Building a new simulation platform with Qarnot and Microsoft Azure
Societe Generale provides financial services to 31 million individuals and businesses worldwide. Its corporate and investment banking business - SG CIB - provides global access to markets through solutions in equities, bonds and currencies, commodities and alternative investments.
A simple, scalable and resilient platform
Societe Generale partnered with Qarnot computing and Microsoft Azure to build its new flexible, scalable and environmentally friendly financial simulation platform.
\"The challenge for Qarnot Computing and Societe Generale was to deliver a modular, scalable and resilient application in a very short timeframe.\"
Financial simulations typically involve several thousand computational tasks, each taking from a few seconds to several minutes to calculate. They also rely on a lot of data. Because Société Générale has offices all over the world, jobs can be run at any time. Some of the simulations also have strong computation time constraints.
Société Générale chose Qarnot for its low-carbon and distributed infrastructure, optimized for batch computing and thus particularly adapted to financial compute.
We designed a solution that:
- exposes a simple REST API to client applications within Société Générale,
- handles compute ranging from a few tasks to several thousand,
- provides caching of financial data for efficient task distribution,
- is available 24/7.
All this is achieved in a context where delivery must be frequent (simulation libraries are constantly evolving).
The key components of this solution are:
- An https web gateway exposing simulation services via REST API.
- A set of microservices managing data caching and simulation job orchestration, from task allocation to result in retrieval.
- Multiple cloud providers on a single grid (new providers can be added very easily).
- The Web gateway and the microservices are native Service Fabric applications, all deployed in a scalable cluster in Azure Cloud.
Results:
Now that the simulation platform is in production, the team is focused on integrating new types of simulation and scaling up so that more and more client applications can move away from Legacy systems.
In addition to the ability to simply absorb the ramp-up in risk compute, this bank now has a detailed report of its footprint on this activity. It also knows the energy savings, which in this case was 62.36%.