Tenstorrent, the company vying to compete with Nvidia is now valued at $2.6 billion after Amazon founder Jeff Bezos joined South Korean tech giant Samsung in a $700 million bet on the AI chip startup.

Bloomberg reported this funding as a significant development in the AI industry. Founder and semiconductor pioneer Jim Keller said in an interview that the feat was achieved in a funding round led by South Korea’s AFW Partners and Samsung Securities, aimed at raising capital for Tenstorrent. The AI startup hopes to create a chip to try to break Nvidia’s stranglehold on the AI business.

Tenstorrent and other smaller firms are emerging, pressuring Nvidia

Bezos Expeditions joined LG Electronics and Fidelity in that financing, betting on Keller’s pedigree and the booming opportunity in AI tech. The AI startup will use the funding to expand its engineering team, invest in its global supply chain, and build large AI training servers to help demonstrate its technology.

This comes as smaller companies are sprouting up trying to snatch market share from Nvidia’s power-hungry chip. The startups are coming in with the gospel of more power and cost efficiency in AI technology.

Tenstorrent is one of these new startups. As part of these new startups, Tenstorrent is Nvidia’s neighbor in Santa Clara, California, and is one of the many engineering solutions now aimed at delivering a more affordable path to AI development.

Nvidia favors complex and pricey components like high bandwidth memory, but Tenstorrent is built on open-source and commonplace technology which avoids the Nvidia route.

“You cannot beat Nvidia if you use HBM, because Nvidia buys the most HBM and has a cost advantage,” Keller said. “But they will never be able to bring the price down the way HBM is built into their products and their sockets,” added Keller.

In order to give customers value for money, Nvidia promises that all parts work better because they were designed in concert. As a result, the chipmaker offers developers a full suite of proprietary technology covering everything from the chips to interconnects and even data center layouts.

Nvidia’s rivals like Advanced Micro Devices (AMD) and startup Tenstorrent are looking to give developers chips with greater interoperability with other technology providers, whether through shared industry standards or opening their designs for others to use.

Based on an open standard called RISC-V, Tenstorrent is a proponent of the alternative kind of logic processor. The technology poses a significant challenge to Arm Holdings, and Jim Keller—renowned for his silicon design contributions at Apple, AMD, and Tesla—is a strong advocate for its adoption.

“Open source helps you build a bigger platform. It attracts engineers, and yes, it’s a little bit of a passion project.”

– Keller

“In the past, I worked with proprietary technology and it was really tough,” added Keller.

Keller’s approach differs from Nvidia’s strategy of providing a vertically integrated solution from chip to data center design.

Tenstorrent to release new AI processor

Tenstorrent like its peers, RISC-V and Japanese partners Rapdius Corp, still has much to do to prove themselves. With the competition rife, Nvidia has tens of billions of dollars of data center revenue each quarter, whilst Tenstorrent to date, the nascent company has signed contracts with customers totaling nearly $150 million.

Keller also revealed that Tenstorrent aims to release a new AI processor every two years whilst Nvidia on the other hand, aims to refresh its AI chip offerings on an annual cycle according to its boss Jensen Haung’s statement in June.

Managing director Bonil Koo said AFW Partners invested after hearing positive feedback from Korean companies already collaborating with Tenstorrent such as LG.

Export Development Canada, Healthcare of Ontario Pension Plan, Hyundai Motor Group, and Baille Gifford were the other investors in this funding round.

Tenstorrent said Samsung Electronics and Taiwan Semiconductor Manufacturing Co. (TSMC) will make the next round of funding after GlobalFoundries made the first iteration. The startup has begun designing for cutting-edge 2-nanometer fabrication.

From Zero to Web3 Pro: Your 90-Day Career Launch Plan