aiScaler Ltd. is a multinational software company founded in 2008. It develops application delivery controllers designed to allow dynamic web pages to scale content by intelligently caching frequently requested content. A number of websites in the Alexa top 1000 use aiScaler to manage their traffic.^[1][2]

aiScaler software can be deployed either on public cloud computing platforms such as Amazon Web Services^[3] or private virtual environments. aiScaler software is considered an edge device as it proxies traffic, augmenting or replacing content delivery networks endpoints.^[4]

History

aiScaler started as a project in 1994 by the web development company WBS. The project was called "Jxel", short for Java Accelerator. The technology was Java-based and intended to be run on a Java Virtual Machine sharing the same computer system as the HTTP server. It was re-written in 2009 using the C computer language, occupying its own dedicated server. The new software was rewritten to run on Linux only, taking advantage of changes in the input/output model based on epoll. In July 2008, aiScaler Ltd acquired all technology of WBS for $3.8 million.^{[citation needed]}

Until 2013, aiScaler was known as "aiCache", producing a product called aiScaler. The company took over the name of its main product, phasing out the brand name aiCache.^[5]

Products

All aiScaler products can be categorized as Application Delivery Controllers

• aiScaler is an HTTP accelerator that provides application delivery control, in addition to scaling and acceleration of content delivery ^[6]
• aiProtect offers protection against DDoS attacks and SQL injections^[7]
• aiMobile is a Mobile content management system
• aiCDN is a cloud-based Application Delivery Network that allows scaling of dynamic web applications.
• aiScaler and Dell offer a hardware Application Delivery Controller, which fits in a standard rack unit server rack.^[8]

aiScaler is based on epoll technology allowing it to employ a right-threaded (only the specified number of workers process requests, no matter how many clients are connected), non-blocking, multiplexed IO design.^[9]