Dell CFO David Kennedy: budget tokens, chase speed

Dell's message comes as enterprise AI moves from pilot projects into production, and financial services firms are among the first to show it can pay off at scale. Kennedy's advice reflects that shift: rather than obsessing over the price of a token, CFOs should track token consumption as a budget line and focus on how efficiently models are used. He pointed to model placement as the lever — running repetitive finance workflows on open-weight models on-premises turns variable token costs into fixed ones.

Dell itself is betting on that broader demand. The company counts 6,500 AI factories across its customer base and in May expanded its AI Factory with Nvidia to help more customers close the gap between experiment and outcome. Kennedy said demand is spreading well beyond GPUs into traditional servers and storage, suggesting the AI buildout is pulling along more of Dell's product line.

Kennedy's read on adoption speed is that financial services, especially high-frequency trading, led the way, with pharmaceutical companies likely next. His framing — that there are enough proof points that AI works, and the open question is whether companies can do it at scale — puts the emphasis on execution and cost discipline rather than experimentation.