Research Notes

IBM’s LLM-Guided Evolutionary Framework for Quantum Error Correction Code Discovery

Steve McDowell June 16, 2026

IBM Research recently published a paper on an LLM-guided evolutionary framework for discovering quantum error-correction (QEC) codes, accompanied by the open-source release of OpenEvolve on GitHub. The framework applies evolutionary artificial intelligence techniques, originally developed for general program synthesis, to the domain of quantum low-density parity-check (qLDPC) codes.

Microsoft’s Majorana 2 Topological Quantum Chip

Steve McDowell June 9, 2026

Microsoft unveiled Majorana 2 at its recent Build developer conference in San Francisco, introducing its second-generation topological quantum chip and announcing the general availability of Microsoft Discovery, an agentic AI platform for scientific research and development.

Majorana 2 delivers qubits that are 1,000 times more reliable than those in its predecessor, Majorana 1, with a mean qubit lifetime of 20 seconds and peak instances approaching 1 minute.

Qualcomm’s Dragonwing IQ10 Robotics Reference Design

Steve McDowell June 9, 2026

At the recent Computex 2026, Qualcomm introduced the Dragonwing IQ10 Robotics Reference Design (RRD), the company’s most concrete commitment to the physical AI market to date. The new platform consolidates compute, sensor interfaces, deterministic I/O, networking, and a layered software stack into a single enclosed reference design for autonomous mobile robots (AMRs), industrial robotics, and humanoid platforms.

Microsoft: Project Solara & the Agentic Operating Plane for the Non-NVIDIA World

Steve McDowell June 7, 2026

At its recent Build 2026 event, Microsoft introduced Project Solara, a chip-to-cloud platform built specifically for devices whose primary interface is an AI agent. Solara

Cisco’s Full-Stack Play for Agentic Enterprise AI

Steve McDowell June 7, 2026

At Cisco Live 2026 in Las Vegas, Cisco unveiled Cisco Cloud Control, a unified management plane designed for human operators and AI agents to jointly run and defend enterprise infrastructure. The platform anchors Cisco’s AgenticOps operating model, unifying networking, security, compute, observability, and collaboration under a single login and a shared data layer.

IBM & Red Hat: Project Lightwell, Protecting Open Source Software

Steve McDowell May 29, 2026

IBM and Red Hat announced Project Lightwell, a $5 billion initiative to establish a trusted enterprise clearinghouse for open-source software security. The project deploys more than 20,000 engineers, augmented by AI, to identify, triage, validate, and remediate vulnerabilities across open-source supply chains at a scale that exceeds what most enterprises can achieve independently.