成為服務的 IT 功能,都應包含在內。例如啟動 (provisioning) 儲存裝置、啟動網絡/虛擬局域網段 (vLAN segment)、啟動裸機伺服器 (bare metal server) 等。

Continue reading

「談雲說 IT」第三回:中介服務

(文章亦同時被 Unwire Pro引用)


過去兩年來,當我們和眾多企業的 CxO 級主管交談,一個道理呼之欲出:世上沒有放諸四海通用的雲系統。這些主管往往要和爛 IT 搏鬥,因 IT 情況逐漸失控而感到沮喪。普遍觀察所得:1. 雲技術是個進程,不可能瞬間完成;2. 許多企業極需制訂混合雲策略。今次,讓我們探討雲端中介服務 (Cloud Broker Service) 為何重要。


  1. 雲端平台選擇眾多,企業漸多採用混合雲策略。雲端的選擇,包括服務供應商﹝例如IBM Cloud、AWS、Azure等﹞,及多種平台部署﹝IaaS、PaaS、SaaS等﹞。
  2. 非IT業務範圍 (LOB) 的企業,逐漸加入購買IT產品和服務。
  3. 令不少企業煩惱的問題是,舊有負荷 (唯讀應用) 該怎樣存放,新負荷又該怎樣存放?

Continue reading

古迹結合科技 傳承文化瑰寶



雲運算與古蹟名勝,是否風馬牛不相及?步進意大利的米蘭大教堂(Duomo Di Milano),第一印象是135座歌德式尖塔、逾3,500座聖人和動物雕像和大理石十字座堂的彩色花格玻璃窗。搭電梯或走上去,可登頂俯瞰整個米蘭,鴿子在大教堂廣場飛翔,遠眺見到阿爾卑斯山風光。

米蘭大教堂的建築工程歷時六百年,由大教堂的聖工社(Veneranda Fabbrica del Duomo di Milano)負責施工,揉合歷代建築風格,到1950年才告完成,規格和工藝超越了絕大部分的歐洲教堂。

Continue reading

「談雲說 IT」第二回:協同服務

(文章亦同時被 Unwire Pro引用)


上回筆者介紹過企業資訊科技中的混合雲 (Hybrid cloud),本篇則探討協同服務 (Orchestrator) 的重要性。首先澄清,無論公共或私有的雲端系統,都可分為不同的雲端部署模式(deployment models)。下圖是基礎設施即服務 (Infrastructure as a service, IaaS)、平台即服務 (Platform as a service, PaaS),軟件即服務 (Software as a service, SaaS) 的對比,及它們跟傳統 IT 有何分別。

Continue reading

自動售賣機 AI推廣更貼心

(文章於2017年9月16日在 iMoney刊登)


不論是大品牌或小商戶,要在香港開店,租金都是一大問題,因此近年有售賣各種貨品的自動售賣機出現,「其實我不太鍾意自動售賣機這個名字,因為立即聯想汽水機,其實我是想創造Automatic Retail(全自動零售)的新方向。」SmartRetail創辦人蘇俊豐在2009年創立公司,一改自動售賣機只賣汽水、零食的形象,變成賣健康零食,並配合機上屏幕做不同宣傳活動。貫徹健康的主題,最近他與IBM合作,採用對方開發的人工智能技術,售賣Fitbit產品。客人走近售賣機,機上的前置鏡頭能自動偵測其性別及年齡範圍,介紹相應的商品,客人可即時以信用卡、八達通等電子貨幣付款。配合不同商戶的市場推廣需要,屏幕可有相應的用途,例如客人可選擇觀看廣告以獲取折扣等。

Continue reading

商界夥學界 推動創科研發應用



IBM 與 MIT 將共同建立新型 AI 研究實驗室,這也是歴來最大宗的商界學界研究合作項目之一。


回想 2005 年,當時 IBM 與 Google 和學界合作,共同投資於一種嶄新互聯網運算模式及培訓相關人材。項目投資額為四千萬美元,用於建設數據中心及開設教學課程。教材源自 IBM 在數據中心領域上的知識、及 Google 在開發互聯網應用的經驗。項目於 2007 年擴展至包括麻省理工(MIT)及哈佛等多家學府,亦促成了今天的雲運算技術。

Continue reading