📖 This is Part 1 of a three-part series on Jensen Huang. Bookmark it - it's a long read.

TL;DR

• NVIDIA didn't become NVIDIA by "believing harder." It survived because Jensen Huang did a rare CEO move: tell the truth early, even when the truth threatens your biggest contract.
• In the mid-1990s, SEGA was not a random client. It was a cultural and commercial superpower-the Pepsi to Nintendo's Coke-and winning that partnership looked like winning the future. ✓ SEGA's Genesis/Mega Drive era and the early Saturn years made SEGA a kingmaker in console credibility. (Source)
• NVIDIA's first major architecture (NV1/NV2 lineage) leaned into quadrilateral-based rendering, right as Microsoft's Direct3D standardized the industry around triangles. ✓ NV1's design and market reception are documented; it used an unusual rendering approach compared with the industry trend. (Source)
• When the bet went bad, Jensen didn't just pivot. He did it while nearly out of cash-sources describe NVIDIA having only months (even "30 days" in some retellings) of runway. ✓ Multiple profiles recount the near-bankruptcy period and drastic cuts. (Source)
• The "rebirth" product, RIVA 128, wasn't a lucky hit. It was a compressed design cycle, a standards-aligned reset, and an operational decision: buy speed (simulation/emulation) even when it's unproven. ✓ RIVA 128's 1997 release is a well-covered inflection in NVIDIA's early history. (Source)

---

NVIDIA Admitted It Was Wrong, Took $5M, and Survived

You can tell a founder story in two ways.

1. The Hollywood version: "We had a vision. Everyone doubted us. We persisted. We won."

2. The CEO version: "We made a technical bet. The market standardized against us. Our biggest customer was about to find out. We had months of cash. We had to choose between ego and survival."

💬 FORKED thesis: Jensen Huang's early years aren't inspirational. They're instructional-because they're about pride, standards, and cash, not about vibes.

There's a concept that runs through Naval Ravikant's writing: pride is the most expensive emotion in business, because it blocks the one thing that keeps you alive—learning. (Source: Naval Ravikant on The Knowledge Project)

⚡ In hardware, that tax is even steeper. Software can patch. Silicon can't. Every week you spend defending a wrong architecture, you're burning cash on pride.

---

Hook: the rarest CEO skill is not conviction - it's timed humility

A lot of founders can sell conviction.

Almost no one can sell a confession.

Yet Jensen Huang's most consequential early decision was essentially:

"Our architecture is wrong."

Not in a postmortem. Not after the board forces a reset. Not after the customer churns.

He said it to the customer who was keeping NVIDIA alive.

✓ NVIDIA was founded in 1993 by Jensen Huang, Chris Malachowsky, and Curtis Priem. (Source) ✓ NVIDIA's origin story is famously tied to a Denny's meeting. NVIDIA itself has published a retelling of the Denny's roots. (Source)

⚡ The "Denny's myth" matters less than the underlying mechanism: small teams survive by making faster truth-to-decision loops than incumbents.

💬 Most founders treat humility as personality. Great CEOs treat it as a weapon-deployed at the exact moment it buys time, trust, or leverage.

---

Background (1993-1996): a market nobody wanted (yet)

Today "GPUs" sound inevitable. In the early 1990s, they sounded like an over-engineered hobby.

The Denny's table: three careers, one constraint

The lore says "Denny's." The more useful detail is who was at that table: three engineers who had already seen how big companies move.

✓ NVIDIA was founded by Jensen Huang, Chris Malachowsky, and Curtis Priem, who had worked at companies like Sun Microsystems; Jensen also worked at LSI Logic and AMD earlier in his career. (Source) ✓ Jensen Huang's biography is widely documented, including his early engineering path before NVIDIA. (Source)

⚡ They weren't naive. They were choosing to trade stability for speed.

Then there's the constraint that makes the story feel like a CEO decision rather than a founder fantasy: money.

✓ Early funding for NVIDIA is described in historical summaries, with Sequoia Capital and Sutter Hill Ventures often mentioned among early backers. (Source)

⚡ Venture money doesn't remove constraints. It changes them. In semiconductors, you don't "run out of motivation." You run out of cash, and the cash burn is tied to physics: masks, wafers, validation cycles.

💬 This is why Jensen's later intensity about runway ("we're always close to dying") isn't drama. It's a scar tissue response to how hardware companies actually die.

The real gap NVIDIA saw

✓ Early 1990s PC gaming showcased growing demand for 3D-like experiences; in that era, dedicated 3D acceleration on consumer PCs was still emerging, while high-end 3D graphics was associated with expensive workstations (e.g., SGI). (Source)

⚡ NVIDIA's founding bet wasn't "games will be popular." Games were already popular. The bet was: 3D graphics will move from rare/expensive to common/cheap, and whoever makes that transition will own a layer of the stack.

💬 That is a classic "sell shovels" move. But the shovel market wasn't mature. It was a knife fight.

It wasn't just NVIDIA. It was a swarm.

✓ Retellings of the era often note dozens of graphics-chip startups; some sources cite ~90. (Source)

⚡ In "swarm markets," the main risk is not competition-it's standardization. The moment the industry converges on a standard, half the swarm dies overnight.

💬 Your product roadmap doesn't matter if the platform owner picks a different physics.

---

Why SEGA mattered (so much): the 90s console war was cultural power

If you're under 30, SEGA might feel like a retro logo.

In the early-to-mid 1990s, SEGA was closer to a media empire in the youth imagination.

SEGA vs Nintendo was Coke vs Pepsi-played out in living rooms

✓ The Sega Genesis (Mega Drive) was Sega's 16-bit home console and a centerpiece in the early 1990s "console war." (Source)

⚡ That's the polite Wikipedia summary. The lived reality was brand tribalism.

• SEGA marketed as faster, edgier, more "teen."
• Nintendo marketed as family-friendly, iconic characters, and quality control.

💬 It wasn't "which console has more teraflops." It was identity: what kind of kid are you?

The North America swing: why developers cared

The single most important thing to understand about 90s SEGA is this: for a moment, it beat Nintendo on Nintendo's turf-American kids.

✓ The Sega Genesis page notes that the Genesis outsold the SNES in the United States nearly two-to-one during the 1991 holiday season, and that Sega controlled 65% of the 16-bit console market in January 1992-the first time Nintendo had not been the console leader since 1985. (Source) ✓ The broader "console war" history also notes Nintendo publicly acknowledged by 1992 it was no longer in the dominant position. (Source)

⚡ Read what those two facts mean in business terms:

• Developers follow the installed base. If Genesis is the machine in more living rooms, that's where game budgets go.
• Retail follows velocity. Shelf space is a zero-sum game; whoever sells faster becomes "the default."
• Media follows heat. In a pre-social era, magazines and TV were distribution; SEGA was a storyline.
• Accessories follow standards. Controllers, cartridges, peripherals-an ecosystem forms around the winning platform.

💬 This is why young founders underestimate SEGA's leverage. SEGA wasn't just a customer with a purchase order. It was an ecosystem gatekeeper at a moment when ecosystems were smaller, more centralized, and more brutal.

⚡ When a platform wins the living room, it also wins:

• developer mindshare
• retail shelf space
• magazine coverage
• and, crucially, "default choice" inertia

💬 If you're a small chip company in 1993-1995, a console deal isn't "revenue." It's legitimacy. It's an API to culture.

Saturn wasn't just a console - it was an industrial project

✓ The Sega Saturn launched in Japan in 1994. (Source)

⚡ Saturn-era SEGA also reflects how brutal the industry was becoming: 3D transitions, rising R&D, and hardware cycles that demanded huge bets.

💬 Which is exactly why NVIDIA's SEGA partnership was both an opportunity and a trap: SEGA needed a technical edge; NVIDIA needed a lifeline.

---

Product #1 (NV1): when "differentiation" quietly becomes incompatibility

NVIDIA's first product attempt is a perfect startup cautionary tale:

• It's technically interesting.
• It's different.
• And it's wrong for the direction the ecosystem chose.

What NV1 was

✓ NV1 was NVIDIA's first product (mid-1990s) and is documented as using a rendering approach that differed from the triangle-based methods that became common later. (Source) ✓ NV1 was sold via board partners (e.g., Diamond Multimedia's Diamond Edge 3D). (Source)

⚡ NV1 tried to be too many things at once-graphics, audio, even input integration in some accounts.

💬 Hardware founders: beware the "Swiss Army chip." Integration looks like leverage until it becomes the reason nobody can adopt you.

The bet underneath NV1

⚡ NVIDIA's differentiation was: "We can render 3D differently (quadrilaterals) and maybe better for certain models."

But the ecosystem bet that mattered was: "What will the platform standardize?"

And in the mid-90s, the platform wasn't a console.

It was Microsoft.

---

The standard that killed them: Direct3D's triangle world

✓ Microsoft introduced DirectX in the mid-1990s, and Direct3D became a key API for 3D graphics on Windows PCs. (Source)

⚡ The strategic violence of standards is that they don't have to be "better." They just have to be adopted.

Once game developers build around the triangle pipeline, your fancy quadrilateral approach becomes an adoption tax.

💬 Startups die from adoption taxes more than from inferior tech.

---

The SEGA gamble (NV2): when your biggest customer becomes your biggest constraint

NV2 is where the story becomes pure CEO decision-making.

Because it wasn't just: "Our tech is wrong."

It was: "Our tech is wrong and our roadmap is entangled with the one customer we cannot lose."

✓ NV2 is commonly described in retrospective profiles as a planned follow-on tied to a Sega-related project; retellings emphasize that the shift in industry standards and Sega's strategic moves undermined the plan. (Source)

⚡ Here's the trap pattern:

1. A startup wins a marquee client.
2. The startup custom-builds for that client.
3. The client's needs become the startup's architecture.
4. A platform standard changes.
5. The startup can't pivot because it's "locked" to the marquee client.

💬 Marquee contracts can be golden handcuffs. They pay you in cash and delusion.

---

Near-bankruptcy (1996): the runway math that turns philosophy into action

You can romanticize pivots when you have cash.

When you don't, pivots become a question of time.

✓ Various profiles describe NVIDIA's mid-1990s cash crisis and layoffs; the company reduced headcount significantly during the downturn. (Source) ✓ Another account describes NVIDIA being close to running out of cash, framing it as a "months of runway" situation. (Source)

⚡ In a cash crisis, the CEO's job collapses into three verbs:

• Choose (what not to do)
• Cut (what to stop funding)
• Confess (what is true now)

⚡ When you have 6–9 months of runway, you don't get to "explore." You get to focus every ounce of attention on the one constraint that kills you first. Naval calls attention "the only real currency"—and NVIDIA was almost bankrupt in every sense.

---

The decision: tell SEGA the architecture is wrong (and still ask for the fee)

This is the moment that defines Part 1.

What happened

✓ Retellings of NVIDIA's early history describe Jensen informing SEGA leadership that NVIDIA's approach wasn't going to work and that the project should be terminated; some accounts mention a termination payment that helped extend NVIDIA's runway. (Source) ✓ Jensen has publicly described early NVIDIA products as technically poor in later speeches/interviews, reinforcing the "we were wrong" framing. (Source)

⚡ Notice the duality:

• He admits the truth (a relationship move).
• He enforces the contract (a survival move).

💬 Most founders can do one of these. Very few can do both.

Why this confession was not "nice"-it was strategic

If you're a big client and a vendor lies to you, you punish them.

If they tell you the truth early, you might still be angry-but you can also respect the competence.

⚡ Early confession buys you something more valuable than goodwill: it buys you reality alignment. That's the prerequisite for speed.

💬 In hardware, reality is undefeated. Confessing early is not morality. It's math.

Why this confession worked (and most don't)

Here's the thing about pride in a chip company—it doesn't look like ego. It looks like inventory. It looks like "staying the course." If Jensen had let pride drive the decision, NVIDIA would have:

• insisting quadrilaterals were the future
• forcing the world to meet you where you are
• shipping more silicon into an ecosystem that no longer wants it

💬 The pride tax isn't emotional—it's physical. Every unsold chip is pride you can hold in your hand.

---

Rebuild mode: RIVA 128 as a compressed reset

A lot of "pivots" are actually rebrands.

RIVA 128 wasn't.

It was a violent rewrite.

✓ NVIDIA's history highlights RIVA 128 (1997) as a major step that helped the company recover and compete in the PC graphics market. (Source)

What they changed (the non-negotiables)

1. Align to the standard (Direct3D/triangles)

✓ Direct3D's role as a Windows 3D API standard is central to the era's PC gaming stack. (Source)

⚡ RIVA 128 wasn't "our new idea." It was "we're going to win inside the rules that now exist."

💬 If you're early, you can try to invent rules. If you're dying, you play the rules that already have distribution.

1. Buy iteration speed

✓ Retellings describe NVIDIA using emulation/simulation approaches to accelerate design and shorten cycles compared to industry norms. (Source)

⚡ They spent scarce cash on tools that made the team faster. That is a founder move that looks reckless until you realize time is the only resource you can't raise.

💬 When you're in a death spiral, "unproven software" can be safer than "proven slowness."

1. Ship what matters, not everything

✓ Accounts of RIVA 128 emphasize performance and competitiveness even if feature coverage wasn't exhaustive in every mode or edge case. (Source)

⚡ This is the underrated pivot skill: deciding which completeness you can postpone.

💬 Great teams don't ship perfection. They ship leverage.

What RIVA 128 proved (beyond the product)

✓ Some accounts cite rapid sales traction after release, framing it as a turning point. (Source)

⚡ The deeper proof was cultural: NVIDIA could now iterate fast enough to survive a market where new chips appeared constantly.

💬 You don't win hardware by being right once. You win by being less wrong, faster.

---

The meta-pattern: "one thousand iterations" is not "one thousand hours"

This is another Naval-flavored lens from the research framing:

💬 There's a difference between repeating effort and repeating learning.

⚡ NVIDIA's early generations weren't just "hard work." They were iterated failure-NV1 to NV2 to RIVA 128-each failure clarifying what the market would actually adopt.

✓ The NV1 story and the later shift toward mainstream 3D APIs appear in multiple histories of NVIDIA's early years. (Source)

💬 The uncomfortable truth: you can be a brilliant engineer and still be irrelevant if you're incompatible with the platform.

---

So what exactly was Jensen's "bet-the-company pivot"?

A lot of people treat the pivot as "NVIDIA switched products."

That's shallow.

Here's the pivot at the CEO-decision level:

1. He redefined success from "proving our architecture is superior" to "winning inside the new standard."
2. He chose speed over elegance (emulation/simulation acceleration).
3. He used truth to preserve optionality with a key customer.
4. He traded identity for survival-and survival for compounding.

✓ NVIDIA later went public in 1999. (Source)

⚡ RIVA 128 wasn't the end. It was the return to the game.

💬 In Part 2 we'll cover how NVIDIA turned "back in the game" into "define the category," including GeForce and the early GPU framing.

---

FORKED Scorecard: Bet-The-Company Pivots (When Your Architecture Is Wrong)

💬 Use this when you face a pivot that isn't optional-because the standard moved, the customer changed, or the tech bet failed.

How to score

• Score each criterion 1-5.
• Be specific. If you can't point to evidence, cap yourself at 2.

Interpretation bands

• 26-35 = Hard pivot now. You're already dead in the old path; speed is mercy.
• 18-25 = Pilot pivot. Cut scope, keep runway, run a "truth sprint" to validate.
• ≤17 = Fix core first. You may be panicking. Stabilize cash and execution.

⚡ Total for NVIDIA here is roughly 32/35.

💬 Translation: this wasn't "a brave pivot." It was the only rational move left.

The kill criteria (even if your score is high)

💬 Don't pivot (yet) if:

• ⚡ Your rebuild cycle is longer than your runway and you can't extend runway (raise, renegotiate, cut deeper).
• ⚡ The "standard shift" is actually temporary noise (e.g., one game engine's fad).
• ⚡ Your team is psychologically incapable of admitting the old path is dead (you'll half-pivot and die slower).

---

The counter-intuitive insight: Jensen's confession wasn't humility - it was a speed hack

Here's the non-obvious part.

Most people interpret the SEGA moment as moral character: "He was honest."

That's true-but incomplete.

⚡ It was also an engineering move.

Because honesty collapses the search space.

• If you lie, you have to maintain two realities.
• If you confess, you have one.

💬 In a death spiral, reducing the number of realities you're juggling is how you buy time.

---

Hidden costs: what the legend leaves out

Every CEO hero story hides the price.

1) The reputational blast radius

⚡ When you tell a titan client you're wrong, you're also broadcasting: "We're not safe."

💬 You can't demand the market trust you while you're rebuilding. You have to earn it again, with shipped performance.

2) The internal morale tax

⚡ Imagine being an engineer who spent years building NV1/NV2 logic and then hearing, "We're burning it down."

💬 Great CEOs don't just pivot products. They pivot identities. That's emotionally expensive.

3) The temptation to fetishize speed

⚡ NVIDIA's later culture is famous for intensity. The origin story rewards speed because speed saved them.

💬 Survival intensity is useful. Permanent intensity is a different thing-it can become the company's blind spot.

4) The "standards hostage" risk

⚡ Once you align to a standard (Direct3D), you're also vulnerable to whoever controls its evolution.

✓ Microsoft's DirectX/Direct3D platform position was central to the era. (Source)

💬 You escaped one form of incompatibility by accepting another dependency.

---

What Would You Do?

You're Jensen Huang in 1996.

• Your first big product path is incompatible with where the ecosystem is going.
• Your biggest client is also your lifeline.
• Your runway is short.

💬 Do you confess early and rebuild-or do you try to "ship something" and hope the standard bends back toward you?

⚡ Action you can steal: write two memos.

1. The Pride Memo: "If we refuse to admit we're wrong, what are we protecting?"
2. The Runway Memo: "If we have X months, what is the fastest path to a product that fits the standard?"

Then show both to someone who will tell you the truth.

---

FAQ

Q1: When was NVIDIA founded, and by whom? ✓ NVIDIA was founded in 1993 by Jensen Huang, Chris Malachowsky, and Curtis Priem. (Source)

Q2: Where did NVIDIA's founding story take place? ✓ NVIDIA has published that its founders met at a Denny's, a story widely repeated in profiles of the company. (Source)

Q3: What was NV1, and why didn't it win? ✓ NV1 was NVIDIA's first product and used a rendering approach that differed from the triangle-based pipelines that became standard; it was sold through partners such as Diamond Multimedia. (Source) ⚡ NV1 didn't just "lose on performance." It lost on ecosystem fit: developers and APIs converged elsewhere.

Q4: What is DirectX/Direct3D and why did it matter to NVIDIA's early years? ✓ DirectX (including Direct3D) became a major Windows multimedia and graphics API suite that influenced how PC games targeted 3D acceleration. (Source) ⚡ Once the API target is standardized, "alternative primitives" become adoption taxes.

Q5: Why was SEGA such an important partner in the 1990s? ✓ Sega's Genesis/Mega Drive era made it a leading console platform in the early 1990s and a major player in the console war. (Source) ⚡ A SEGA partnership in that era wasn't just revenue; it signaled legitimacy in consumer graphics.

Q6: How close was NVIDIA to bankruptcy in the mid-1990s? ✓ Multiple retrospective accounts describe NVIDIA as extremely cash-constrained in the mid-1990s, involving layoffs and a near-bankruptcy moment. (Source)

Q7: What was RIVA 128, and why is it seen as a turning point? ✓ RIVA 128 (1997) is widely cited in NVIDIA's history as a product that helped the company recover and compete effectively in PC graphics. (Source)

Q8: What is the key leadership lesson from Jensen Huang's SEGA moment? ⚡ The lesson isn't "be honest." It's: tell the truth early enough that it changes your options, and do it in a way that preserves runway.

Q9: Is this the same Jensen Huang who later led NVIDIA to define the GPU era? ✓ NVIDIA introduced the GeForce brand and later described GeForce 256 as the first "GPU," and NVIDIA went public in 1999-events that set up the next phase of the company. (Source)

---

Related Reads

• From Snowboards to $292B: Shopify's 5 Pivot Decisions (With a Scorecard You Can Steal)
• Canva's Melanie Perkins: The Decision Discipline Behind 100 No's
• DuPont CEO Spent ~$20B to Escape Oil
• Luckin Coffee Faked $310M, Got Delisted, Then Crushed Starbucks

---

Sources

• NVIDIA corporate blog - Denny's origin story: https://blogs.nvidia.com/blog/nvidia-dennys-trillion/
• NVIDIA corporate blog - Huang NTU commencement (Jensen remarks on early products): https://blogs.nvidia.com/blog/huang-ntu-commencement/
• Wikipedia - NVIDIA: https://en.wikipedia.org/wiki/Nvidia
• Wikipedia - Jensen Huang: https://en.wikipedia.org/wiki/Jensen_Huang
• Wikipedia - NV1: https://en.wikipedia.org/wiki/NV1
• Wikipedia - DirectX: https://en.wikipedia.org/wiki/DirectX
• Wikipedia - Sega Genesis: https://en.wikipedia.org/wiki/Sega_Genesis
• Wikipedia - Sega Saturn: https://en.wikipedia.org/wiki/Sega_Saturn
• Inc.com - Against a Wall (near-bankruptcy + SEGA episode): https://www.inc.com/peter-cohan/against-a-wall-how-jensen-huang-saved-nvidia-in-the-1990s/91065579
• Acquired Briefing - NVIDIA I (1993-): https://www.acquiredbriefing.com/p/nvidia-i
• BFS Invest - Jensen Huang profile (near-bankruptcy retellings): https://www.bfsinvest.com/nvidias-jensen-huang-the-ultimate-entrepreneur