Skip to main content

The Rise of NVIDIA, Part 5: The Xbox Gamble

The Xbox Gamble

🎬 Prefer to watch? Here's the 60-second version:

Watch on YouTube ▶

🎮 The Rise of NVIDIA — a 20-part series. See all parts »  |  « Part 4: Inventing the GPU

By the summer of 2000, Nvidia had a swagger it had earned the hard way. It had survived the graphics-chip bloodbath of the late 1990s, coined the word "GPU," and buried most of its rivals. So when Microsoft came shopping for a graphics processor to power a secret console codenamed "Xbox," Nvidia did not just want the deal — it wanted to win it, the way it won everything else. What it got was the biggest contract in its short history, and one of the hardest lessons in the fine print of business.

Winning the console nobody had built yet

Microsoft had never made a game console. That was exactly the point: the company was terrified of a future where Sony's PlayStation became the living-room computer and Windows became irrelevant. To build the Xbox fast, Microsoft chose off-the-shelf PC brains — a 733 MHz Intel Pentium III for the CPU and a custom Nvidia chip for the graphics. It was a PC in a black box, and that was the strategy.

Getting the graphics contract was not a formality. A scrappy design house called GigaPixel had been working closely with Microsoft and nearly landed the job; when it lost out, the ailing 3D pioneer 3dfx swooped in and bought GigaPixel in early 2000, partly on the strength of that console ambition. Nvidia beat them all. And in a fitting coda, when 3dfx collapsed later that year, Nvidia bought its assets — patents, brand, and the GigaPixel technology included — for roughly $112 million. The company that lost the Xbox bid effectively got absorbed by the one that won it.

The prize was enormous. Nvidia reportedly received around $200 million up front as a down payment against future chip sales and licensing. It would supply two custom parts: the NV2A graphics processor, a cousin of the GeForce 3 clocked at 233 MHz, and the MCPX media and communications processor that handled audio and I/O. For a company Nvidia's size, this was a firehose of revenue attached to a product Microsoft intended to sell by the tens of millions.

The trap inside the contract

The Xbox launched in North America on November 15, 2001 at $299, and it did the thing consoles do: it sold at a loss. Microsoft's model — still the industry playbook — is to lose money on the hardware and make it back on games and services. That works only if the hardware gets cheaper over time, so the losses shrink with every price cut. And the single most expensive component to shrink was the silicon.

Here was the problem. Nvidia's supply agreement locked in chip pricing, but Microsoft's whole business depended on those prices falling. As the console aged and Microsoft slashed the retail price to compete with Sony, it wanted its component costs to fall in step. Nvidia, sitting on a fixed-price contract it had negotiated from strength, saw a guaranteed margin it had no reason to give away. Two companies, one console, and directly opposed incentives baked into the paperwork.

It turned into open conflict. On April 23, 2002, Microsoft escalated the pricing fight into binding arbitration, seeking damages and a court-ordered reduction in what it paid per chipset. Nvidia, for its part, began deferring the disputed revenue — the gap between what Microsoft was paying and what it insisted it should pay. During the process the arbitration panel issued an interim ruling that Nvidia had to keep supplying Microsoft's reasonable requirements of chips, so the Xboxes kept rolling off the line even as the lawyers argued over the price tag on their guts.

A costly education

The two sides finally called a truce. On February 6, 2003, Nvidia and Microsoft announced they had settled every issue tied to the pricing of the Xbox GPU and MCP and ended the arbitration. Nvidia agreed to collaborate with Microsoft on future cost reductions — the very thing the fight had been about — and the settlement let it recognize about $40.4 million in additional revenue in the fourth quarter of its fiscal 2003. On paper, Nvidia came out whole. In practice, it had spent a year at war with its single largest customer.

The Xbox was a genuine success: roughly 24 million units sold over its life, each one carrying Nvidia silicon, with Halo turning the console into a cultural phenomenon. For Nvidia, the deal delivered a mountain of revenue at a formative moment. But it also taught a lesson the company never forgot — that the console business is a game of guaranteed price erosion, where the customer's success depends on squeezing you a little harder every year.

The bill came due when Microsoft built its next machine. For the Xbox 360, Microsoft walked away from Nvidia and hired rival ATI to design the graphics, this time insisting on owning the intellectual property so it could shop the manufacturing around and drive costs down on its own terms. The pricing fight had consequences that outlasted the settlement. Nvidia would return to consoles later — in Sony's PlayStation 3 and, triumphantly, in the Nintendo Switch — but it went in with its eyes open, having learned exactly what it costs to power someone else's dream.

And the timing of the Xbox fight could hardly have been worse, because Nvidia was about to stumble into a disaster of its own making — a graphics card so loud and so late that gamers gave it a nickname. Next: The Dustbuster, the GeForce FX flop and the war with ATI.


🔗 Explore more from Syncster

Comments

Popular posts from this blog

Cursor AI Review: Is the AI Code Editor Worth It?

I've been using Cursor as my main code editor for a while now, and enough people have asked whether it's worth switching to that a proper review felt overdue. Short version: for me, yes — but with caveats. What is Cursor? Cursor is an AI-first code editor built as a fork of VS Code. That means every extension, theme, and keybinding you already use in VS Code works here, but with AI woven directly into the editing experience instead of bolted on as a plugin. It's made by Anysphere and can run models from OpenAI and Anthropic under the hood. What I like Tab completion is uncanny. Cursor predicts your next edit — not just the rest of the line, but the next change across the file. Once you get used to hitting Tab, going back to a plain editor feels slow. The Composer / Agent mode. You describe a change in plain language and it edits multiple files at once, showing you a diff to accept or reject. For refactors and boilerplate, this saves real time. It unde...

How I used Google Sheets and Apps Script

Google Sheet is one of the most powerful spreadsheet application that exists online, rivaling with Microsoft's Excel. One of the main strengths is its strong support for collaboration with other users, much easier and popular than collaboration tools with Microsoft Office. Aside from plain spreadsheet, it also supports extensions such as macro. If you are familiar with macros on other office tools, they work almost the same. However, the most extension I use and tinker with is the Apps Scipt . Apps Script Extension One of the challenges I faced recently is how do I track or monitor reports in our department if they are submitted on time or worst, forgotten due to lack of better monitoring tools. So I thought if there can be simple applications that can be deployed or use by a more general user to allow reminding periodically what reports are approaching due dates or those that are past dues. Then I looked for a way, instead of creating a full blown app from scratc...

MacBook Pro M5 vs M5 Pro: Which One Should You Actually Buy?

Apple's latest 14-inch MacBook Pro comes in two very different flavors: the base M5 and the step-up M5 Pro . On paper they look similar — same gorgeous Liquid Retina XDR display, same design — but under the hood the gap is bigger than the names suggest. Here's a clear, no-hype breakdown, with concrete use cases so you can match the chip to your work. Quick spec comparison Spec M5 M5 Pro CPU 10-core (4 performance + 6 efficiency) Up to 18-core (6 performance + 12 efficiency) GPU 10-core Up to 20-core Neural Engine 16-core 16-core Memory bandwidth 153 GB/s 307 GB/s (roughly double) Unified memory 16 / 24 / 32 GB 24 / 48 / 64 GB Max storage Up to 4 TB SSD Up to 8 TB SSD Battery (video playback) Up to 24 hours Up to 22 hours Media engines Single encode/ProRes engine More encode/ProRes engines (higher configs) What actually changes between them More cores — the M5 Pro nearly doubles CPU cores and adds GPU cores, so sustained, multi-threaded work finishe...