Nvidia asks TSMC for more chips

Presented by

☕ Good morning,

There's something almost zen about Nvidia's position right now. They can't fulfil demand. Supply is tight. Competitors are circling. Stock sceptics are questioning the AI boom. And Jensen Huang's response is essentially "we need more chips because demand keeps growing faster than we can keep up."

^{That's not a problem, that's a flex disguised as logistics.}

_{—Here’s to the first sip.}

TODAY IN AI
OpenAI urges US to expand AI tax credits

OpenAI is asking the US government to expand the Chips Act tax credit so it helps more than just chipmakers. The company wants the 35% tax break to also cover AI data centers, servers, and power equipment like transformers, basically everything needed to run big AI systems.

The idea is to make building AI infrastructure cheaper and faster, while attracting more private investment and keeping up with China’s growing AI industry.

OpenAI plans to spend around $1.4 trillion over the next eight years on chips and data centers, so this tax credit would help lower some of those costs.

There was some noise earlier last week after OpenAI’s CFO mentioned the idea of government-backed loans. Sam Altman clarified that OpenAI isn’t asking for a bailout, just incentives that help the entire AI ecosystem grow.

If this change goes through, it could make AI expansion in the US a lot more affordable and speed up how fast companies can build the hardware behind the technology.

TECH BARISTA
Nvidia wants more chips from TSMC

Nvidia CEO Jensen Huang says AI demand isn’t slowing down anytime soon in fact, it’s getting stronger every month. While visiting TSMC’s headquarters in Taiwan, Huang confirmed that he’s asked the chipmaker for more wafer supply to keep up with the surge in orders for Nvidia’s AI chips.

He also mentioned that Nvidia’s memory partners Samsung, SK Hynix, and Micron have all expanded production to support the growing demand. TSMC’s CEO C.C. Wei backed that up, saying the company expects to hit record sales again this year, despite supply still being tight.

Even with recent dips in tech stocks and some skepticism around the AI boom, Huang remains confident, saying Nvidia’s growth is accelerating. Qualcomm’s CEO Cristiano Amon echoed the same sentiment earlier this week, saying people are still underestimating how massive AI will become.

Huang also took a moment to credit TSMC for Nvidia’s success, saying, “No TSMC, no Nvidia.” The partnership is clearly vital, especially as other chipmakers are trying to compete for the same limited capacity.

PRESENTED BY LINDY AI

The Simplest Way to Create and Launch AI Agents and Apps

You know that AI can help you automate your work, but you just don't know how to get started.

With Lindy, you can build AI agents and apps in minutes simply by describing what you want in plain English.

→ "Create a booking platform for my business."
→ "Automate my sales outreach."
→ "Create a weekly summary about each employee's performance and send it as an email."

From inbound lead qualification to AI-powered customer support and full-blown apps, Lindy has hundreds of agents that are ready to work for you 24/7/365.

Stop doing repetitive tasks manually. Let Lindy automate workflows, save time, and grow your business

Get $20 Worth of Free Credits

GADGETS BARISTA
Poco Pad M1 spotted

The Poco Pad M1 just showed up on Geekbench, confirming it’s powered by the Snapdragon 7s Gen 4 chip with 8GB of RAM. The test unit was running Android 15, which suggests the final version might not ship with Android 16 out of the box, a bit disappointing considering the newer version has been out for months.

According to leaks, this isn’t an all-new device but rather a rebranded Redmi Pad 2 Pro, which launched earlier this year. That means we’re looking at a 12.1-inch LCD display with a 2560x1600 resolution, 120Hz refresh rate, 8MP front and rear cameras, and a 12,000mAh battery that supports 33W fast charging and 27W reverse charging.

No word yet on the official launch date, but it’s likely to arrive alongside the upcoming Poco F8 Pro and Poco F8 Ultra.

STARTUP BAR
Inception raises $50M to build faster diffusion-based AI

A new AI startup called Inception popped up with a fresh idea and a massive $50 million seed round to back it. The company is led by Stanford professor Stefano Ermon, who’s betting on diffusion models instead of the usual language model approach everyone’s using right now.

If you’ve heard of Stable Diffusion or Midjourney, you already know what diffusion models are about. They refine outputs step by step instead of generating things word by word like GPT or Gemini. Inception’s taking that same concept and applying it to software development and coding tasks. Their latest model, Mercury, is already being plugged into dev tools like ProxyAI, Buildglare, and Kilo Code.

The big difference is speed. Ermon says diffusion models can run way faster because they don’t need to process data sequentially they can handle multiple operations at once. Inception claims over 1,000 tokens per second, which crushes the usual LLM benchmarks.

The funding came from big names like Menlo Ventures, Microsoft’s M12, Snowflake Ventures, Databricks, Nvidia’s NVentures, plus backing from Andrew Ng and Andrej Karpathy.