Skip to main content
FINOS Open Source · HPC

High‑performance computing,
built for modern finance.

In the era of AI-driven finance, high-performance computing (HPC) is the backbone of innovation—enabling real-time risk modeling, algorithmic trading, generative AI, and large-scale data analytics. At FINOS, we’re pioneering open-source HPC solutions that combine scalability, cost-efficiency, and sustainability to meet the demands of modern financial workloads. Our suite of HPC projects empowers organizations to accelerate AI/ML training, optimize resource allocation, and deploy cloud-native computing at scale. Whether you’re running AI inference workloads, distributed data processing, or dynamic grid computing, FINOS provides the tools to boost performance, reduce costs, and minimize environmental impact.

Join the community
Grid schedulingCloud‑native workloadsClimate‑aware utilization
HPC

Open Standard for Grid Resource Scheduling

Standardize how compute jobs move across grids, clusters, and clouds. OpenGRIS gives you a common language for describing capacity, priorities, and policies so teams can share resources without sharing infrastructure. Open project

Cloud-Native High Throughput Compute Grid

HTC-Grid is a container-based, cloud-native HPC/Grid environment that provides a reference architecture for building modern high-throughput computing solutions. Leveraging AWS services, it enables users to submit high volumes of short and long-running tasks while dynamically scaling environments. Open project

Intelligent Time-Based Scheduling for Kubernetes

5 Spot is a Kubernetes operator that provides intelligent, time-based scheduling for physical machine deployments. It optimizes resource allocation by aligning workloads with availability windows, reducing costs, and improving efficiency. Open project

Kubernetes Integration for IBM Symphony

The Open Resource Broker extends IBM Symphony’s HostFactory to provision Symphony compute nodes on Kubernetes. This provider enables seamless integration between legacy HPC workloads and modern Kubernetes-based infrastructures. Open project

How it fits into your HPC landscape

Start small, plug into what you already run today, and grow toward a shared, open HPC fabric across desks, regions, and cloud providers.

01

Describe your resources

Model clusters, queues, and policies in OpenGRIS so schedulers and tooling share a common view of capacity, SLAs, and constraints.

02

Connect existing schedulers

Integrate on-prem and cloud schedulers, then surface a unified job routing layer for quant, risk, and analytics teams.

03

Optimize for cost & climate

Shift workloads based on price, energy mix, or regulatory boundaries while keeping observability and governance in one place.

Who is this for?

Built for the teams that run compute‑heavy workloads and the teams that govern platforms, cost, and sustainability.

Quant / Research

Faster iteration on models

Run bursty simulations and training jobs without fighting queue policies or guessing where capacity lives.

Risk / Analytics

Predictable batch at scale

Standardize scheduling and routing so nightly risk and analytics runs hit SLAs across regions and platforms.

Platform / Infra

Composable HPC platform

Integrate existing schedulers and clusters behind a common model while keeping governance and observability consistent.

MLOps

Reliable training & inference

Standardize how GPU/CPU capacity is scheduled so pipelines, training, and inference can scale without fragile, one-off integrations.

Supporters

Supported by industry leaders

These organizations support the project and its mission to build an open, modern HPC stack for finance.

CitiMorgan StanleyRBCAWSOracle