HP outlined a shift toward hybrid computing at The Six Five Summit: AI Unleashed 2026, emphasizing the accelerating adoption of on-device artificial intelligence in enterprise workflows. Ketan Patel, President of Personal Systems at HP, noted that nearly 72% of organizations are either deploying or piloting AI agents, a trend he described as transitioning from a cloud-centric model to a hybrid approach.
The agent-based AI market is projected to reach a $6 trillion total addressable market by 2028, Patel said, while on-device AI capabilities accounted for just 2% of software development in 2024. Current laptops can locally host open-source models with up to 20 billion parameters, with next-generation silicon architecture expected to support models as large as 120 billion parameters. On-premises infrastructure, Patel added, can already accommodate models exceeding a trillion parameters.
Enterprise cost structures are being reshaped by the move toward local inference. A case study cited by HP involved a customer whose annual cloud AI spending rose from $20 million to between $200 million and $300 million after expanding model access for employees. Local inference, Patel argued, can reduce costs by $650,000 per 1,000 employees for basic models, scaling to $2 million per 1,000 employees as more advanced models are deployed on-site.
HP’s own performance has reflected broader sector momentum, with shares up 61% over the past six months and 36% year-to-date. Revenue growth for the company reached 5.74% in the latest reporting period. Patel also referenced the seven-year history of The Six Five Summit, underscoring the rapid evolution of AI integration in enterprise and consumer devices.
