By including Zing, Cinnober improves overall response time predictability in its trading, clearing and risk management solutions. It will also result in greatly reduced JVM tuning requirements, which enables faster deliveries of new capabilities.

Collaboration with cutting-edge partners such as Azul is considered an important part of Cinnober's own research and development. Implementing Zing complements the Swedish company's innovation with orchestrated memory management, achieving pause-free execution in servers that are critical for low response times.

"Our TRADExpress Ultra technology already enables trading and clearing systems with consistent response times and the lowest latency available, at single-digit microseconds," says Veronica Augustsson, CEO of Cinnober. "Now Azul's supplementary real-time JVM can provide every one of our offerings with consistent response time. Zing is an outstanding JVM with regard to predictability in the demanding domain of financial IT. We look forward to working together with Azul on further developing Java for financial IT - improving latency and adding features, functionality and add-ons faster to build competitive advantages."

Cinnober's solutions benefit from Java's speed and flexibility as well as its rich feature set and its ability to enhance developer productivity. With Zing, Cinnober's customers can eliminate latency outliers and deliver smooth scalability as order rates, in-memory data set sizes or user loads grow. Developers can also utilize all available Java libraries and functions.

Scott Sellers, Azul Systems President and CEO, says: "Competition in capital markets is fierce today and Zing is the only JVM that solves long-standing Java performance and operational issues like persistent latency outliers, Java's 'warm-up' problems, and burdensome tuning requirements. Cinnober's choice illustrates why Zing has become the new performance standard for Java-based applications within the financial sector."

Zing can be implemented with TRADExpress solutions simply by replacing a system's current JVM - without recompiling or re-architecting. This allows the system to meet the most stringent service level agreements with regard to predictability and eliminating latency outliers.