Intel (NASDAQ:INTC) (NEOE:INTC:CA), one of the world’s leading chipmakers, did report poor second-quarter financial results as data centers appear to be underwhelmed with the chipmaker’s current products. Nonetheless, the company still seems to be well-positioned to benefit from increased demand for AI PCs, while its Gaudi 3 AI chips may generate significant, needle-moving revenue for Intel. What’s more, over the longer term, the quality of its data center chips could very well rebound meaningfully, while the company’s foundry efforts may succeed.
In light of all of these points, I recommend selling Intel stock for now. But given its multiple, potential, positive catalysts, I think it’s worth keeping an eye on the chipmaker in order to determine if these possible developments do start to show signs of materializing. If there are indications that the firm is capitalizing on these opportunities, Intel stock could become worth buying down the road.
Poor Q2 Results as the Data Center Unit Underperformed
Intel’s top line fell 0.9% in the second quarter versus the same period a year earlier to $12.83 billion, while its adjusted earnings per share came in at 2 cents, down from 13 cents in Q2 of 2023. Analysts’ average revenue estimate was $150 million higher than the firm’s top line, while their mean EPS outlook was 8 cents higher. The firm expects Q3 sales of $12.5 billion to $13.5 billion, far below the mean outlook at the time of $14.39 billion. It expects an adjusted per-share loss of 3 cents, far below the mean outlook at the time of an adjusted profit of 32 cents per share.
It’s become quite fashionable to portray Intel as a dead-in-the-water company whose technologically backwards products are constantly losing tremendous market share. But the firm’s PC chip business actually performed quite well in Q2, as its revenue from desktop chips increased a healthy 6.6% year-over-year, while its sales from notebook chips climbed a very impressive 15% YOY.
The biggest culprit behind the dismal results, which likely negatively impacted its Q3 guidance as well, was its Data Center and AI business. Despite the rapid expansion of data centers and AI, the unit’s top line actually dropped 3% year-over-year.
Potential Positive Catalysts in the Datacenter Space
Intel is starting to produce less dense data center-oriented chips called Intel 3, and the tech giant should eventually benefit from being part of a “duopoly” with AMD (AMD) in the “lucrative” datacenter space, commentator Kumquat Research noted recently.
Moreover, the company’s Gaudi 3 AI chip, which was first made available to server makers last quarter, could start positively moving the needle for the firm in Q3 and even more so in Q4. According to Intel, Gaudi 3 provides “50% on average better inference and 40% on average better power efficiency2 than Nvidia H100 – at a fraction of the cost.” Given these points, along with the chronic shortages of AI chips, I would not be surprised to see Gaudi 3 rack up sizeable revenue in Q3 and Q4.
Among the major companies that committed to utilize the chips were IBM (IBM), the large German auto parts maker, Bosch, and the huge Indian telecom company Bharti Airtel. These announced customer wins show that there is some demand among large companies for the offering. Moreover, the fact that Nvidia has delayed shipping its new Blackwell chip until January 2025 could also meaningfully boost the demand for Intel’s Gaudi 3 AI chips this year.
Other Potential, Positive Catalysts for Intel
AI PCs could ultimately become a positive catalyst for Intel and Intel stock. On the company’s Q2 earnings call, CEO Pat Gelsinger noted that the Intel Core Ultra chip was “already powering AI capabilities across more than 300 applications and 500 AI models,” while the firm had shipped over 15 million AI PCs between December and early July. Further, the company expects to ship over 40 million such devices by the end of 2024 and over 100 million in total by the end of 2025.
Some are concerned about the declining margins of the company’s AI PC chips. But I believe that history shows that, although the Street worries about margins in the short term, over the long term, firms’ top and bottom lines are more important. After all, the shares of many firms with low margins, such as General Motors (GM) and Exxon Mobil (XOM), have risen a great deal in recent years
And the company’s efforts to make semiconductors for other chipmakers should eventually boost Intel stock. The business, known as Intel Foundry, is currently losing significant amounts of money. But Microsoft (MSFT), chip giant Broadcom (AVGO), and major telecom-equipment maker Ericsson (ERIC) have been announced as Intel Foundry’s customers. That’s an impressive start. Further, in July Gelsinger reported that the unit’s customers have already committed to spending $15 billion on Intel Foundry, and he added that “There’s a lot more in the pipeline.”
What’s more, a huge gorilla in the chip space, NVIDIA (NVDA) reportedly may start using Intel Foundry to make its chips. Foundry is not expected to become profitable until 2030, but investors could become excited about its revenue gains and outlook well before then, particularly if it does land Nvidia as a customer.
Intel’s decision to sell its stake in chip designer Arm Holdings (ARM) for slightly over $147 million at the end of Q1 will give Intel a relatively small amount of extra cash that it can use to reduce the impact of Intel Foundry’s losses and help pay for the capital expenditures needed to expand the initiative.
Valuation and the Bottom Line
Intel is now changing hands at a fairly low forward Enterprise Value/Sales ratio of 2.24 times. Since analysts, on average, expect Intel to generate low EPS of 26 cents this year and $1.24 in 2025, I believe that using its revenue to value the shares is a more logical approach at this point. Enterprise value provides a fuller picture of companies’ true worth because, unlike market capitalization, it takes into account cash and debt.
Taking a look at the forward EV/Sales ratio of other chipmakers, Micron (MU) weighs in at 4.97, while Broadcom’s forward EV/Sales multiple is 16.2 and NXP Semiconductors’ (NXPI) is 5.6.
Also, noteworthy is that Intel’s PC business is performing fairly well. Still, the struggles of the company’s Data Center/AI unit are worrisome. However, the chipmaker does have multiple, positive, potential catalysts. But it’s still too early to know if and when those possible drivers will start emerging and being appreciated by the market. Consequently, I recommend selling the shares but keeping a close eye on the firm’s progress.
Read the full article here
