Read time: 2 minutes 54 seconds

4 weeks ago, we announced a temporary shift from weekly emails to 1x email every 2x weeks, to prioritise long-term vision over short-term progress.

Not an easy decision. 

For those doing napkin math at home, that’s a 50% revenue haircut. 
For an early-stage, bootstrapped, indie newsletter business. 
For ~3 months. 

Ouch!

But there’s a magical nervous momentum that happens when you give yourself permission to entertain your favourite side-quests, despite the opportunity cost.

2 weeks from today, we’re launching something fun. 

Well, technically 3 things:

1. A rebrand.

2. A new teammate.

3. A creative experiment.

6 months in the works. 2 weeks from launch.

Buzzing.

— Tom

P.S. We wrote today’s article before the Yen carry trade tanked $NVIDIA stock. The strategy still stands strong. No recency bias here. But still, this article is not financial advice (obviously).

P.S.S. Pretty unique initiatives from our sponsors today, including some involvement from yours truly! Enjoy.

Chess Move

The what: A TLDR explanation of the strategy

Sometimes easy to forget, Silicon Valley gets its name from the homonymous element that conducts electricity, making all technology innovation possible.

The business of semiconductors – the end product of silicon manufacturing – is a complex one for many reasons: 

• Buyers are often also suppliers 

• Customers aren’t individuals, they’re giant multinational conglomerates

• Unpredictable supply shocks create massive shortages in downstream shipments

• National governments have vested interests influencing global control of semiconductors

One company has navigated those challenges better than anyone over the past half-decade:

Nvidia. 

Rewind to 2016 and Nvidia was starting their transformation from an average semiconductor manufacturer in a crowded landscape to a global aristocrat. Their Graphics Processing Units (GPUs) were being widely adopted by the gaming community due to their ability to produce high-resolution visuals without latency. 

They were also coveted by another, lesser known population: bitcoin miners. Around that time, the cryptocurrency community realised that Nvidia’s powerful GPU chips could be used to mine precious bitcoins. 

At that time, this segment of their business made up 80% of Nvidia’s total revenue. 

In the background, Nvidia had spent billions of dollars developing the DGX-1, an ultra-powerful processor embedded with chips designed specifically for machine and deep learning. 

What did Nvidia do when the DGX was ready for primetime? 

They decided to give it away for free to OpenAI, who was only 6-months old at the time.

Breakdown

The how: The strategic playbook boiled down to 3x key takeaways

1.  Generosity gets rewarded

The gesture between two longtime friends sent shockwaves around Silicon Valley and the entire industry.  

According to their 2016 10-K, Nvidia had more than 6,500 full-time employees assigned to research, and they spent more than $1.3 billion on that function throughout the year.

How could they be willing to give away a processor that retailed for $129K and took literal years and billions to make? 

Nvidia saw the move as an investment, not a gift.

They understood that the faster AI companies like OpenAI companies could develop their Generative Pre-Trained Transformers (GPTs) and other transformers, the faster those companies could become huge customers of Nvidia. 

Nvidia also gained valuable intel around the business needs of OpenAI and how they could improve their products in preparation for more AI customers. 

OpenAI would go on to be ~acquired(ish) by Microsoft in 2023.

Nivida’s largest customer last year? 

You guessed it. 

Microsoft. 

2. Advertise your kindness

Up until 2016, Nvidia had a difficult time separating themselves from the pack of other ”Fabless” chip companies. 

Their GPUs were perceived as valuable for specific use cases like gaming or video content services, but perhaps not strong enough to support the enterprise requirements that were typically delivered by incumbents Qualcomm and Broadcom. 

Intel even sanctioned a research study during the launch of their Xeon Phi, a competitor to the DGX, to highlight the futility of Nvidia’s flagship product. 

Jensen Huang knew it was important to control the narrative around the capabilities of his supercomputer. Huang also knew the timing of his photo opp was precarious – only days before the annual Intel Developer Conference, where Intel was set to unveil their Xeon.

This kicked-off a roadshow of sorts, with Huang taking a proactive role in transforming Nvidia’s image from a supplier to the gaming industry into a global superpower. 

He developed close personal relationships with the other leaders of the “Magnificent 7,” along with bosses at: 

• Mercedes

• Volkswagen

• MIT

• IBM

• Stanford

• University of Toronto

Nvidia is now a global phenomenon.

And Huang is tech’s Taylor Swift. 

3. Recurring gifts = recurring results

The relationship with OpenAI yielded promising early results. By the time OpenAI unveiled GPT-3 (ChatGPT) in 2022, Nvidia was certain they had a winner on their hands.

But that didn’t stop them from extending more generosity. 

Along the way, they co-developed more than 200 research papers with OpenAI, where Nvidia contributed:

• Hardware developments

• Software integrations

• Dedicated headcount

They now boast similar arrangements with IBM and Meta, amongst others. 

And Nvidia is poised to follow their playbook from the initial DGX-1 as they prepare to launch their much anticipated HGX-100 and 200 products this year.  

By constantly gifting their latest product to their strategic partners, Nvidia has helped to create an AI ecosystem that can only succeed with the support of chipmakers… like Nvidia. 

Early returns on that strategy seem to be promising. Or even historic. Depends who you ask.

Rabbit Hole

The where: 3x high-signal resources to learn more

Our Tech Stack

Before we sign-off, here are 3 toys in the Strategy Breakdowns toolbox that we use, love, and recommend trying:

See the full set of tools we use at Strategy Breakdowns here.