Assemble War Rooms: How Jensen Huang and Lisa Su Rally Small Teams to Solve Existential Threats Fast
Eight ingredients for mobilizing when the stakes are existential.
Matt Hunter • March 3, 2026
If you’re leading a company right now, you probably have at least one issue that’s quietly compounding damage.
A stalled launch.
A reliability incident.
A sales dip no one can quite explain.
A competitor moving faster than you expected.
Most teams handle these problems inside normal operating cadence—weekly meetings, Slack threads, scattered ownership.
That’s a mistake.
Some problems don’t need better coordination.
They need a war room.
A war room is a temporary, ad hoc group assembled to solve an important and urgent problem, fast. Throughout history, existential threats, when addressed by the right people in a focused war room, have become engines of innovation.
Successful companies know how to rapidly mobilize around critical issues. The team is small and cross functional, with clear decision rights, a central scoreboard, live data, a tight meeting cadence, and mutual accountability supported by clean, timestamped commitments. It’s the opposite of normal day-to-day operations; it focuses on solving urgent problems quickly. It has explicit exit criteria and disbands when the job is done. When a good war room convenes, problems that might otherwise linger for weeks get solved in days.
The war room should be reserved for your “must-win” battles. Don’t create one for routine operations or keep it running continuously, or you’ll burn people out and dilute its power. Convene only when the stakes are high and delay compounds damage: an existential competitive threat, a failing launch, a major customer at risk, a reliability incident, or a baffling sales dip. The essential purpose is to provide a handful of smart, dedicated people with what they need to solve the problem as quickly as possible. As the anthropologist Margaret Mead is often credited with saying, “Never doubt that a small group of thoughtful, committed citizens can change the world; indeed, it’s the only thing that ever has.”
At NVIDIA the war room shows up in a practice known as Swarming, where Huang pulls the right people into the room, regardless of hierarchy, to tackle urgent problems head-on. NVIDIA’s culture is no-nonsense; problems are solved immediately. Huang has little tolerance for formal meetings; when he identifies an urgent need, he assembles people, makes decisions on the spot, and works diligently until the problem is resolved. The war room is toughness operationalized.
There are eight ingredients for running an effective war room:
1. Rally Around a Goal
Successful war rooms work because everyone knows exactly what they’re fighting for: a clear, bold target that energizes the team and brings everyone’s efforts into alignment. War room goals often plug leaks or seize fleeting opportunities. It should be something you can describe in one sentence that includes a number and a deadline: “Restore checkout uptime to 99.95 percent within seventy-two hours;” “Rebuild Q4 pipeline to +$3 million by October 15.” Clearly state the exit criteria that signal when the war room disbands: “Incident closed and uptime greater than 99.95 percent for seven consecutive days”; “$3 million in new qualified pipeline logged and verified in CRM.”
2. Pick the Right Team
Assemble a small, cross-functional team with just enough people and the right capabilities to get the job done. Follow Amazon’s “two-pizza rule” for assembling problem-solving groups: If two pizzas can’t feed the team, it’s too big (aim for five to seven people or fewer). For a major outage, that might include engineers to fix the issue, marketing to craft messaging, and customer success to handle customers. Invite only doers with bandwidth; everyone else is a stakeholder who gets asynchronous updates. Also apply the uninvite rule: If someone has no owned deliverables or hasn’t spoken for two meetings, they’re likely out. No tourists, only contributors. And remember this: Titles don’t fix problems; people do. Steve Jobs routinely bypassed hierarchy and pulled frontline engineers into small war room squads when they were the ones who could solve the issue fastest.
3. Appoint the Directly Responsible Individual
Every war room needs a DRI, the person accountable for solving the problem from start to finish. The DRI makes the calls, sets the cadence, assigns work-stream owners, and brings in additional expertise only when a missing capability blocks progress. Assign an ops partner or chief of staff to join the meetings and keep everything organized, updating KPIs and dashboards, capturing decisions, recording “who does what by when,” chasing follow-through, and handling reminders. This frees the DRI to stay on the critical path.
4. Light a Fire
At kickoff and any time progress stalls, deliver a quick but powerful brief explaining why this urgently matters, what’s at risk, and what success unlocks. Use your language to tap intrinsic motivators: connection to the mission, pride in craft, an at-stake reputation, and a competitive spirit. As a leader, you don’t always need to bring the thunder. Sometimes the problem is so glaring that everyone already feels the heat.
5. Clear Decision Rights
At kickoff, begin by outlining who decides what. Who is authorized ship or commit code? How much money can be spent without further approval? Who can communicate with customers? Delegate decision-making power, reserving the final say on important decisions for the DRI. Borrow Amazon’s language and sort decisions into two buckets: Two-way doors are easy to reverse and should be made quickly by the closest owner, while one-way doors (brand, compliance, big spend) are hard to reverse and should be escalated to the DRI for a deliberate call. Decisions should be made quickly and communicated continuously. Information flow maintains strong momentum and keeps the team aligned.
6. Establish a Cadence
Set a tempo and stick to it. Hold fixed daily stand-ups (as many as three times a day if it’s hot). Your agenda should focus on changes in the numbers, key decisions on deck, barriers to progress, and new commitments. Track goals visibly, and hold people accountable. Between stand-ups, execute. This rhythm provides rapid feedback, allowing you to course correct as you go. If you can’t meet in person, conduct the war room online with the same urgency.
7. Build a Battlefield Picture
Centralize the data so everyone can see it through a single shared dashboard and comms channel. Show only the vital signs: three to five key indicators and your main goal that update hourly or daily. Keep this dashboard visible in the room or virtual space, and pin the link in the channel where all updates and relevant files live. If you’re running an important campaign, for example, track sessions, sign-ups, paid conversions, customer acquisition cost, and return on ad spend.
8. End It on Purpose
Close the war room the moment the exit criteria are met: The site is working again, and customers can place orders without issues. We’ve fully caught up on shipping, and no orders are delayed. Say it plainly (“War room is closed”), then run a debrief: What did we learn? Which decisions were slow (and why)? Which practices should we standardize? Steve Jobs was a great believer in postmortems, gathering honest feedback on what worked and what didn’t. Assign owners and deadlines to any follow-ups, and archive the rest. Express appreciation for the team’s effort, urgency, and sacrifice by being specific, naming names, and spotlighting above-and-beyond moments. Publish a one-page summary, return decisions to normal governance, and allow the team a period of recovery time.
War Room: Lisa Su and AMD
In the early days of AMD (Advanced Micro Devices Inc.), the computing company struggled financially due to fierce competition from Intel. Over the years, AMD has undergone numerous leadership changes, yet it has managed to survive. Today, under the leadership of Dr. Lisa Su, the company is thriving. AMD became a recognized leader in the chip space due to a critical decision made by Su that turned the company around.
When Su became CEO in 2014, AMD had fallen far behind. The stock price was at a historic low, while competitors such as Intel and NVIDIA surged ahead with breakthrough innovations. Su knew that an innovative new strategy was necessary for their survival. A visionary with deep technical expertise, she set her sights on the company’s product and technology roadmaps. Su challenged her leadership team to bring fresh ideas to the table.
As Moore’s Law—the observation that the number of components on a microchip doubles roughly every two years—became increasingly difficult to sustain by simply adding more transistors, the team began exploring chiplets, an entirely new approach to building chips. Although researchers had long considered breaking chips into smaller chiplets, no company had been bold enough to attempt the technique to create real products, let alone build an entire roadmap around it. The theory was untested and would require rethinking the entire design, manufacturing, and assembly processes.
Su didn’t wait for conditions to improve to test the idea—she quickly spun up a war room with her best leaders and engineers. She set a bold, high-stakes goal; brought the right people together; encouraged honest, passionate debate; ensured unambiguous accountability; maintained a disciplined meeting cadence; and adopted a direct, courageous style that inspired deep commitment to the bet they were making. The team explored how chiplets could enable better performance and greater customization while also improving manufacturing yields. They also evaluated whether this approach would be effective at scale. As she listened carefully to her team’s ideas and examined the data, she cemented her conviction that this bold decision would be the key to reestablishing AMD’s competitive edge.
Once the decision to pursue chiplets was finalized, the real work began. Over the next several years, Su worked tirelessly to ensure that the company executed with precision and met every milestone for its customers, holding leaders accountable for their commitments. Reflecting on this crucial period at AMD, Su realized it had been “almost a ‘bet the company’ decision”—but that bet paid off.
With revenue of around $5.5 billion and a market cap near $3 billion in 2014 (when Su became CEO), AMD reported revenue of $32.0 billion for the twelve months ending in September 2025—a 482 percent increase—and a market cap of $421 billion in November 2025. Su’s direct and bold style has been a powerful force in sustaining AMD’s phenomenal success. In 2024 Time magazine named her CEO of the Year, recognizing the impact of her vision and execution.