The Specialist Services stream in Risk Focus is designed to pair very specific client needs with deep technology and domain expertise. In this case we are leveraging skills and expertise gained at Google and applying them to Enterprise Risk Management solutions in Capital Markets.
The goal of this service is to help clients analyze and optimize their cluster applications:
- Reduce the size of compute clusters whilst maintaining performance & SLA’s
- Increase trades processed & scenarios run without impacting hardware costs
- Reduce the time taken to run end of day or intraday compute jobs
Optimizing your cluster applications will get you latency and speed gains and may reduce overall cpu usage. But fine tuning is hard, there are lots of variables and sometimes they conflict with each other. Furthermore, the environment is dynamic which means what works today might not work tomorrow. Our goal is to help you get the most out of your compute cluster.
Our High Performance Computing service offers a comprehensive optimization strategy that focuses on 3 specific areas:
- Job Characterization: Can your jobs be better understood and managed to ensure optimal use of your machines?
- Caching Strategies: Can you use caching or optimize your caching strategy to reduce latency and increase throughput?
- GPU’s: Can you use GPUs cost effectively to increase your compute capacity?
The service deliverables follow a simple development life cycle and you can engage at any point in the life cycle, depending on where you are in your own optimization process.
- Discover: Collect and Analyze the metrics and data points required to assess current state.
- Design: Recommend and Document areas of investment and expected returns.
- Deliver: Implement recommended changes and Measure their impact and ROI.
To get started we offer a free introductory workshop. This includes a detailed presentation delivered at QCon NY2013 (received 94% approval rating). We will provide an overview of the various streams of work and the methods that can be applied to the optimization of your compute infrastructure as well as some of the potential pitfalls. Afterwards there will be a detailed Q&A session for you to drill into specific topics and where we will gather the data required to provide sizing estimates for the various phases of work described above.
Clive started his career at Application Networks™, developers of JRisk, a Java based risk framework, along with Vassil Avramov (Risk Focus co-founder) before leaving for an opportunity to work at Google™. Clive has now joined Risk Focus with a view to applying the skills and knowledge gained at Google™ to the challenge of building scalable and cost effective Enterprise Risk solutions for our Capital Markets Clients.