Case studies

High Performance Stream Processing Platform for Financial Services
Many financial institutions have built automated trading systems that consume high volume market data feeds. Their business success depends on millisecond level responses to approximately a half million messages per second. They typically have a home-grown purpose-built solution with hundreds of servers. The problem: Market data providers are projecting more than 5x traffic increase during the next two years. Their current systems cannot cost-effectively scale to handle the anticipated load. They need to improve end-to-end latencies for competitive advantage:

• a software stack designed to provide hardware independence
• automation for configuring & monitoring thousands of processes on a thousand nodes
• a scalable hardware/software solution that will meet expanding needs for years to come while staying within existing data center power and space limits

Solution
The solution is a stream processing platform for financial markets. It will be demonstrated in two configurations, one using generic HW and one leveraging the full power of a Blue Gene supercomputer. Each will run the same System S hosted streaming application designed to mimic the functions & services used by sample applications.

Benefits

• Improves competitiveness
  • Ultra high speed & ultra low latency processing
  • Scales easily to addresses future workloads
  • Lower infrastructure TCO
• Removes hardware dependency
• Enables greater flexibility
  • Real-time monitoring
  • Global optimization
• Conforms to existing data center power and space limits
• 21x performance improvement on the volume of data consumed by financial trading systems

Press