Architect Labs raises $24M to challenge Broadcom in chip design

Key Takeaways

• Architect Labs raised $24 million in seed funding to build AI tools that speed up custom chip design
• The startup targets a market where chip design takes two years and costs hundreds of millions of dollars
• Google DeepMind's Jeff Dean and executives from OpenAI and Nvidia invested in the company

Architect Labs has raised $24 million in seed funding to build AI tools that accelerate custom chip design. The Palo Alto startup is targeting the lucrative market currently dominated by Broadcom and Marvell, where designing a single chip takes roughly two years and costs hundreds of millions of dollars.

Kindred Ventures led the round, with participation from TQ Ventures, Race Capital, and Together Fund. The investor list includes notable names from the AI industry: Google DeepMind Chief Scientist Jeff Dean, along with executives from OpenAI and Nvidia.

What problem is Architect Labs solving?

Custom chips, or ASICs, have become essential for cloud giants like Amazon and Google seeking alternatives to Nvidia's expensive GPUs. Broadcom designs Google's TPUs. Marvell works with Amazon on custom AWS chips. These relationships generate tens of billions of dollars in revenue.

But the design process remains painfully slow. Companies spend years and hundreds of millions in labor and R&D before a chip ever reaches manufacturing. Architect Labs believes AI can compress that timeline and open custom silicon to companies that couldn't previously afford it.

“Their biggest problem today is not necessarily the backend execution or the layout. Their biggest thing is how can I take this workload that I want to deliver to the world, whether it be AI or robotics or anything like that, and how can I build the chip architecture.”

— Ebrahim Hussain, Co-founder, Architect Labs

Who are the founders?

Ebrahim Hussain and Aaditya Subedi founded the company. They've assembled a team of about 18 people, split between machine learning engineers and hardware specialists. The goal, according to Subedi, is to make chip design as accessible as Taiwan's TSMC has made chip manufacturing.

That's an ambitious comparison. TSMC democratized manufacturing by allowing fabless companies to design chips without owning billion-dollar fabs. Architect wants to do the same for the design side, letting software companies create custom silicon without maintaining massive chip architecture teams.

Who will use these AI design tools?

Architect plans to target two customer segments. First, existing chip companies that want to accelerate their design workflows. Second, software companies that could benefit from custom chips but lack the expertise or budget to build them traditionally.

The second group is more interesting. A startup building robotics software, for instance, might achieve significant performance gains with purpose-built silicon. Today, that would require hiring an expensive chip team or partnering with Broadcom or Marvell. AI design tools could change that calculus.

The competitive landscape

Architect Labs isn't alone in applying AI to chip design. Google has published research on using machine learning for chip layout. Synopsys, the established EDA tool vendor, has added AI features to its software. Several startups are attacking different parts of the design workflow.

The difference is ambition. Most AI-for-chip tools focus on optimizing specific steps like layout or verification. Architect is going after the harder problem: translating a software workload into chip architecture. That's the creative, expensive part that keeps design cycles so long.

Whether AI can actually perform this kind of high-level architectural work remains unproven. Chip design combines deep domain expertise, physical constraints, and engineering tradeoffs that are difficult to automate. But $24 million in seed funding from investors who understand both AI and semiconductors suggests the approach is worth testing.

Frequently Asked Questions

How much funding did Architect Labs raise?

Architect Labs raised $24 million in seed funding led by Kindred Ventures, with participation from TQ Ventures, Race Capital, and Together Fund.

Who invested in Architect Labs?

Beyond the venture firms, Google DeepMind Chief Scientist Jeff Dean invested, along with executives from OpenAI and Nvidia.

What does Architect Labs do?

The company builds AI tools to accelerate custom chip design, targeting a process that currently takes about two years and costs hundreds of millions of dollars.

How does Architect Labs compete with Broadcom and Marvell?

Rather than designing chips directly, Architect aims to sell AI tools that make design faster and cheaper, potentially enabling more companies to create custom silicon.

Source: Tech-Economic Times / ET