Nvidia has reportedly barred several small cloud providers, known as NeoCloud companies, from attending the GTC Berlin conference. According to data from the analytical firm SemiAnalysis, this decision stems from these providers integrating hardware from competitors, including AMD and Google TPU, into their service offerings.
Small cloud providers specializing in AI computing resources face significant pressure due to their reliance on Nvidia. While some companies have begun offering alternative accelerators, they often do so discreetly to avoid potential retaliation from the dominant market player. Larger cloud platforms, however, appear to be less affected by these restrictions.
This development follows industry discussions from June 2026 regarding the potential impact of using third-party networking equipment. Concerns were raised that such choices could jeopardize access to new Nvidia chips or support during critical business milestones, such as IPOs or venture capital fundraising efforts.
The relationship between Nvidia and these providers is further complicated by direct financial ties. In several instances, Nvidia has acted as an investor or assisted these companies in securing funding. This dependency makes the adoption of rival hardware a significant risk for those seeking to maintain their ongoing partnership with the chip manufacturer.
