AI data center infrastructure by FOTOGRIN via Shutterstock
AI infrastructure is becoming a massive business as companies race to build the computing capacity needed to train and run AI. Supermicro Computer (SMCI) and Coreweave (CRWV) are two key players in the AI infrastructure boom. SMCI supplies the servers, cooling, networking, and other equipment behind those data centers, while CoreWeave operates GPU-powered cloud infrastructure that customers use to run AI workloads.
In other words, SMCI makes the infrastructure while CoreWeave operates it. The question now is which business model has more room to grow as AI demand explodes.
Valued at a market cap of $24.8 billion, Supermicro Computer builds the physical machinery, including GPU and CPU servers, storage, networking, liquid-cooling systems, and increasingly complete rack-scale data-center solutions, that makes AI data centers work. Basically, the company's business is to package these components together so customers can deploy AI infrastructure faster instead of assembling everything separately.
www.barchart.com
The enormous number of data centers and AI systems that are built now is boosting Supermicro's business. The company ended fiscal 2026 strong with a 78% year-over-year (YoY) increase to $39.1 billion. Adjusted earnings for the fiscal year also increased by 76% to $3.63 per share. In the fourth quarter alone, the company added $60 billion in new orders to its backlog. SMCI is increasingly selling complete data center building-block solutions rather than isolated products, which allows customers to bring AI data centers online much faster.
The company is also diversifying from its traditional customer mix. Notably, enterprise and channel revenue increased 172% YoY to $5.6 billion, now accounting for 50% of total revenue, compared to 28% in the previous quarter. This diversification is important so that SMCI doesn't rely entirely on a handful of enormous data center projects. According to management, a more favorable customer and product mix with lower tariff costs led to a higher adjusted gross margin of 17.6% in Q4, compared to 10% in Q3.
The company expects its worldwide production capacity to surpass 6,000 racks a month, with more than 3,000 of them equipped with direct liquid cooling. SMCI has also announced a 32-acre data center building-block campus in Silicon Valley, taking its U.S. footprint to nearly 4 million square feet. While this scale-up is aggressive, it could also lead to an increase in fiscal 2027 sales to $65 billion to $72 billion, compared to $39 billion in fiscal 2026. The balance sheet also supports the aggressive expansion. SMCI ended the quarter with $7.5 billion in cash and $1.2 billion in net debt.
If AI infrastructure spending continues expanding, Super Micro has a chance to sell more systems, capture more of the infrastructure stack, and expand its enterprise customer base. However, its long-term investment case rests on how much of this opportunity it can turn into consistent profits and cash generation.
Therefore, Wall Street remains cautiously optimistic of SMCI stock, rating it a consensus "Hold." Of the 20 analysts who cover the stock, four rate it a "Strong Buy," two say it is a "Moderate Buy," 11 rate it a "Hold," one says it is a "Moderate Sell," and two say it is a "Strong Sell." The stock is up 27% year-to-date (YTD) and is trading close to its average target price. But the Street-high estimate of $60 suggests a potential 56% jump over the next 12 months.
www.barchart.com
The Case for Coreweave (CRWV)
With a market cap of $47.8 billion, CoreWeave's model is different from SMCI's because it owns and operates AI cloud infrastructure while selling processing power and related services to consumers.
www.barchart.com
In the most recent second quarter, total revenue reached $2.6 billion, up 112% YoY. The revenue backlog even reached $104 billion, an increase of 246% from the year earlier. Furthermore, $25 billion of additional customer commitments were signed early in the third quarter. If customers continue running workloads on Coreweave's platform, the same infrastructure can generate revenue over multiple years. Not only that, Coreweave also generates revenue from businesses beyond GPUs, including storage, CPU, networking, and software, which exceeded $400 million in annualized recurring revenue (ARR) in the second quarter. Furthermore, managed inference ARR grew from $1 million to more than $100 million in just a few months, and the company expects at least $250 million of managed inference ARR by the end of 2026.
CoreWeave added nearly 500 megawatts of active capacity in Q2, taking its total to 1.5 gigawatts. Management remains committed to reaching at least 8 gigawatts by 2030. Contracted power stood at 3.7 gigawatts at quarter-end, and the company also has additional potential capacity from powered land, site options, and other arrangements. Building an AI data center isn't just about buying GPUs. CoreWeave also needs electricity, land, cooling systems, equipment, and money to build and operate those facilities.
This is one of CoreWeave's biggest advantages: it is getting the things ready that are needed to build AI data centers years ahead of demand. While the potential of Coreweave's business strategy is huge, so is the cost. Its business model requires enormous upfront investment. It spent $9.4 billion on capital expenditures in the quarter and expects full-year 2026 capital expenditures of $35 billion to $39 billion. It reported a net loss of $626 million in the quarter. At the end of the quarter, it had $6.9 billion in cash, cash equivalents, restricted cash, and marketable securities and raised roughly $18 billion through debt, convertibles, and equity during the quarter.
Wall Street rates CRWV stock as a consensus "Moderate Buy." Out of 35 analysts covering the stock, 22 have a "Strong Buy" rating, one suggests a "Moderate Buy," 11 rate it a "Hold," and one says it is a "Strong Sell." The average target price of $140.30 indicates 67% potential upside from current levels. Meanwhile, the high target price of $250 implies a potential rally of 198% over the next 12 months.
www.barchart.com
Which AI Infrastructure Model Has the Bigger Opportunity?
Finally, while SMCI may have the stronger value today, Coreweave has the more compelling long-term growth model. CoreWeave is trying to capture the ongoing demand for computing after the infrastructure is deployed. If AI inference becomes a huge recurring workload over the next decade, CoreWeave has the potential to keep monetizing its infrastructure while adding services around it, assuming it can properly manage its debt and massive capital spending.
On the date of publication, Sushree Mohanty did not have (either directly or indirectly) positions in any of the securities mentioned in this article. All information and data in this article is solely for informational purposes. This article was originally published on Barchart.com
