Which AI infrastructure option actually holds a boundary?
The article promotes a Washington dinner on sovereign AI infrastructure, where technology and risk leaders can question HPE and NVIDIA representatives about how different architectures protect data and jurisdictional boundaries. The discussion matters because infrastructure choices involve trade-offs that may only become clear when a pilot moves into production.
The dinner is scheduled for 29 October, from 6.00pm to 8.30pm at Fiola Mare in Washington DC, and will be chaired by The Register’s editor-in-chief under the Chatham House Rule. HPE will send Bharath Ramesh and Tracy Mills, while NVIDIA will send Bert Condensa. Topics include capital costs, capability, elasticity, operational responsibility and what evidence auditors may accept; the event is aimed at leaders in sectors including healthcare, finance, government and higher education.
- HPE and NVIDIA representatives will discuss sovereign AI infrastructure in Washington.
- The dinner will examine costs, data boundaries and operational trade-offs.
- It is scheduled for 29 October under the Chatham House Rule.
New here? Start with this
Many organisations are concerned about where their data goes when they use artificial intelligence and which countries' laws apply to it. The choice of which AI systems and computing infrastructure they use can determine whether data stays within their national borders or moves abroad, affecting which regulations matter most.
Two major technology companies, HPE and NVIDIA, are hosting a discussion in Washington where business and government leaders can explore how different technical approaches protect data and keep it within particular jurisdictions. The event brings together decision-makers from healthcare, finance, government and higher education to question company representatives about the practical implications of these choices.
The discussion matters because organisations must often make long-term technical decisions before they fully understand the consequences, which typically only become clear once systems are actually in operation at scale. Topics will include the costs and capabilities of different options, as well as what evidence regulators and auditors might require to verify that boundaries are truly being respected.