AIToday
Large Language ModelsOpen-Source AIAI Business & IndustrySiliconANGLE AIPublished: Nov 18, 2026, 01:00 JST

Dell's O'Sullivan: token budgets spent in weeks, not months

Send on LINE
Dell's O'Sullivan: token budgets spent in weeks, not months

3 Key Points

  1. What happened

    Dell's Helen O'Sullivan said Bay Area customers run through token budgets allotted for months in a matter of weeks, and Intel's Tim Wood said existing data-center CPUs may have headroom.

  2. Why it matters

    That budget pressure is pushing enterprises to keep frontier models but lean on open-source and on-premise infrastructure, which may ease the token stress O'Sullivan described.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Context & Analysis

The AI Leadership Symposium conversation, as reported by theCUBE, centered on the practical friction of moving enterprise AI from pilot to production. Helen O'Sullivan's account of token budgets evaporating in weeks rather than months illustrates why cost control has become a board-level topic, and Intel's Tim Wood pointed to existing data-center CPUs as a possible source of headroom. That pairing of budget pressure and underused hardware suggests enterprises may increasingly blend frontier models with open-source and on-premise options.

A second thread ran through governance. Shiv Agarwal of Singulr AI Inc. described agent actions, such as fetching data from the internet or emailing a private account, that can be deeply hidden inside an agentic action, making runtime oversight necessary. Sri Viswanath of Sycamore Labs Inc. said CIOs often favor virtual private clouds to keep sensitive data inside a controlled environment, reflecting a distrust of letting data leave.

The build-it-yourself culture of Silicon Valley surfaced in Deepgram's plan to run voice AI models natively and air-gapped on customer devices with Dell and Fortanix Inc. Dell's Sean O'Connor said the partner ecosystem is starting to bleed into private equity and venture capital, a sign that Dell is widening how it reaches technically sophisticated buyers who prefer to assemble their own solutions.

FAQ
What is driving enterprises to rethink their AI budgets?
Cloud token bills remain a major pressure point, and Dell's Helen O'Sullivan said some Bay Area customers run through a token budget allotted for months in a matter of weeks.
What does Deepgram plan to do with its voice AI models?
Deepgram is working with Dell and Fortanix Inc. to bring its voice AI models beyond the cloud and onto customer devices, aiming for them to run natively and air-gapped.
What governance concern do companies deploying AI agents face?
Shiv Agarwal of Singulr AI Inc. said runtime governance is needed to assess AI agent behavior during execution, because actions like reaching out to the internet can be deeply hidden inside an agentic action.
SiliconANGLE AIRead Original Article
Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articleAI server upgrades push Taiwan PCB, CCL, IC substrate supply chain upmarket