close
close

How Fabric Plans to Make Advanced Cryptography Ubiquitous

Fabric Cryptography, a hardware startup run by MIT and Stanford dropouts (and husband-and-wife team) Michael Gao and Tina Ju, wants to make modern cryptographic techniques like zero-knowledge proofs (which let you prove something without revealing exactly what you know) and fully homomorphic encryption (which lets you work with encrypted data without decrypting it) commonplace. The co-founders say it will ease what they see as a fundamental tension between trust and privacy at a time when companies are collecting more and more data about consumers but are increasingly unable to protect it.

To prove this, the team has developed a custom RISC-V-based integrated circuit that is optimized to run the algorithms necessary to establish zero-knowledge proofs and enable fully homomorphic encryption. Fabric calls this a “verifiable processing unit” (VPU). Existing hardware, the team argues, is simply too slow to make things like fully homomorphic encryption common, so custom integrated circuits are needed to make it happen.

A number of venture capitalists are now backing the idea, with Blockchain Capital and 1kx leading a $33 million Series A round. Offchain Labs, Polygon, and Matter Labs also participated. It follows a $6 million seed round led by Metaplanet with participation from Inflection and Liquid2 Ventures.

Image sources: Fabric

The fact that there are several crypto investors investing in crypto here is no coincidence. “We couldn’t be more excited to partner with Fabric on their mission to accelerate all crypto operations with the world’s first VPU to help create a world where privacy and verifiability are non-negotiable components of all digital systems,” said Yuan Han Li, an investor at Blockchain Capital.

While Fabric is focused on enterprises, the market that needs its solution the most is blockchain. That’s why its first-generation chip will focus on zero-knowledge proofs.

“With zero-knowledge proofs and fully homomorphic encryption, it’s really starting to look like AI in the 2010s, with exponential growth in computing power,” Gao told me. “You could say that even though not everyone understands what it is, it’s already becoming more and more common in the crypto space. And we think it’s going beyond cryptography. It’s going to be about enterprise cryptography. And so we realized that this is really the major moment, and hardware could be the ultimate accelerator to push it forward and allow cryptographers to fulfill their dreams. So Tina and I came together and started this company.”

Fabric says it has already received “tens of millions of dollars worth of pre-orders.”

Enterprises also don’t want to deal with “toxic sludge of user data,” Ju noted, because it’s a burden in most cases. So if they can keep that data encrypted even while it’s in use, or don’t even have to download it and instead exchange only what’s necessary via zero-knowledge proofs, that’s a huge win for them.

Gao is no stranger to ambitious hardware startups. He previously co-founded photonic AI startup Luminous Computing. He then co-founded bitcoin mining chip startup Katana in 2021 and took over as CEO in 2022. Katana has since changed its name to Fabric Systems. Fabric Systems is now defunct, but part of its DNA seems to live on in Fabric Cryptography.

But the pair’s experience at Luminous may have been something of a wake-up call. “We were both disappointed that, despite my personal enthusiasm for large language models, and despite the enormous potential of AI in general…, the most practical way to enter the photonics computing market seemed to be giant AI machines for targeted social media ads—a use case that barely got me out of bed in the morning,” Gao said.

Fabric is in the process of shipping its first chips, and the company plans to use the new funding to develop the next generation of its chips and scale its software and cryptography teams. In the long term, Fabric hopes to bring its chips to enterprise data centers and maybe even sell them for hyperscale clouds.