This report is from this week’s The Tech Download newsletter. like what you see? you can subscribe Here.
NVIDIA It has cemented itself at the center of the AI boom with a monopoly on the most powerful chips to train and run models, but a growing crop of startups is ready to challenge the company’s supremacy.
And increasingly, investors are putting huge sums of money behind them. According to Dealroom, in 2026, AI chip startups are expected to raise $8.3 billion in funding globally. Barring a near-complete collapse of the market, record sums are expected to flow into the sector this year.
So what’s causing the spike?
While Nvidia’s graphics processing units (GPUs) – which were originally designed for gaming – have been effectively reused for AI training, the focus is now shifting to the most efficient ways of actually deploying the technology in applications, known as AI inference.
The logic of these chip upstarts is this: GPUs were not purpose-designed for AI, and so, novel system architectures will bring huge savings in energy and cost.
“Now guesswork dominates, and the existing GPU architecture was not built for it in the ways that make the most sense at scale,” Patrick Schneider-Sikorsky, director of the NATO Innovation Fund (NIF), which has invested in UK AI chip startup Fractile, tells me.
Nvidia, which has huge advantages as the world’s most valuable company with a nearly unlimited supply of cash, is still racing to develop new chips to power AI.
In December, the company acquired $20 billion in assets from AI inference startup Grok and announced in March it had invested $4 billion in two companies developing photonics technology.
The chip giant spent more than $18 billion on research and development in its most recent full fiscal year ending January 2026.
startup funding
But that hasn’t stopped investors from putting money behind new, and often untested at scale, AI chip technology.
In the US – where many of the largest rounds have been raised – Cerebrus Systems raised $1 billion in February, and $500 million is set to be raised in 2026 for MatX, Ayar Labs and Etched.
European companies have raised comparatively smaller amounts, but Accelera and Olix have both raised about $200 million this year. Euclid and optolysis, among others According to reports, they told me they are planning at least $100 million in 2026, like Fractile and Arago.
“It’s no longer a great bet,” said Carlos Espinal, managing partner at European VC Seedcamp, which backs chip startup Vare Computing. “It’s becoming a core part of how people think about AI infrastructure.”
latest updates
Anthropic and OpenAI both announced major UK expansion plans. Anthropic unveiled New office space for 800 people, while OpenAI said it will open its first permanent London office with capacity for more than 500 team members.
TSMC on Thursday reported a 58% rise in first-quarter profit. A new record was set, beating estimates due to strong demand for artificial intelligence chips.
OpenAI abandons plans to rent capacity directly from Norwegian data centers, The additional calculations come days after Microsoft confirmed it had halted a similar project in the UK, saying the ChatGPAT maker would rent capacity instead. MicrosoftThe company told CNBC.
Amazon said on Tuesday it would acquire Globalstar The deal is worth about $11.57 billion as it seeks to boost its nascent LEO satellite internet business and compete with Elon Musk’s SpaceX.
uber agreed to buy an additional 4.5% shares of German food delivery firm Delivery Hero on Friday From the company’s largest shareholder process.
stock of the week
ASML stock declined after the company’s earnings report on Wednesday.
asml Despite raising its sales forecast for 2026 and beating first-quarter revenue and profit expectations, the stock has been falling since the results were announced on Wednesday.
The astronomical expectations surrounding the AI boom were likely to cause a negative reaction, along with tighter restrictions on export controls, leading to a decline in the percentage of net sales to China.
