- The Prohuman
- Posts
- Nvidia makes room for custom AI chips
Nvidia makes room for custom AI chips
Plus: Grok gets 1.7 million potential users
Rimsha Bhardwaj, Hasan Toor & Ihtesham Haider
September 01, 2026
In partnership with
Hello, Prohuman
Today, we will talk about these stories:
Nvidia wants more than chips
Grok enters the Pentagon’s AI stack
Nvidia adapts to the custom chip push
Stop Paying for 10 Tools. One AI Does It All.
Most e-commerce sellers are running their store across 6 to 10 separate tools — and spending more time managing software than growing their business. StoreClaw replaces your entire stack with one autonomous AI engine that monitors competitors, optimizes listings, automates marketing, and tracks real profit across Shopify, Amazon, and beyond.
It doesn't wait for you to ask. It runs 24/7 in the background, so you wake up to a full dashboard instead of a list of things you forgot to check.
Connect your store, and StoreClaw gets to work — no prompts, no complex setup, no six-app stack.
Free to start. No credit card required.
Nvidia wants custom chips on its network
Image Credits: NVIDIA.
Nvidia just put $3.5 billion into MediaTek.
The two companies are expanding their partnership across data centers, PCs, and cars, with MediaTek adopting Nvidia’s NVLink Fusion platform for custom AI accelerators. That gives cloud providers and model developers a way to build their own XPUs while keeping them connected to Nvidia’s rack-scale systems.
I think that is the important part.
Large AI customers increasingly want custom silicon for cost, power, and performance reasons, so Nvidia has an incentive to make those chips work comfortably inside its infrastructure. If customers stop buying every accelerator from Nvidia, the company can still sell the networking, memory architecture, software, and rack design around them.
A server rack full of different chips could still carry Nvidia connections between each board.
MediaTek also gains a larger role in AI infrastructure while extending its Nvidia work in PCs and automotive systems.
The question is how much of the AI data center Nvidia can keep controlling when the main compute chip comes from someone else.
The military’s AI platform is getting crowded
Grok just got a very large workplace.
The Department of War has added Starshield AI’s Grok for Government to GenAI.mil, its internal generative AI platform for military personnel. The system is approved for Controlled Unclassified Information at Impact Level 5.
The bigger number is 1.7 million.
That is how many unique users GenAI.mil has onboarded in nine months, out of more than 3 million department personnel. At that scale, adding another model is less about experimentation and more about making AI part of routine work, from acquisition research to supply-chain planning.
I think the vendor mix matters too. The department says Grok helps reduce vendor lock-in, which suggests it wants several frontier models available rather than one provider becoming the default.
Picture a logistics officer opening the same government portal and choosing between different AI systems for the day’s work.
The next question is which model employees actually keep choosing.
Nvidia wants the layer above its GPUs
Image Credits: The Economic Times
Nvidia may spend $12.9 billion to buy Hugging Face, the platform where thousands of AI models, libraries, and developer tools already live.
That timing matters.
OpenAI, Google, Anthropic, and others are working on their own chips because paying Nvidia for compute remains one of their largest costs. OpenAI’s new chip reportedly processes tokens 3.6 times faster than Nvidia’s GB300.
Nvidia sees the pressure.
Owning Hugging Face would give Nvidia much closer contact with the developers deciding which models, tools, and hardware become standard across AI projects. It could also make Hugging Face less neutral, especially for startups whose work suddenly sits inside a major supplier’s corporate portfolio.
You can almost hear the server racks humming as both sides move closer to controlling more of the same production chain.
The open question is whether developers keep treating Hugging Face as their home once Nvidia owns the keys.
Prohuman team
Covers emerging technology, AI models, and the people building the next layer of the internet. |  Founder |
Writes about how new interfaces, reasoning models, and automation are reshaping human work. |  Founder |
Free Guides
Explore our free guides and products to get into AI and master it.
All of them are free to access and would stay free for you.
Feeling generous?
You know someone who loves breakthroughs as much as you do.
Share The Prohuman it’s how smart people stay one update ahead.