# IBM studies open-model agent factory for chip design

> IBM Research tested an agent factory using open-weight models and distributed inference to search hardware designs for a quantum error-correction decoder. Under the same 100-hypothesis search budget, the company reported inference costs of about $26 for the open-model configuration and $299 for the proprietary-model configuration.

- Canonical URL: https://newsverum.com/en/news/ibm-ai-agent-factory-quantum-chip-design-costs
- Publisher: Newsverum
- Section: Science
- Country: United States
- Published: 2026-10-15T08:58:14+00:00
- Updated: 2026-10-15T08:58:14+00:00
- Event / data period: 2026-10-13
- Other editions: [Türkçe](https://newsverum.com/tr/news/ibm-ai-agent-factory-quantum-chip-design-costs.md), [العربية](https://newsverum.com/ar/news/ibm-ai-agent-factory-quantum-chip-design-costs.md), [Français](https://newsverum.com/fr/news/ibm-ai-agent-factory-quantum-chip-design-costs.md), [Español](https://newsverum.com/es/news/ibm-ai-agent-factory-quantum-chip-design-costs.md), [Deutsch](https://newsverum.com/de/news/ibm-ai-agent-factory-quantum-chip-design-costs.md), [Português](https://newsverum.com/pt/news/ibm-ai-agent-factory-quantum-chip-design-costs.md)

The open-model configuration achieved a 52x speedup over the baseline in synthesis estimates, while the proprietary configuration reported 204x. Each configuration was evaluated in a single search run; the study shows a cost-performance trade-off on one engineering task, not a universal model ranking.

Researchers also reported that extending the open search from 100 to 200 hypotheses raised the speedup from 52x to 60x, with diminishing gains. Inference cost is not the cost of the full hardware-development workflow; the results are IBM-reported synthesis estimates.

## Data

### Demonstration result

IBM Research tested agent-based design search for quantum error-correction decoder hardware using open models and llm-d.

Methodology: The event is IBM Research’s company demonstration dated October 13, 2026. Cost and speedup figures describe model inference expense and high-level synthesis estimates in the study, not measurements on fabricated silicon.

| date | label | description | status |
|---|---|---|---|
| 2026-10-13 | IBM studies open-model agent factory for chip design | IBM Research tested agent-based design search for quantum error-correction decoder hardware using open models and llm-d. | occurred |

Sources: <https://research.ibm.com/blog/open-agent-factories-and-distributed-inference-for-chip-design> · Download data: [JSON](https://newsverum.com/api/v1/articles/ibm-ai-agent-factory-quantum-chip-design-costs/data/event-summary?lang=en&format=json) · [CSV](https://newsverum.com/api/v1/articles/ibm-ai-agent-factory-quantum-chip-design-costs/data/event-summary?lang=en&format=csv)

## Sources

- [IBM Research](<https://research.ibm.com/blog/open-agent-factories-and-distributed-inference-for-chip-design>) (Source checked: 2026-10-15; Published: 2026-10-13)

---

Cite the canonical URL above. [Editorial standards and corrections](https://newsverum.com/en/editorial)