AI Is Now Hunting for Better Chips — And It Could Reshape the Hardware Powering Your Business

The race to build faster, cooler, and more efficient AI chips just got a new contender — and this one uses artificial intelligence to do the searching.

Discovered Materials, a startup focused on AI-driven materials discovery, has raised $9 million in funding to accelerate the hunt for novel materials that could be used to build more efficient semiconductor chips. The round signals growing investor confidence that AI can do more than just run on chips — it can help design the next generation of them. The news was reported by Tim Fernholz at TechCrunch AI.

What Discovered Materials Is Actually Doing

The core challenge in chip design is heat. As processors get faster and more powerful, they generate more heat, which limits performance, increases energy costs, and shortens hardware lifespan. Finding materials that conduct electricity more efficiently while generating less heat has been a decades-long scientific problem.

Discovered Materials is applying AI to that search in a methodical, iterative way — what TechCrunch describes as a kind of "whack-a-mole" approach. The company uses AI models to rapidly identify promising material candidates, test them computationally, eliminate the ones that don't work, and move on to the next option. This dramatically compresses what would otherwise be years of laboratory trial-and-error into a much faster cycle of discovery.

The $9 million in funding will allow the company to expand that search and push more candidate materials through its pipeline.

Why This Matters Beyond the Lab

It is easy to read this story as purely a deep-tech or materials science item. It is not. The downstream effects of breakthroughs in chip materials will be felt across the entire business technology stack.

Here is the chain of impact worth paying attention to:

More efficient chips mean AI inference and training become cheaper. That translates directly to lower API costs for businesses using AI tools, more capable models running on smaller hardware, and faster response times in production applications. It also means the data centers that power cloud services — the ones your business depends on daily — become less expensive to operate and less carbon-intensive to run.

For small and mid-sized businesses especially, the cost of AI access remains a real barrier. When a startup asks whether they can afford to run AI agents continuously or process large data sets, the answer often comes back to compute costs. Cheaper, cooler chips do not just benefit hyperscalers like Microsoft or Google. They flow downstream to every SaaS product, cloud service, and AI tools for business that SMBs rely on.

The Broader Trend: AI Accelerating Its Own Infrastructure

What Discovered Materials represents is a broader pattern that deserves serious attention from business strategists. AI is increasingly being turned on the problems that limit AI itself.

We have seen this in chip architecture, where AI-assisted design tools are helping engineers create more efficient processor layouts. We are seeing it in energy research, where AI models are being applied to battery chemistry and grid optimization. And now we are seeing it in materials science, where the combinatorial complexity of finding new semiconducting compounds is exactly the kind of problem machine learning is well suited to tackle.

This recursive quality — AI improving AI's own hardware and infrastructure — is likely to compress the timeline on capabilities that previously seemed years or decades away. Business leaders who are still treating AI as a future consideration rather than a current operational priority may find the gap widens faster than expected.

What Business Teams Should Take Away

The practical takeaway here is not that businesses need to understand materials science. It is that the infrastructure costs and capability ceilings that currently shape what AI can do for your organization are not fixed. They are being actively worked on by well-funded teams using AI to solve hard problems.

That means the AI automation tools available to your team in two or three years will likely be significantly more powerful and affordable than what exists today. Planning your AI adoption strategy with that trajectory in mind — rather than benchmarking only against current limitations — is the smarter move.

Platforms like WRRK.ai are built with that future in mind, helping business teams put AI to work today while remaining adaptable as the underlying infrastructure continues to evolve.

Original reporting by Tim Fernholz, published August 10, 2026, on TechCrunch AI. Read the original article at TechCrunch.

Frequently Asked Questions

What is Discovered Materials and what did they raise funding for?

Discovered Materials is a startup using artificial intelligence to identify novel materials that could be used to build more efficient semiconductor chips. The company raised $9 million to expand its AI-driven materials discovery pipeline, which uses iterative computational testing to rapidly evaluate and eliminate candidate materials in the search for better chip components.

How does chip efficiency affect the cost of AI for businesses?

More efficient chips reduce the energy and cooling requirements of the data centers that run AI models. This lowers the operating costs for cloud providers and AI companies, which over time translates to lower API pricing and more accessible AI services for businesses of all sizes, including small and mid-sized companies that are sensitive to compute costs.

Why is AI being used to discover new materials for chips?

Finding materials with the right electrical and thermal properties for advanced chips involves evaluating an enormous number of possible chemical compounds. AI can screen and test these candidates computationally far faster than traditional laboratory methods, compressing years of research into a much shorter discovery cycle.

AI Is Now Hunting for Better Chips — And It Could Reshape the Hardware Powering Your Business