International Games System Co., Ltd.
– Arcade Game Division Assistant Manager Paddy Yang
Introduction of the company
International Games System (IGS) was founded in 1991. Soon after establishment, owing to the vast commercial gaming market, IGS aspired to develop its own brand and key technologies, to break the long-term monopoly of Japanese gaming giants, and to gain popularity among Chinese ethnic groups. While the commercial gaming industry was booming, IGS completed the development of exclusive 2D gaming consoles in 1996. 2 years later, it was first recognized by CommonWealth Magazine as one of the Top 1000 Manufacturers and became the largest game developer in Greater China. In the era of cross-platform development from 2000 to 2003, IGS held onto the business strategy of ‘Focus on the essential and autonomous developments’. It has achieved major breakthroughs in personal computer games, physical gaming consoles, PolyGame Master, online games, touch screen games, the Atomiswave, and other game development platforms and technologies. With successive excellent results, IGS has become an all-around game developer in Greater China.
Since 2004, IGS has bolstered its own brand and autonomous developments. It has expanded internationally and has hoped to attract more talents. In 2006, it was formally listed on the stock exchange and laid down 3 business goals: firstly, to have the largest market share of simulators worldwide; secondly, to become the best provider of online games; lastly, to build the best team for game development. The long-term objective is to meet international and world-class standards through overseas deployment.
The challenges facing the enterprise
More than 3 decades ago, IGS started its business by manufacturing electronic circuit boards and gaming console boards as well as by providing hardware maintenance. 3-5 years after foundation, it commenced autonomous developments in gambling, mah-jong, and fighting games. It entered the market in mainland China as early as 20 years ago. Few game developers had the capacity of hardware at the time. With its capability of advanced developments, IGS launched the first fighting game that was solely developed by Chinese and was cheaper than its Japanese competitors. After that, it moved on to the European and American markets as the US and Italy developed gambling products. In the meantime, entertainment games, such as 3D, rhythm and dance machines, were developed in the mainland. About a decade ago, IGS began developing online games. In addition to the pre-existing office of gaming consoles, the office of online games was set up. Over the past 5 years, all the offices have started developing App games as driven by the market demand. The office of gaming consoles went beyond Taiwan and targeted the global market once it introduced App games. It was therefore in search of a solution that was compatible with global deployment and flexible resource allocation.
With regard to the IDC hosting services previously employed by the online office, different resources were required for online games and App games. Large-scale online games were in need of hardware resources with greater data processing capacity and performance to support more online players. By contrast, not as many resources were necessary for App games, hence less flexibility in costs should physical hosts be purchased. When the grades of hosts were adjusted, the conventional execution processes within IDC were so complicated that promptness and flexible timing was a huge concern in case of emergency. Cloud services became IGS’s first choice as they met the requirements for cost optimization, flexible host adjustments, immediate responses, and so on.
There are a wide variety of projects and proposals as well as data worth analyzing within IGS. From IGS’s point of view, the log of players’ activities, the status of resource utilization, the changes in revenue and number of players together with other data can all allow the gaming platform to get hold of the conditions of events, which in turn inform operational analysis and strategies. In the past, the internal programmers moved the online backend data to AWS, disposed of Mango data, and then presented them in the form of Excel spreadsheets. On most occasions of substantial statistical information, it was only possible to present the data on the next day, hence the low degree of promptness. It was rather difficult to make timely responses to the status of events so a highly efficient tool for operational analysis that could be visualized was deemed absolutely necessary.
Before formal subscription, IGS evaluated the Cloud services provided by various domestic and overseas companies. Some of them could only offer a low degree of autonomy while other facilities and service adjustments were dependent on the staff’s capacity of maintenance. IGS eventually opted for Amazon Web Services (AWS) for the reasons below:
- AWS developed early and has become the leading provider of Cloud services, thereby offering more diverse supporting services.
- AWS has been more open and transparent than other providers of Cloud services in terms of resource allocation and pricing.
- AWS has comprised a wide range of integrating services with higher flexibility. It has been able to offer comprehensive cost estimates for the benefit of deployment.
IGS has subscribed to the following AWS services and products:
- Its own database was built with the help of the integrating Amazon EC2. It has also flexibly mounted the fundamental hardware to bring about the advantages of cost flexibility and high usability.
- IGS has regarded AWS EBS services as the cornerstone of capacity and backup mechanisms as the requirements for uninterrupted, encrypted, and highly usable services have been met.
- Amazon Route 53 has been used to manage the separation of domains.
- Amazon S3 and Amazon CloudFront CDN have been used for backend backps and front-end data deployment.
- In the future Amazon ElastiCache will be introduced to IGS so as to speed up data access.
IGS first came across Splunk while attending a Nextlink’s seminar on data monitoring and BI module sharing. It subscribed to Splunk after just a few courses. The staff members responsible for the project presented the data, which other colleagues would like to analyze, through dashboards. Splunk’s quick function of chart creation helped IGS promptly grasp the data. Through Splunk’s dashboards, the machine behavior was monitored and the players’ data, for instance, the extent of participation in events, and the status of spending on games, was observed. Incidents could be instinctively represented in a visualized manner as soon as they took place. The promptness enabled IGS to make loud and clear strategies. For example, it was estimated to take less than 10 seconds to find an opponent in fights between players but it actually took 5 minutes to do so as shown by the dashboards. The size of deficit in players, the average waiting time, and other relevant data could all effectively help IGS take prompt measures to improve gaming experiences.
The Splunk units used within IGS have so far been dependent on the provision of existing data. They are mainly for resource monitoring for the time being as all kinds of phenomena in the games are monitored. Other data, such as the numbers of players in different countries, and the players’ spending, may be imported later on to assist the sales and other internal units in applications and event budgeting.
- Execution performance and capacity has been raised. The number of projects handled by a single person has increased from 4 to about 10.
- The management processes have been simplified so the hardware workloads have been reduced.
- The time needed for building hardware has been cut from a month to less than a week.
- Costs can be flexibly adjusted in accordance with the practical needs.
- The sources are applicable to plenty of file types.
- Swift chart production can save time and human costs.
- The monitoring function meets the requirement for promptness.
The application products and services
Amazon EC2, Amazon EBS, Amazon Route 53, Amazon S3, Amazon CloudFront, Amazon ElastiCache, Splunk