Nvidia Vs. Intel: The Growth Story Versus The Turnaround Stock
Contributor since 2020 / 10415 Followers
Summary:
- Nvidia Corporation has overtaken Intel as the most profitable data center chip vendor due to the rise of AI. It is delivering strong growth and does not have an excessive valuation.
- Intel is struggling financially (due to a manufacturing business operating at a 66% operating loss) but is on a comeback path technologically, presenting a turnaround investment opportunity.
- Both Nvidia and Intel offer investment opportunities, with Nvidia being a reliable growth stock and Intel presenting a potential turnaround play.
- Its low valuation provides Intel with the strongest potential upside.
Robert Way
Investment Thesis
In our initial coverage of Nvidia Corporation (NASDAQ:NVDA) vs. Intel (INTC) in 2020, we observed that the two companies, which historically operated in distinct markets, were becoming competitors. Several years later, this has become reality, with Nvidia quickly overtaking Intel as the most profitable data center chip vendor due to the rise of AI. Meanwhile, Intel struggles financially, on an expensive turnaround path after a decade of mismanagement. But this is also how the stock is priced anyway, and technologically, Intel is on a comeback war path, so it shouldn’t be written off (yet).
As such, actually both companies present a legitimate investment case, each with a different profile. Since the saying goes that real men have fabs, the Intel fab turnaround remains a valid investment strategy, with the chance for outsized returns if Intel can finally ramp EUV to get them profitable (from currently woefully unprofitable with a -66% operating profit margin). Secondly, Nvidia has demonstrated reliable and consistent QoQ (hyper)growth. While growth is obviously slowing down, which should reduce the possible alpha that might be obtained from investing in the stock, it is therefore the less risky, more reliable investment.
Note: Nvidia releases its earnings next week, on August 28th.
Background
The original comparison in 2020: Nvidia Vs. Intel: The Semi Battle Of The Decade.
Intel’s latest quarterly report discussion: Intel: Financials Stagnate.
Nvidia’s latest quarterly report discussion: Nvidia: Another Strong Quarter Reaffirms Investment Case.
Company profile: Nvidia
Nvidia is from origin (mostly) a GPU company. However, with the advent of the modern era of AI (deep learning) since 2012, the company has begun investing in chips specifically for this market. Most notably, its GPU business started seeing an initial surge around 2017 with the launch of Volta, which contained Tensor Cores to speed up performance by something like 5x. Through various techniques (in combination with Moore’s Law), performance has further increased over the generations (with Ampere in 2020, Hopper in 2022 and Blackwell early next year according to the latest rumors). Nvidia further acquired Mellanox several years ago for its networking portfolio.
Although Nvidia already saw a large increase in revenue through the COVID-19 period, revenue started seeing a surge from the second quarter of last year (2023) as it became clear large language models (“LLMs”) were becoming the next big thing in tech (kickstarted by OpenAI’s ChatGPT in late 2022).
This surge in revenue has been on the order of a (quite linear) $4B increase in (data center) revenue per quarter. Due to leverage, given the almost software-like margins Nvidia has achieved from its data center GPUs (with their very high prices, especially since Hopper), the profit and free cash flow margins have increased strongly as well, so the increase in EPS has far exceeded (about doubled or so) the revenue growth.
Nvidia has been supply limited, and according to its statements this may continue well into next year, so the likelihood of a continued rise in revenue seems quite assured. Of course, with quarterly revenue now exceeding $20B, a roughly $4B increase in revenue per quarter means the revenue growth is slowing down to more normalized (although still high) levels. In addition, the benefit Nvidia saw over the last year from increased leverage (higher FCF and NI margins) is likely tapering out, so earnings growth going forward should be more in line with revenue growth (rather than outpace it).
In terms of valuation and stock price, the stock is up nearly 160% YTD already. As discussed in previous coverage, Nvidia had been a clear buy since the forward P/E ratio had decreased to around 30x (since the stock had remained quite flat in the second half of 2023), which given its growth rate made the stock relatively cheap.
After its rally this year, though, and recently bouncing back from a dip back near all-time highs, the forward P/E has increased to 45x. While this is certainly less cheap, even slowly growing companies like Apple (AAPL) are often valued at or near 30x. As such, paying a 50% premium to get much stronger growth still seems like a solid investment case.
In addition, the forward (aka 2024) P/E now also includes two reported quarters. Due to its strong QoQ earnings performance (although, as noted, also strong stock performance), the run rate P/E based on its Q2 results is 50x. For comparison, another semiconductor company that has regularly traded at such valuations is ASML Holding (ASML), and clearly Nvidia is delivering stronger growth.
In terms of forward growth outlook, in addition to the reported continued shortage, as mentioned, Nvidia has already announced its successor, Blackwell, so Nvidia is likely to (continue to) benefit from both higher volume and higher prices.
The main risk, which as discussed is quite certainly not soon, is that at some point the build out of AI infrastructure may reach its peak. It could perhaps even see a decline if it turns out there had been an industry overinvestment. An argument often pointed out is that while a lot of money is being poured into AI, the actual revenue that is being generated from AI applications is much lower, which seems or may be unsustainable. On the other hand, bulls would regard AI as a generational tech trend that might see a continued increase in investments for many years.
Overall, Nvidia presents a quite standard growth stock investment, with the benefit (compared to many growth stocks) of being very profitable. The valuation is higher than most S&P 500 (SP500) stocks, but so too is its growth profile.
Company profile: Intel
Intel has three main businesses, each mostly based on the traditional CPU. It participates in the mostly stagnant PC market, but also in two/three potential growth markets with its data center and networking/edge segments.
In the aftermath of COVID, there had been a downturn that started in 2022. The PC market has recovered to about pre-COVID levels, while (on the other hand) both data center and networking/edge seem to have stabilized at a lower level. The data center sector is likely experiencing some pressure as customer investments are and seem to remain on the AI (accelerator) buildout side for the foreseeable future. Given the Q2 report, it seems there is no (full) recovery in demand in sight, with revenue (currently) stabilizing at a much lower level than even pre-COVID (although Intel has since already divested about a dozen businesses, some of which like modem and NAND were quite sizeable).
Intel has further been trying to diversify its business over the last decade or more, but many of these efforts have failed and/or as mentioned have been spun off, or are still in progress. The latter ones are GPUs (both for consumers and the data center), autonomous driving with Mobileye (MBLY) and the foundry business that was (re)started by Pat Gelsinger.
Overall, in the near-term, there are indeed little if any catalysts for revenue growth.
What has also impacted Intel (in addition to the downturn and perhaps the current AI investments), especially on the profitability side, is its erosion of technology leadership over the past half a decade (and more), in particular since the TSMC (TSM) N7 node, in late 2018 for Apple (AAPL) and in 2019+ for others like AMD (AMD). This has most visibly led to market share losses in the data center, likely in addition to pricing power erosion, as its chips became less competitive.
The reason for this loss of technology leadership, as an IDM that manufactures its chips instead of outsourcing to a foundry, is that it has seen significant delays in all three of its last major nodes (14nm, 10nm/Intel 7 and 7nm/Intel 4) due to yield issues (defects during manufacturing). This has caused it to slip from being two years ahead to being three years behind TSMC.
As such, after announcing the 7nm (Intel 4) delay in mid-2020, Intel has hired Pat Gelsinger as CEO to lead a turnaround effort to regain process leadership. The conclusion so far, four years after that 7nm delay, is that Intel is on track to the roadmap Pat Gelsinger had announced (“5 nodes in 4 years”), which cultivates with 18A in 2025 to regain process leadership.
The reason for this being so significant is that Moore’s Law is an economical law. Quite evidently, while the number of transistors per area doubles over time, the cost of a fab (or wafer) increases much less, resulting in a lower cost per transistor over time. Hence, not just the performance (and power) competitiveness of Intel’s chips has been under pressure, but also its fundamental economics. And that is in addition to the yield issues of these nodes as well (as lower yield results in less chips per wafer, requiring more wafers to achieve the same volume, resulting in an increase in chip manufacturing cost).
Basically, with inferior process technology with relatively low yield, its chips are significantly pricier to fabricate in the first place, and then, due to their lack of competitiveness, have also to be sold at lower prices as well. The aim of the turnaround is to undo this trend by regaining leadership.
The financial impact of this decade of manufacturing issues has only recently really become clear, as Intel has introduced its “internal foundry model” reporting this year, which separates the economics of Intel Products and Intel Foundry. Intel going forward is operating as two virtually independent companies, with (for segment reporting purposes) Intel Products buying Foundry wafers at the same prices that external foundry customers must pay.
Intel has announced it expects peak losses this year. In Q2, the operating loss of Intel Foundry was $2.8B ($11.2B annualized). Note that some of this loss is attributable to costs associated with the investment in 5 nodes in 4 years. Clearly, having a significant deficit in process technology and relatively low yield makes for a very unprofitable business. Given the $4.3B revenue, this makes for a 66% operating loss.
For example, and comparison, TSMC is reporting in the range of 50-60% gross margin and 40-50% operating margin. Hence, this shows the Intel investment thesis. As it restores its industry leadership, there is no reason why Intel Foundry couldn’t achieve similar operating metrics, in the worst-case a bit lower due to lower pricing to gain market share. Based on the Q2 report, this implies on the order of $15B annualized improvement in profitability. For the company overall, the goal is to achieve 60% gross and 40% operating margin by the end of the decade.
Given the current market cap of $90B after the latest stock price decline, that means Intel might currently be trading at less than 5x or even 4x 2030 P/E. Intel has claimed it can achieve its profitability targets without much if any revenue growth at all (which this 5x estimate incorporates). As mentioned, though, Intel has no less than three potential businesses that could become significant over time and hence drive revenue growth (foundry, GPU and autonomous driving). This is also in addition to possible growth if data center and network/edge recover, and perhaps market share and pricing power gains in the data center as it restores its competitiveness against AMD.
In terms of risks, this is arguably far smaller than the stock price would indicate. As discussed, Intel has not had any further delays for the last four years now, and hence is on track to regain industry technology leadership next year. Note, though, that it generally takes several years to fully ramp a process node and overtake the old ones in terms of volume. As such, even after 18A is introduced next year, it will still take a few years for the financials (of Intel Foundry) to (completely) reflect this.
Instead, the main risk is the size of the overall total addressable market, or TAM. Its market share in the PC has already stabilized, and likely also in the data center going forward. If these markets do not really grow (which the PC quite certainly pretty much won’t), and if it isn’t successful in any of its three emerging markets, then (contrary to what management has claimed) it might nevertheless have difficulty achieving the operating metrics it has put out as (2030) target. It is simply the case that both in Foundry and Products, there are (fixed) costs such as R&D that could be more readily absorbed with higher revenue.
In the worst-case, as the recently announced reorganization shows, Intel might simply be unable to sufficiently fund its fundamental development efforts (even though Intel claimed the layoffs would spare those areas), perhaps leading to falling behind yet again.
Overall, given to where its valuation has sunk, Intel stock presents a clear turnaround investment that does not even require much if any revenue growth, and which given the fundamental technology progress (with 18A on track to start ramping in the first half of 2025) is mostly derisked already. After 18A starts ramping, it will take several more years for this to become fully reflected in the financial results, which represents the minimum time horizon for the investment. Further potential upside could come from revenue growth from GPUs, foundry and robotaxis. If all three of these emerging businesses become successful, this growth could in principle also be quite sizeable.
This combination of strong profitability growth (somewhat offset by the current 80x forward P/E) and at least the potential for decent revenue growth could deliver strong investment returns.
Real men (should) have fabs
As mentioned, pursuing Moore’s Law is costly, requiring quite substantial revenue. There is an old saying that real men have fabs, and one could envision, at least in principle, that a company like Nvidia could certainly benefit from controlling this (process technology, packaging and manufacturing) aspect of the stack as well.
Given its current financial results and the current Intel stock price, paying perhaps on the order of $150B to gain access to what will soon be the industry’s most advanced fabs seems like a compelling possibility. Given Nvidia’s financial status, this could derisk Intel’s ability to continue to sufficiently invest in its businesses (mostly CPU and manufacturing) as well.
Investor Takeaway
Arguably, both stocks present a potentially quite lucrative investment thesis.
Nvidia as a traditional and very profitable growth stock, where the main question is how large the market for AI silicon will become. But given its (high but) not excessive valuation as well as the current demand outlook, the risk seems manageable.
Intel is a turnaround play of a company that aspires to (again) become very profitable (with the fabs currently at a 66% operating loss) and potentially return to revenue growth as well, with several compelling emerging businesses. Given its current valuation and the progress of its technology, the investment seems largely derisked. The returns could already be considerable if the revenue and profitability (and hence stock price) would just return to levels seen several years ago, which for the stock price would be a 3x return to $60.
Overall, Nvidia would be the more reliable investment. Intel arguably has the stronger upside case, as the days of Nvidia hypergrowth are slowly ending. As argued, Intel’s strong technology execution should put it on a quite reliable path towards improved profitability (although there is a larger question mark regarding revenue growth, but given the current valuation less required).
Editor’s Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.
Analyst’s Disclosure: I/we have a beneficial long position in the shares of INTC either through stock ownership, options, or other derivatives. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Seeking Alpha’s Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.