AI chip startup Cerebras files for IPO, in market dominated by Nvidia
By Bill Peters
During a funding round in 2021, Cerebras said it was worth $4 billion
Cerebras Systems Inc., a maker of chips and other tech infrastructure for artificial intelligence, filed for an initial public offering on Monday, as more startups look to make a name for themselves in a space occupied by tech juggernauts like Nvidia Corp.
The number of shares to be offered, or their price, was not yet clear. The Sunnyvale, Calif.-based company, in a filing, said it had applied to list its shares on the Nasdaq Global Market under the ticker "CBRS."
Citigroup Global Markets and Barclays Capital are acting as joint book-running managers for the offering. During a funding round in 2021, Cerebras said it was worth $4 billion.
Founded in 2016, Cerebras is trying to build bigger chips to handle the demands of AI, where questions have grown about the costs of rolling out the technology and the impact of its energy demands on the power grid.
The filing also arrives as speculation grows over a possible bubble for the technology. However, many analysts remain bullish on the long-term potential for the technology to upend the labor market and generate loads of profit.
The company, in its IPO filing, said it had solved a decades-long problem in the computer industry's history - developing chips as big as full silicon wafers. Its third-generation Cerebras Wafer-Scale Engine, it said was "57 times larger" than the "leading" graphics processing unit, with far more memory and bandwidth.
"The sheer size of the wafer-scale chip allows us to keep more work on-silicon and minimize the time-consuming, power-hungry movement of data," the company said in its IPO filing. "This enables Cerebras customers to solve problems in less time and using less power."
"Our AI compute platform combines processors, systems, software and AI expert services to deliver massive acceleration on even the largest, most capable AI models," it continued. "It substantially reduces training times and inference latencies, while reducing programming complexity."
Cerebras, in its prospectus, said it makes processors for AI training, and other systems to funnel data into those processors. Its software connects those systems into AI supercomputers, which it also makes, the prospectus added.
Customers, it said, use those computers to train AI models. The company also develops AI cloud services, which it uses to deliver the technology to customers. Last month, Cerebras introduced a new AI cloud service intended to undercut Nvidia (NVDA) on price.
The company logged sales of $78.7 million last year, well up from the $24.6 million in sales it made in 2022. During the six months ending on June 30, Cerebras had $136.4 million in revenue.
Cerebras lost $127.2 million last year, narrower than the $177.7 million it lost in 2022. The company lost $66.6 million during the first half of this year.
The company warned that it got a "significant majority" - more than 80% - of its sales from one customer, G42, an AI company in the United Arab Emirates. It also said it had identified "material weaknesses" in its financial reporting protocols.
-Bill Peters
This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.
(END) Dow Jones Newswires
09-30-24 1750ET
Copyright (c) 2024 Dow Jones & Company, Inc.-
5 Stocks to Buy Instead of Overpriced US Equities
-
Q4 Stock Market Outlook: Where We See Opportunities for Investors
-
Markets Brief: Non-Farm Payrolls in the Spotlight Again
-
6 Top-Performing Large-Growth Funds
-
What’s the Difference Between the CPI and PCE Indexes?
-
Micron Earnings: Great Guidance but Stock Now Looks Fairly Valued
-
August PCE Report Forecasts Show More Good News on Inflation
-
AI Stocks May Be Down, but Don’t Count Them Out
-
The 10 Best Companies to Invest in Now
-
New 4-Star Stocks
-
Morningstar’s Guide to Investing in Stocks
-
Our Top Pick for Investing in US Renewable Energy
-
How to Measure a Stock’s Uncertainty
-
How to Determine Whether a Stock Is Cheap, Expensive, or Fairly Valued
-
Why a Company’s Management and Capital Allocation Matter
-
How to Determine What a Stock Is Worth