On May 28, 2026, FuriosaAI and Broadcom announced a partnership to build rack-scale inference clusters (source).

On May 28, 2026, FuriosaAI and Broadcom announced a partnership to build rack-scale inference clusters (source). The collaboration aims to deliver high-performance AI inference solutions, leveraging Broadcom's networking and custom chip expertise alongside FuriosaAI's AI accelerators. This strengthens Broadcom's position in the AI inference market, potentially driving increased demand for its networking and ASIC products. The partnership is timely given the growing need for efficient inference infrastructure among hyperscalers like Google, Amazon, and Microsoft, which are heavy consumers of such clusters. Broadcom's upcoming Q2 2026 earnings call on June 3, 2026 (source) may provide additional context on revenue implications. While no financial terms were disclosed, the move signals Broadcom's deepening involvement in AI inference, aligning with industry trends toward custom silicon and dense cluster deployments.