VC
Value Add VC
⚡HomePulse⚡Helpful Apps📝Blog
← Value Add PulseBIG TECHCustom silicon

OpenAI, SpaceX and the Scramble to Build Custom AI Chips -- and Escape Nvidia

From OpenAI to SpaceX, the most compute-hungry companies in tech are now designing their own silicon, intensifying pressure on Nvidia's dominance of AI hardware. OpenAI recently unveiled its first custom chip built with Broadcom, and the broader move toward in-house accelerators reflects a strategic push to cut costs, secure supply, and avoid paying Nvidia's margins on every training run and inference call.

In-house AI accelerators
Trend
Broadcom
OpenAI Partner
Nvidia GPU dependence
Target
Cost, supply, control
Goal
Google, Amazon, Meta, SpaceX
Also Building
TC
Trace Cohen
Early-stage VC & angel · Founder, New York Venture Partners
June 26, 2026
2 min read
KEY TAKEAWAYS FOR VCs & FOUNDERS
1

Custom silicon attacks Nvidia's ~75% margins -- the single biggest cost line in the AI buildout

2

Vertical integration of compute is becoming a strategic necessity, not a nice-to-have, for frontier players

3

It validates Broadcom and Marvell as the picks-and-shovels winners of the custom-accelerator wave

4

Control of chips, like control of power and data, is where durable AI moats are now being built

TC
The VC Read · Trace's TakeTrace Cohen

The number every Nvidia bull should stare at is ~75% gross margin -- because that's the prize the entire AI industry is now organizing to take back. When your biggest customers are also designing the chips to replace you, the question stops being 'will demand stay strong' and becomes 'how long until margin compresses.' The real winners of this wave are the unglamorous arms dealers -- Broadcom, Marvell, TSMC -- who get paid no matter which logo ends up on the accelerator. CUDA still protects Nvidia for now, but the strategic direction is set: in AI, owning your silicon is becoming as important as owning your model.

⚡ AI Chip Wars →🤖 AI Landscape →

The biggest names in AI are racing to design their own chips, and the collective effort is turning up the heat on Nvidia, whose GPUs have been the near-mandatory currency of the AI boom. OpenAI recently unveiled its first custom chip, developed with Broadcom, joining a list of hyperscalers and frontier companies -- from Google and Amazon to Meta and now SpaceX-adjacent efforts -- that have concluded buying all their compute from a single vendor is both expensive and strategically dangerous.

The logic is straightforward economics. Nvidia commands gross margins that have run around 75%, meaning every dollar of GPU spend carries a large premium that flows straight to Nvidia's bottom line. For companies spending tens of billions a year on compute, even partial substitution with in-house accelerators tuned to their specific workloads can save enormous sums and, just as importantly, free them from the supply queues and allocation politics that have defined the GPU shortage.

The pattern is now industry-wide. Google has its TPUs, Amazon its Trainium and Inferentia, Meta its MTIA line, and OpenAI's Broadcom collaboration brings the most prominent model lab into the custom-silicon club. The enabling players are the merchant chip designers -- Broadcom and Marvell chief among them -- that provide the design IP and packaging expertise to turn a software company's requirements into working silicon. They, alongside TSMC's manufacturing, are quietly becoming the indispensable layer beneath the AI economy.

“Nvidia commands gross margins that have run around 75%, meaning every dollar of GPU spend carries a large premium that flows straight to Nvidia's bottom line.”

This is not Nvidia's death notice -- far from it. Its CUDA software ecosystem remains a deep moat, its newest GPUs still set the performance bar, and demand continues to outstrip supply. Custom chips tend to target specific, high-volume workloads like inference rather than the full breadth of training, so the near-term effect is margin pressure and diversification rather than displacement. But the direction of travel is unmistakable: the largest buyers are systematically reducing their dependence.

The strategic read connects to a broader theme running across the AI landscape -- that durable advantage is migrating to whoever controls the physical layer. Just as Amazon and rivals are racing to lock up data-center capacity and power, controlling your own silicon is a way to own your cost structure and your destiny rather than renting them. It's the same instinct that has Musk's world buying optical-interconnect startups and building data centers: vertical integration of compute.

The bear case for the challengers is that designing competitive chips is brutally hard, expensive, and slow, and many in-house efforts underperform Nvidia's roadmap or arrive late. CUDA lock-in is real, and software portability remains a genuine obstacle. What to watch: how much of their own inference the hyperscalers actually move onto custom silicon, whether OpenAI's Broadcom chip ships at scale, and whether Nvidia's margins finally compress as the buyers it depends on become its competitors.

ShareXLinkedInEmail
More onSpaceX →OpenAI →Nvidia →

Originally reported by TechCrunch. Analysis and editorial commentary by Value Add Pulse.

← Back to Pulse

Markets Now

live
SPCX▲+0.73%
$233.10
CBRS▲+0.66%
$259.80
SPY▲+0.07%
5,952.40
QQQ▲+0.12%
20,061.30
NVDA▼-0.98%
$152.10
MSFT▲+0.29%
$482.60
GOOGL▲+0.34%
$209.10
META▲+0.23%
$655.40

Read Next

BIG TECHMesh acquisition

FTC Clears Musk to Acquire Mesh, the SpaceX-Alumni Optical-Interconnect Startup

The FTC expedited its antitrust review and cleared Elon Musk to acquire Mesh Optical Technologies, a startup founded by three former SpaceX engineers that builds high-speed optical transceivers for data-center communications. Mesh emerged from stealth in February with a $50M Series A led by Thrive Capital; the acquisition slots directly into Musk's push to make his AI data centers faster and more energy-efficient.

BIG TECH0.7nm node

IBM Debuts the World's First Sub-1-Nanometer Chip Technology at the 0.7nm Node

IBM unveiled what it calls the world's first sub-1-nanometer chip technology -- a new transistor architecture at the 0.7nm (7-angstrom) node that packs nearly 100 billion transistors onto a fingernail-sized chip. The breakthrough, built on IBM's 3D nanostack architecture, promises up to 50% higher performance or 70% greater energy efficiency versus its 2nm chips, with a path to production within five years.

BIG TECH$13B / $48B total

Amazon Commits Another $13B to India for AI and Cloud, Pushing Its Bet to $48B by 2030

Amazon CEO Andy Jassy, meeting Prime Minister Narendra Modi in New Delhi, announced an additional $13 billion investment in India for AI and cloud infrastructure by 2030, lifting Amazon's total committed India spend to about $48 billion -- more than $21 billion of it earmarked for AI and cloud. The money expands AWS data-center capacity in Mumbai and Hyderabad as the hyperscalers race to own the infrastructure layer of the world's fastest-growing internet market.

@Trace_Cohen·t@nyvp.com