Etched Emerges From Stealth With $800 Million in Funding and $1 Billion in Customer Contracts
AI inference hardware startup Etched emerged from stealth on Wednesday, announcing $800 million in raised funding and more than $1 billion in signed customer contracts. The company completed its A0 tapeout and built its first full racks, which are scheduled to ship to enterprise customers this summer. Etched says early tests show the custom silicon achieves state-of-the-art throughput, latency, and power efficiency for running large language models.
Founded by Gavin Uberti and Robert Wachen, the startup differentiated its architecture by hardcoding attention mechanisms directly into silicon to address the thermal and memory bandwidth bottlenecks that typically limit traditional GPUs. The hardware stack, which targets both chip compute and cluster-scale memory transfer, has attracted investment from Jane Street, a venture affiliate of Taiwan Semiconductor Manufacturing Co., and a wide network of AI researchers and industry veterans.
From the sources (25 posts)
@kimmonismusEtched is coming out of stealth with $800M raised, $1B+ in customer contracts, first racks shipping this summer, and claims of SOTA inference throughput, latency, and power efficiency. But holy, look who backed the funding. The who-is-who
@patrick_oshagThree years ago, two Harvard dropouts set out to build a better AI chip than the largest companies in the world. Almost everyone I called at the time said it was impossible. Today, Etched (@Etched) comes out of stealth with $800M total r
@lpolovetsRT @patrick_oshag: Three years ago, two Harvard dropouts set out to build a better AI chip than the largest companies in the world. Almos…
@mtsliveSITUATION DETECTED: Etched, a startup building custom AI inference chips, has emerged from stealth, announcing $1B+ in customer contracts and $800M raised after a successful chip tapeout.
@techmemeAI chip startup Etched says it has raised $800M from investors including Jane Street and a venture firm linked to TSMC, and has signed sales contracts worth $1B (@dinabass / Bloomberg) (Visit Techmeme dot com for the link and full context!
@edludlowRT @business: AI chip startup Etched says it has raise $800 million and has $1 billion in sales contracts.
@stevehou👀
@brendanfoody@UbertiGavin, @robertwachen, and @czhu1729 are some of the best founders in Silicon Valley. Excited to start using @Etched chips this summer 🎉
@jasonrshumanThis company is going to put a dent in the universe. Off the charts performance. Insane time to market. Eye popping customer demand. World-Class Team.
@ml_hardwareRT @Etched: We're coming out of stealth. We've built our first racks after a successful A0 tapeout, $1B+ in customer contracts, and $800m…
@tri_daoIt's wild how quickly Etched designed and got the chips out, all within 2 years. They went deep, hardcoding attention into silicon and getting very high MFU. This kind of hardware tailored made for LLM inference is soon gonna bring cost of
@millionintChips are a substrate for computation that the layer of intelligence we're all building relies on. Like every layer of the stack, it requires dedication, hard work, taking risks and innovating. Etched Sohu has all those in spades and I'm
@aiexplainedytI remember saying on the channel, in Dec 2023, 'if Etched take over the world, you heard it here first.' They do seem to be somewhat on their way.
@jubertiThe world needs faster and cheaper inference compute, and it's amazing to see Etched releasing their 10x solution. Great work @UbertiGavin and team!!!
@mathemagic1anBeen watching demos of Etched for years now and still blows me away every time Crazy ambitious strategy to scale inference. Cluster-scale memory w/ latency of SRAM but cheap like HBM. Proud to be a backer of these guys, absolute rocketshi
@mattshumer_Nvidia is worth $5T because the world realized inference is the new oil. If Etched’s chips are meaningfully better than Nvidia’s for inference… What is Etched worth? This is a worldview-shattering event.
@mattshumer_I’m a very early investor in Etched, as I’ve believed for years that inference will be the biggest market we’ve ever seen. Incredibly impressed by the team’s execution and so excited to use models powered by their hardware!
@himanshustwtsnew ai inference hardware company, just dropped!
@mattshumer_Nvidia has some serious competition coming their way.
@jasonrshumanRT @patrick_oshag: Inference will never be the same: Etched invented two new ways to massively improve compute cost, speed, and per watt ef…
@edludlowRT @austinsemis: At the end of the interview, Gavin mentions @Etched VP of Platform Brian Loiler, who was an early employee at Nvidia and h…
@colossusmagRT @patrick_oshag: Three years ago, two Harvard dropouts set out to build a better AI chip than the largest companies in the world. Almos…
@dmdohanBig if true, and afaict it's true congrats to the Etched team on getting inference running on their new chips!
@jordanjacobs10RT @tri_dao: It's wild how quickly Etched designed and got the chips out, all within 2 years. They went deep, hardcoding attention into sil…
@jordanjacobs10RT @Etched: Vertically integrating the full inference hardware stack is extremely hard. Hear @UbertiGavin and @robertwachen share three ye…