Blaize is preparing to bring a different kind of AI chip story to public markets. The company, founded by former Intel engineers and focused on edge applications, announced on Monday that it is set to go public on the Nasdaq in a SPAC deal on Tuesday.
The move comes as Nvidia’s rise has renewed investor attention around AI chip startups. But Blaize is not trying to tell the same data center story. Its pitch is that AI chips will increasingly sit inside physical products, not only inside vast centralized computing facilities.
A public-market test for an edge AI chip company
Launched in 2011, Blaize has raised $335 million from investors including Samsung and Mercedes-Benz. The company is headquartered in El Dorado Hills, California, and says its manufacturing is done in the U.S.
Blaize focuses on AI chips for edge applications. In plain terms, that means chips designed to operate inside devices and equipment closer to where data is created or used. The source article describes examples including security cameras, drones, and industrial robots.
That distinction matters because much of the AI chip conversation has centered on data centers. Nvidia’s chips are described in the source as being mostly used in vast data centers. Blaize is instead betting on chips integrated into smart products that operate in the physical world.
“AI-powered edge computing is the future due to its low power consumption, low latency, cost-effectiveness, and data privacy advantages,” CEO Dinakar Munagala, who previously worked almost 12 years for Intel, said in a statement to TechCrunch.
The numbers show both ambition and risk
Blaize is still a small player in a very large AI chip industry. Its most recently available financials show the company losing $87.5 million on only $3.8 million in revenue in 2023, according to its prospectus.
Those figures make the public-market debut a high-stakes moment. Investors are not being asked to buy into a mature, profitable business. They are being asked to consider whether a capital-intensive chip company can grow into the market it is targeting.
Munagala framed the economics of chipmaking as a long investment cycle that can require heavy spending before scale arrives.
“As you can imagine, [as a] chip company you do a massive amount of investment and when the hockey stick comes, it climbs,” Munagala told TechCrunch.
The company is also pointing to future demand. Blaize is touting $400 million in deals in the pipeline. Its investor deck includes a signed purchase order of up to $104 million with an unnamed EMEA “defense entity,” likely in the Middle East, for a system that can identify unknown or friendly troops, spot small boats, and detect drones. Munagala declined to say exactly which country.
Why the SPAC valuation matters
Munagala told TechCrunch he expects Blaize to be worth $1.2 billion after its SPAC merger. That valuation places Blaize below some other closely watched AI chip names discussed in the source article.
For comparison, TechCrunch previously reported that Cerebras, another AI chipmaker, filed for an IPO last fall and was seeking to double its $4 billion valuation. Cerebras has not yet gone public. The source article notes that some investors had qualms over its over-reliance on a single Middle Eastern customer, according to investors who told CNBC.
The comparison is not exact because the companies focus on different parts of the market. Cerebras focuses on data center chips, while Blaize is emphasizing edge AI chips for physical products. Still, the contrast highlights a broader question for investors: which parts of the AI chip market can support major public companies beyond the data center boom?
The bigger bet: AI moves into products
Blaize’s public-market story depends on the idea that AI hardware will become more embedded in everyday and industrial systems. That does not mean the data center becomes unimportant. It means Blaize is arguing that real-world applications need chips closer to devices, machines, cameras, drones, and robots.
The logic is straightforward. If AI is used in equipment that needs quick responses, lower power use, lower costs, or greater data privacy, then processing at the edge can be attractive. Those are the advantages Munagala emphasized in his statement to TechCrunch.
The company’s examples also show how broad that edge AI pitch can become. Security cameras, drones, industrial robots, defense systems, boat detection, drone detection, and troop identification all point to AI being used outside conventional computing environments.
Munagala argued that the market’s attention has been too concentrated on data centers.
“All of the AI hype is happening in the data center. Interestingly, they’ve totally neglected and forgotten about real physical world use cases that are very real, that are touching people’s lives and are happening now and making money,” Munagala told TechCrunch. “We’re focused on the practical use of AI in the physical world.”
That is the core of the Blaize thesis. The company is not presenting itself as another data center chip challenger. It is asking public investors to believe that AI chip demand will spread into physical products and that edge computing will become a meaningful part of the AI infrastructure market.
The risk is equally clear. Blaize has raised substantial capital, has a large claimed pipeline, and is entering Nasdaq with a specific view of where AI hardware is going. But its latest available revenue remains small, and its losses are large. The SPAC deal will test whether investors are ready to back that edge AI future before it is fully reflected in the company’s financials.