IBM Research

IBM Research is IBM’s global R&D organization, with 3,000+ scientists and researchers across a network of labs worldwide.

A long track record of foundational computing work

IBM Research has been active since the mid-20th century and is widely credited with multiple breakthrough technologies that shaped modern computing and IT infrastructure.

• Storage + hardware foundations — IBM Research teams pioneered core storage technologies, including early hard disk drive systems.
• Semiconductors + systems — IBM researchers helped create foundational building blocks used across modern computing.
• Awards + impact — IBM’s research community has a long history of major academic and industry recognition.

Open source and developer ecosystems

IBM Research regularly turns research into developer-facing, community-driven projects.

• Qiskit — IBM’s open-source quantum SDK and ecosystem for building and running quantum programs.
• Granite open models — IBM has open-sourced parts of the Granite family (including code-focused releases) and continues investing in practical, governable AI for production systems.

CUGA and open-sourcing enterprise agents

In 2025, IBM Research introduced CUGA (Computer Using Generalist Agent / ConfigUrable Generalist Agent) as a configurable, enterprise-focused generalist agent and open-sourced it under the cuga-project GitHub organization.

CUGA is designed for real-world digital work across web + API tasks and is built to integrate with modern tool interfaces like MCP, alongside common enterprise constraints such as auditability and governance.

Recent highlights

• NorthPole — a research prototype chip architecture aimed at significantly improving AI inference efficiency.
• Granite + tooling — continued open releases and tooling aimed at making enterprise AI more transparent and usable in practice.

For current projects and publications, see https://research.ibm.com/.