CoreWeave got another investment! This startup company, known as the "AI computing scalper," has doubled its revenue 70 times in three years! What started out as a good mining company has become the world's fastest-growing cloud computing provider.
Nvidia's heavily invested private cloud computing provider CoreWeave has also received institutional investments such as Fidelity, valued at up to $7 billion.
Only four months have passed since their first equity financing this year, but the valuation has risen 3-4 times!
According to statistics from public channels, the private cloud computing provider has completed two rounds of financing from April this year to now, plus this one under way, the annual content financing will reach three times, totaling more than 2.7 billion US dollars!
In April this year, they completed a round of equity financing amounting to 420 million yuan. In August, they used their Nvidia H100 GPU as collateral to obtain a $2.3 billion loan directly from financial institutions.
"Our explosive growth trajectory has been recognized by top institutional investors, and this transaction highlights our differentiation of market-leading results, significant technology edge and strong customer adoption," Michael Intrator, CoreWeave co-founder and CEO, said in an emailed statement.
The AI industry is at a turning point, and the company plays a central role by providing customers with the "most differentiated" AI infrastructure.
Why is this five-year-old private cloud computing provider able to win Nvidia's favor and continue to receive investment after this year's generative AI explosion?
On the website, which also received Nvidia's investment in paperspace, another private cloud provider, there is a detailed comparison between CoreWeave and Amazon Cloud Technologies, the world's largest cloud service provider, for computing services in the United States.
As you can see, CoreWeave not only offers a more diverse configuration, but it is also much cheaper than AWS!
Taking the NVIDIA A100 80G GPU service as an example, CoreWeave offers clusters of 1-8 cards at a cost of only $2.61 per hour per card. AWS, by contrast, offers an eight-card option and costs nearly twice as much as CoreWeave.
It is undeniable that AWS is mainly aimed at large enterprise customers, and stability may indeed have more advantages in all aspects, but for users with limited budgets who do not have such large computing needs, CoreWeave is really "affordable and more choices."
CoreWeave claims that they can "deliver high-quality computing services 35 times faster and 80 percent cheaper than mainstream computing companies."
CoreWeave started with mining and achieved "class transition" with NVIDIA. CoreWeave has three founders: Michael Intrator, Brian Venturo and Brannin McBee. The trio initially worked in finance as commodity traders, hedge funds and family offices.
When they were still managing funds in New York, the cryptocurrency mining boom had not subsided, initially just to earn extra income, they bought the first GPU, and then bought more and more, and Wall Street desks were full of GPUs.
"In 2016, we bought our first GPU, plugged it in, put it on the pool table in our office in Lower Manhattan overlooking the East River, and mined the first block on the Ethereum network." Michael Intrator, CEO of CoreWeave, recalled in a blog post from 2021.
After consideration, they decided to move GPU hardware into a garage in suburban New Jersey on the east coast and start a business, mainly operating Ethereum cryptocurrency mining. The following year, CoreWeave became one of the largest Ethereum miners in North America-holding more than 50,000 GPUs, accounting for more than 1% of the Ethereum network.
When cryptocurrency prices plummeted in 2018 and 2019, they decided to diversify into other areas that were more stable but also required a lot of GPU computing.
In 2019, they purchased Leonardo Render, a direct-to-consumer rendering platform, and shortly thereafter brought Concierge Render to market. They were committed to building a cloud infrastructure that would enable creators and innovators around the world to access scalable infrastructure at affordable prices.
And from 2019, CoreWeave focused on buying enterprise-class GPU chipsets, building dedicated cloud infrastructure, and aligning the business around NVIDIA chips.
As new businesses get on track, the Ethereum mining business gradually becomes marginalized. The decision to transform proved correct and fortunate, as none of the founders anticipated the heat of the upcoming AI wave, allowing CoreWeave to gradually expand from a small office to a nationwide data center to respond to the expanding AI market demand.
According to one of the founders, CoreWeave has revenue of about $30 million in 2022, is expected to exceed $500 million in 2023, an increase of more than 10 times, and has signed nearly $2 billion in contracts. This year, it announced a $1.6 billion investment in data centers in Texas, with plans to expand to 14 data centers by the end of the year.
In the words of founder Michael Intrator,"I've never worked in a business that's growing so fast myself."
In order to balance the major cloud computing platforms, NVIDIA bet on start-up cloud enterprises and CoreWeave's "instant red", on the one hand, stems from their keen sense of GPU computing cloud demand for continuous research and development and investment.
And more importantly, they were spotted by Nvidia.
As a result, many foreign netizens began to wonder whether NVIDIA had fabricated such a company to "make accounts" for its own performance. Why else would he invest money in a startup from nowhere to buy his own GPU?
Although such conspiracy theories are easily debunked, they do show that Nvidia's preference for CoreWeave has reached the point of "treating it as his own."
After accepting Nvidia's investment, CoreWeave became Nvidia's "direct cloud computing supplier", and even Microsoft had to order cloud computing from them.
Microsoft signed a $1.6 billion contract with CoreWeave in June.
The main reason is that Microsoft itself cannot get enough GPUs from NVIDIA.
A four-year-old start-up company that got GPUs from NVIDIA surpassed giants like Tesla and received almost the same treatment as Amazon, the overlord of the cloud computing industry.
From Nvidia's point of view, the most direct reason why they should take care of such a start-up company is that it can effectively balance potential competitors in the chip industry-Amazon and Google.
They all launched their own AI chips a few years ago to compete with Nvidia.
The AI computing platform is also the most direct "computing power dealer" for enterprise consumers. If Nvidia does not cultivate a cloud computing platform that is completely loyal to itself, if one day "computing power dealers" begin to sell only the computing power generated by their own chips, Nvidia is likely to be "stuck" by cloud computing platforms from "sales channels".
Therefore, supporting its own cloud computing platform and even focusing its computing resources on it is the best way for NVIDIA to deal with Google, Amazon and Microsoft to launch AI chips.
Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.
CTOnews.com, December 14, Motorola announced that it will launch a new limited version of moto x PANTONE Pantone "extraordinary Magenta", which will be launched at 19:30 on December 15, the new Motorola moto X40.
CTOnews.com news on November 15, V Society quietly launched the OLED version of the Steam Deck game handset last week, with some improvement in screen quality and battery life, but V Society announced at that time that the OLED version of Steam Deck
After a period of experience, the Find N3 phone is really a lot of surprises. Let's share with you the real color images trained by Hassel, the super-field-of-view virtual screen that makes it bigger, and the secure VIP mode. Pay attention to bilibili, the house of IT
Thanks to CTOnews.com netizen Colorful M for the clue delivery! CTOnews.com February 2 news, Feifan Automobile today released "Feifan R7 product value system adjustment," said that based on the diversified trend of user intelligent travel scenarios, since