Investment Conclusion
For 2Q24, Intel Corporation (NASDAQ:INTC) came in short of expectations on revenues and earnings. Compared to 2Q23, revenues were flattish, and the company recorded a loss of over a billion dollars. Some believe that Intel is in a free fall with no turning back, and speculation is rife that its foundry business is likely to be divested or expansion plans associated with the segment are likely to be downsized. We don’t know, but the firm appears to have a plan, it has recognized that it is failing, it has a long-term roadmap, and is implementing strategies for a comeback. The tide has not yet turned, but it appears close.
Intel Corporation has upgraded its chip manufacturing process. By YE24, the size of its processors will have shrunk to 1.8nm, 2nm, 3nm, 4nm, and 7nm, from their 10nm specification in 2021. With 1.8nm chips, INTC will reclaim process leadership from Taiwan Semiconductor Manufacturing Company (TSMC), whose 2nm processors will launch next year. In addition, INTC’s processors produced on its 1.8nm node will integrate its proprietary cutting-edge chip technologies, RibbonFET, PowerVia, and Foveros Direct.
Further, to continue to dominate the PC segment of the semiconductor market, Lunar Lake AI CPU will provide 45 TOPS (trillions of operations per second) of neural net processing, and its upcoming Panther Lake AI CPU 50 TOPS of neural net processing vs. Qualcomm’s (QCOM) Snapdragon X Elite’s 48 TOPS and Advanced Micro Device’s (AMD) Ryzen Strix Point’s 45 TOPS of neural net processing.
In regard to server CPU’s , INTC Xeon 6 Sierra Forest E-core built with 128 cores, its forthcoming Granite Rapids P-core CPU with 144 cores, and its Clearwater Forest E-core, coming next year, will have 288 cores, compared to primary competitor AMD’s EPYC server CPU with 128 cores. Regarding AI accelerators, INTC’s Gaudi 3, to be introduced this quarter, boasts a superior price/performance profile to Nvidia’s (NVDA) H-100 GPU and is competitive with AMD’s Instinct MI100X GPU. Moreover, INTC’s upcoming Falcon Shore discrete GPU is positioned as a key competitor to NVDA’s Blackwell range of GPUs. Overall, INTC will have caught up with the competition regarding product prowess by the middle of 2025.
Furthermore, the company has restructured its top leadership, bringing in turnaround artist Patrick Gelsinger as CEO and added industry veterans to head core business segments, as well as instituted the Semiconductor Co-Investment Program ((SCIP) to improve financial flexibility. In addition, INTC has recreated its foundry business, such that external customers (fabless companies) and internal customers (Intel Products) will be treated equally.
Net-net, given the magnitude of efforts underway to turnaround the firm, we don’t view INTC’s 2Q24 financial performance as dramatic. Turnarounds are gradual, there are bound to be puts and takes on the journey to a comeback. In addition, we don’t perceive the disappointing second quarter outcomes and employee downsizing/reduction in operating spending and capital expenditures as related events. In our opinion, the cuts in manpower, operating expenses, and capital expenditure is a logical next step in the execution plan to turnaround the company.
Overall, based on our analysis, INTC’s revenues and earnings growth is likely to rebound over the 2025-2026 time frame. Further, the firm’s long-term financial risk reward profile appears favorable. Consequently, at current levels. INTC’s shares represent an opportunity to generate asymmetrical returns on investment. We’re initiating on INTC with a Buy Rating and $33/share Price Target, based on our 10-year Discounted Cash Flow model, which includes a terminal value.
Investment Thesis
INTC was founded by Robert Noyce and Gordon Moore (of Moore’s Law fame) in 1968 in Mountain View, California. The company’s headquarters are located in Santa Clara, California.INTC’s wafer fabrication plants reside in the U.S., Ireland, and Israel. In addition, the firm has product assembly and testing facilities in the U.S., China, Costa Rica, Malaysia, and Vietnam.
During 2023, Dell Technologies (DELL) accounted for 19% of INTC’s total revenues, followed by Lenovo at 10%, and HP Inc. (HPQ) at 9%. In April 2024, the U.S. Department of Defense, which announced that RAMP-C customers can begin to manufacture and commercialize prototypes on INTC’s 18-A process node, became an additional large INTC customer.
INTC’s primary product is the CPU and chipset, on SoC (system on chip) or multi-chip package, based on INTC architecture. Under its Consumer Computing Group (CCG), the company provides Intel Core processors for notebooks and desktops. Within its Data Center and AI (DCAI) segment, INTC markets Intel Xeon processors. In addition, under its Networking and Edge Computing (NEX) business, INTC provides Intel Core, Intel Xeon, and Intel Atom, branded processors. In addition, Intel Foundry provides semiconductor contract manufacturing for internal and external clients. Further, INTC has two subsidiaries: MobilEye and Altera.
During FY24, the firm’s CCG segment represented 35% of total revenues, the DCAI business 29%, the NEX category 11%, Intel Foundry 2%, and MobilEye 4%. There are two predominant issues driving investor interest in INTC. The first is, will turnaround efforts succeed. The second is related to the company’s long-term financial outlook.We provide answers to both elements, below.
Strong Foundation Being Built Suggests Better Days Ahead
Given the overhaul of INTC’s business structure initiated in 2021, the company’s business is now comprised of two distinct segments: 1) Products and 2) Foundry. Regarding products, Intel is a designer and marketeer of integrated circuits aka processors aka chips, utilized in data centers, enterprise servers, and personal computers, as well as for networking and edge computing. Regarding the foundry business, the enterprise is structured to manufacture chips for Intel Core and Intel Xeon design teams, as well as for fabless chip companies such as NVDA and AMD.
INTC is integrating AI across its entire technology stack, from semiconductor manufacturing, to its chips, to networking, to edge computing, and to its devices.
Delving further into INTC’s product segment, the company primarily competes with NVDA as a manufacturer and marketeer of chips for data centers and AI, as well as for networking and edge computing. Although, based on projections, NVDA has captured 70% to 90% of the AI chip market share, we believe a market shift is likely over the second half of the decade, in favor of predominantly INTC, but additional industry players such as AMD, Cerebras Systems, Tenstorrent Group, and D-Matrix are positioned to benefit as well. Our conviction that INTC’s share of the AI chip market is likely to escalate over upcoming years is based on several factors:
1)Not all AI workloads require excessive processing power and speed, which are the hallmarks of NVDA’s processors. The first step of machine learning, data engineering requires excessive memory capacity, training requires computing power, and inference requires latency. In that regard, large language models ((LLMs)), such as GPT-4 with 1.7 trillion parameters require tens of thousands of GPUs and three to four months to be trained, whereas an inference workload (basically a response to a user prompt) can be processed using two baseboards of eight GPUs each.
In that context, regarding the DCAI segment (although, it’s beside the point, given that INTC is focused on promoting its chips) for: 1) GenAI workloads associated with training medium to smaller language models that is the purview of businesses and sovereign data centers (local language models); 2) for inference workloads related to LLMs; 3) AI PCs; and 4) edge computing; which is distinct from NVDA’s business priority which appears to favor at scale training and inference related to LLMs that belong for the most part to the cloud hyperscalers), it is not as if INTC’s AI chips are significantly inferior in performance to NVDA’s processors.
In that regard, regarding INTC’s DCAI chips, based on specifications provided on NVDA’s website, an 8,192 chip cluster of INTC’s Gaudi 3 AI accelerator trains LLMs at 40% higher speed than a similar set of NVDA’s H100 GPUs. Moreover, a 64 chip baseboard of INTC’s Gaudi 2 system, provides 15% higher throughput than a comparative system of H100 GPUs, on Meta Platform’s (META) Llama2 LLM. In the context of inference workloads, Gaudi 3 appears to be 3x faster in delivering responses than H100 on LLMs such as Llama2 and Mistral. In addition, Gaudi 3 comes close to NVDA’s B100 at 16-bit floating-point precision, while B100 has an advantage regarding the 8-bit floating-point precision, which Gaudi 3 lacks. In addition, Gaudi 3 will compete with AMD’s Instinct MI100X GPU in the data center.
Regarding server CPUs, INTC will pull ahead of AMD’s 128-core EPYC server CPU, with its Xeon 6 Granite Rapid P-core 144 core CPU expected to be launched this quarter, and later with the company’s Xeon 6 Clearwater Forest E-core 288 core CPU slated to be introduced in 2025. INTC’s Xeon 6 E-core chips provide a balance of power requirements and processing speeds, while the Xeon 6 P-core processors are designed for high-performance computing. It is notable that AMD’s 192 core server CPU Turin is anticipated to be launched later in 2024.
In 2025, INTC will introduce its data center focused Falcon Shore discrete GPU. The chip will most likely be produced on INTC’s 14A node, and be integrated with INTC’s proprietary chip technologies, RibbonFET, PowerVia, and Foveros Direct. In addition, Falcon Shore will be a multi-tile processor (permitting CPUs, GPUs, and AI accelerators from varied vendors), with x86 cores for general computing and xe cores for highly parallel AI and high-performance computing.
Regarding INTC’s chips for devices, the firm’s CCG segment provides processors for the PC market, including for laptops, desktops, and tablets. The company is focused on rapidly developing a presence in the AI PC market. In that context, INTC has introduced several generations of AI CPUs, including Raptor Lake produced on Intel 7, Meteor Lake manufactured on Intel 4, and Lunar Lake produced on Intel 3. The latter is a 3nm chip with 100 TOPs of AI performance, including 45 TOPS of neural net processing capacity. 1/3rd of PCs with INTC AI chips will be fitted with Lunar Lake processors. It appears that ~20 original equipment manufacturers with ~80 designs have signed on to secure Lunar Lake processors.
In addition, Microsoft (MSFT) has selected Lunar Lake chips to power its new Co-Pilot Plus AI PCs. Lunar Lake will jockey for market share with AMD’s Ryzen Strix Point CPU with 48 TOPs of neural net processing power and Samsung’s X Elite AI chip with 45 TOPs of neural net processing capacity. In 2025, INTC will debut the Panther Lake AI CPU, built on the Intel 18A node, integrated with RibbonFET, PowerVia, and Foveros Direct, and designed to achieve 120 TOPs of AI performance and 50 TOPs of neural net unit performance.
Net-net, by the middle of 2025, INTC’s products will have caught up with and even improved on competitor offerings, regarding chip size and technology.
2) The power consumption associated with NVDA’s chips is significant, even considering the processing power and speed they provide. Although INTC’s chips produced on legacy nodes, including Intel 7 are disappointing regarding power specifications, INTC’s chips produced on its 20A and 18A node are expected to demonstrate an improved power/performance profile.
3) Thought leaders posit that prices of AI chips have to decline considerably for AI to trickle down to the masses. The premium prices associated with NVDA’s processors is likely to be a gating factor, looking ahead. The demand for AI chips is likely to be limited only by the availability of compute. If the prices of GPUs, CPUs, and AI accelerators, remain largely at current levels, AI will be limited to critical workloads. CPU manufacturers might mitigate a fraction of the shortfall by increasing matrix math on their equipment. Nevertheless, to process all workloads that can supported by AI, clearly, prices of AI processors have to decline.
In that regard, it is noteworthy that a NVDA H100 baseboard of eight GPUs costs $200K, while INTC’s Gaudi 3 baseboard of eight Gaudi 3 AI accelerators is priced at $125K.
4) NVDA has a monolithic approach towards providing AI infrastructure. The company is moving towards highly integrated offerings, comprised of NVDA branded GPUs, CPUs, and AI accelerators, as well as its proprietary: CUDA software for chip programming and InfiniBand for networking.
In converse, INTC has adopted a disaggregated approach. The firm’s strategy is to offer CPUs, GPUs, and AI accelerators, without customers being forced to purchase the entire rack. The approach permits companies and chip designers to mix and match chips to create rack compositions based on their requirements. Considering that semiconductor design is transitioning from monolithic designs to disaggregated designs, moving from a system on a chip to a system of chips approach, INTC’s policy, which allows chips and tiles from external foundries on its systems, positions the firm to benefit vs. NVDA. In addition, the disaggregated platform strategy, also favors INTC’s packaging business, which permits mix and match of processors from any foundry (TSMC and Samsung don’t allow inclusion of external tiles on their systems), supporting unique baseboard designs on potentially large substrates.
Regarding chip optimization strategies, INTC is relying on open sourcing all code and PyTorch. NVDA, on the other hand, has CUDA, chip programming software, proprietary to its processors.
Further, INTC has chosen the Ethernet as its networking medium, as compared to InfiniBand, which is NVDA’s pick. While the Ethernet provides higher bandwidth and is cheaper than InfiniBand, congestion in data flow is a recognized problem. Conversely, Infiband comes at a higher price point and with a lower bandwidth, which is a drawback as higher bandwidths increase AI workload efficiency. Nevertheless, the Ultra Ethernet Consortium comprised of INTC, AMD, META, and MSFT, and others has banded together to invent technologies which would solve the data congestion issue. Looking ahead, given that the Ethernet is being projected as the industry standard for networking, INTC selection of the Ethernet as its networking medium appears prudent.
In addition to the factors described above that are likely to derail NVDA’s success, is the element of supply chain shortage. NVDA’s chips are backordered for a year. The premium pricing associated with NVDA’s processors and the opportunity cost of having to delay AI workloads for an elongated period, is likely to drive absorption of all AI chip supplies provided by alternate sources, including INTC.
Regarding the foundry business, INTC, the quintessential pioneer of chip manufacturing, lost its leadership position within the foundry industry, as it could not deliver enough capacity on the 10nm node (now Intel 7). In addition, TSMC and Samsung moved on to providing their 7nm and 5nm chips for smartphones, while INTC continued to focus on PCs.
However, times have changed. In 2021, the newly elected CEO, Patrick Gelsinger promised to deliver five node updates in four years. He has delivered. INTC’s 7nm, 4nm, and 3nm nodes have been rolled out and the 2nm and 1.8nm nodes are being prepared for launch by YE24. In addition, the firm’s process node roadmap includes 1.4nm chips on the 14A node by 2026, and 1nm chips on the 10A node, shortly thereafter.
In that regard, it is noteworthy that INTC now characterizes chip sizes in angstroms, with the 20A node producing 2nm chips, the 18A node manufacturing 1.8nm processors, etc. Node sizes matter, as the smaller the processing node, the more the transistors that can be fitted per wafer, reflecting in greater number of transistors per chip, and consequently improved performance. INTC chips, beginning with the 1.8 node and beyond, will be integrated with RibbonFET, PowerVia, and Foveros Direct. RibbonFET is based on gate all around transistors, which tackles the problem of current leakage, to improve overall power consumption and performance. PowerVia is a backside power delivery system that resolves the conflict for space between component and power cables, when they both sit on top of the transistor layer, reflecting in signal interference and power integrity issues. Foveros Direct is a technology, which permits vertical stacking of transistors, greatly enhancing the number of transistors per chip, which improves performance.
Comparatively, TSMC and Samsung have rolled out 3nm chips. Both companies anticipate launching 2nm processors in 2025. In addition, TSMC will introduce a 1.6nm chip in 2026, and Samsung its 1.4nm chip in 2027. In addition, TSMC processors already come fitted with gated all around transistors, while Samsung’s do not. Samsung has announced that its 1.4nm chips will be integrated with back-side power, while TSMC has yet to share plans on the technology. Considering that INTC appears on path to deliver 2nm and 1.8nm chips integrated with back-side power, gated all around transistors, and vertical stacking of transistors, by YE24, the company will then have reclaimed its position as the leader of chip manufacturing and technology.
It is important to note that there are only a handful of cutting-edge foundries manufacturing chip components across the world, including INTC’s Fabs in Arizona, Oregon, Israel, and Ireland; TSMC’s plants in Taiwan, Arizona, and Japan; and Samsung’s facilities in South Korea and Texas. Considering that ~80% of the semiconductor supply chain is based in East Asia, and given the geopolitical conditions associated with China and Taiwan, the U.S. administration appears interested in reshoring/friendshoring chip production to the Western hemisphere. In that context, INTC is being courted as a key partner, assured government funding of ~$8.5 billion, to expand the company’s semiconductor production capacity.
INTC is keeping its part of the bargain, building new facilities in Ireland and Ohio, at a cost of $18 billion and $20 billion. In addition, the firm is investing $20 billion to expand manufacturing capacity at its Arizona foundry. Projects to expand production capacity are also underway at INTC Fabs in Oregon and New Mexico. Considering, INTC’s efforts in rendering its processors more competitive as well as its policies to significantly enhance chip manufacturing capacity, it appears that the company is going all in to deliver on its promise to wrestle back control of processor manufacturing from TSMC, and achieve the mantle of the world’s second-largest chip foundry by 2030.
Furthermore, to support turnaround efforts, along with the ascension of Patrick Gelsinger as CEO, key leadership positions have been switched around, with INTC bringing in industry veterans Kevin O. Buckley, Nick McKeown, and Greg Lavender, to head Intel Foundry, the Network and Edge Group, and the Software and Advanced Technology Group. In addition, Patrick Gelsinger promoted INTC veterans Sandra Rivera and Raja Koduri, to lead the Data Center and AI segment, and the Accelerated Computing Systems and Graphics Group.
Crux of the story is that INTC is at the right place at the right time. With compute requirements doubling every 10 months, a commensurate growth in the number of chips shipped is expected to unfold as well. Regarding DCAI and NEX processors, they will be picked by customers based on varied criteria. Some will select for performance, others because of power consumption. Still others will choose to keep scalability top of mind. Nevertheless, the primary drivers of chip choice will be availability, scalability, and price. Given NVDA’s chip’s premium pricing and supply chain shortfall, all supply of INTC’s processors, given their favorable price/performance profile and upcoming ample availability, is likely to be absorbed by the market.
Regarding AI PCs, pent-up demand dynamics are likely to manifest over the upcoming six to eight months. Two factors drive our conviction that rapid customer uptake of AI PCs is around the corner. The first is that the service updates for Windows 10 are scheduled to lapse in October 2025. Considering that enterprises typically initiate refresh cycles for PCs, a year ahead, corporate demand for AI PCs is likely to escalate over the rest of 2024 and 1H25. The second is that independent customer-driven substantial PC demand transpired during Covid-19, four years ago. Therefore, the group appears poised to replace their computing devices with AI PCs. Overall, INTC’s AI processors Lunar Lake, and the upcoming Arrow Lake and Panther Lake will possibly be very popular among PC original equipment manufacturers, in the near future.
Similarly, INTC’s foundry business, given the process leadership roadmap and the integration with cutting-edge chip technology, is likely to experience tailwinds. Considering that these INTC processors will roll out ahead of chips manifesting similar technology from key competitors, INTC’s chips appear well positioned for technological dominance. Therefore, we see little reason for fabless chip companies not to shift some fraction of their semiconductor manufacturing requirements to INTC’s foundry business.
Cumulatively, INTC is undergoing a bottom-up transformation. Will the efforts support the company in selling more chips and in manufacturing more semiconductors for external parties? Considering the strong foundation being laid, and favorable market dynamics, we believe so.
Long-Term Financial Outlook To Improve As Turnaround Efforts Gather Momentum
As we’ve stated above, we’re not overly concerned about the disappointing second quarter results. They were more an outcome of increased spending and export controls related to Chinese sales, impacting revenues. Net-net, given the improvement in financial performance that rolled out in 4Q23 and 1Q24, we view 2Q24 as a throwaway quarter. In addition, hockey stick turnarounds in company fortunes are rare, with most comeback stories meeting resistance in the initial stages of the effort. Therefore, we project a few additional disappointing quarters, before INTC’s financial performance improves significantly.
To be specific, FY23 financial results, with year-over-year revenues down 14% to $54.2 billion, comprised of sales declines of 8%, 20%, and 31% associated with the company’s CCG, DCAI, and NEX segments, and gross margins, operating margins, and profit margins falling by 15%, 24%, and 22%, on a two-year basis, were terrible. However, if one were to glance more closely at recent financial outcomes, green shoots appear to be sprouting. Specifically, in 4Q23, compared to 4Q22, revenues advanced by 10%, gross margins by 6.5%, operating margins by 25%, net income was $2.7 billion compared to a loss of $700 million, EPS came in at $0.63 vs. a LPS of $0.16. Over 1Q24 compared to 1Q23, revenues were up 9%, gross margins improved by 7%, operating margins by 4%, net income by 86%, and LPS was $0.09 vs. $0.66.
Considering business dynamics, including the anticipated launches of:18A and 20A process nodes, Panther Lake, Clearwater Forest, and Falcon Shore, all integrated with RibbonFET, PowerVia, and Foveros Direct, INTC holding an edge over competitors, should see its fortunes look up over the next couple of years and beyond.
In that regard, it is noteworthy that industry publications are projecting strong growth in segments INTC dabbles in. AI PCs (potential business opportunity for Lunar Lake, Arrow Lake, and Panther Lake) are expected to account for 59% of all PC sales by 2026, compared to a projected 22% in 2024. In addition, as per Gartner, by 2026, 100% of enterprise purchases of PCs will be AI PCs. INTC had shipped 15 million AI PCs by the end of the second quarter, and anticipates shipping 40 million and 60 million AI PCs in 2024 and 2025. Combining the competitiveness of the firm’s CPUs built for AI PCs, and that ~80% of PCs are shipped with INTC processors, we believe that AI PCs fitted with Intel AI processors will dominate the market.
In addition, thought leaders are projecting that sales of data center AI processors will expand by a CAGR of 14.5% between 2024 and 2032. Further, semiconductors utilized for networking are expected to witness average sales growth of 13.9%, between 2022 and 2031. Moreover, the edge computing market is estimated to advance at a CAGR of 34% between 2024 and 2032.
Regarding INTC’s foundry business, it is important to note that the semiconductor industry is projected to expand 17.4% in 2024 to $700 billion and to $990 billion by 2030. Trends suggest that key factors driving semiconductor revenue acceleration will be GenAI and processing requirements associated with the development of artificial general intelligence (AGI).
INTC’s foundry will compete on affordability, undercutting competitors’ prices, without compromising on chip performance. Patrick Gelsinger has stated that INTC knows the cost of TSMC’s wafers, the average selling prices of its chips, as well as its process node targets, and plans to utilize the information to its benefit. INTC is targeting $15 billion in foundry revenues associated with external parties by 2030. This figure would support the firm’s objective of usurping Samsung’s second position in the global foundry industry. Considering, INTC’s upcoming node leadership and the integration of cutting-edge chip technologies, we believe INTC will handily achieve its target.
In that context, it is noteworthy that MSFT has announced plans to manufacture its AI chips on INTC’s 18A node. A further six clients have signed on to producing their integrated circuits on the 18A node. Furthermore, consistent with the agreement inked with Arm Holdings (ARM), the 18A node is being optimized to produce ARM chips, opening the door for INTC to produce silicon for smartphones and servers based on ARM architecture. Regarding internal customers, Intel Core and Intel Xeon team’s chips, including Panther Lake and Clearwater Forest are slated to be built on the 18A node. Ultimately, based on management commentary, INTC is likely to bring back to its foundry, the contract manufacturing business it has heretofore provided TSMC etc.
In addition, INTC has several opportunities to improve margins. The average selling prices associated with wafers manufactured on its next-generation nodes is 3x that related to the firm’s legacy nodes, while the costs of inputs remain virtually the same. Further, given that average selling prices associated with chips increases with the number of cores, and that INTC’s recent processors have more cores, the company has a margin expansion opportunity at hand. Additionally, gross margins associated with INTC chips provided for AI PCs is higher than provided for non-AI PCs. Furthermore, it is notable that 2/3 of processors that INTC markets for PCs are produced on the highly uneconomical 7A node. As the company transitions chip manufacturing to its smaller size nodes, the costs of goods is bound to decrease. Furthermore, as INTC brings home chip manufacturing from external foundries, the firm’s product costs are likely to decline.
Further, a massive corporate restructuring is underway at INTC. The company has announced plans to lay-off 15% of its workforce, reflecting in cost savings of $20 billion, over time. Further, significant reductions in annual operating expenses and capital expenditures are underway. Moreover, to maintain balance sheet flexibility, while it invests to turnaround the company, INTC has brought in external capital to partly fund its efforts to expand its chip manufacturing capacity. We believe that management is likely to further streamline the business, including cutting capacity at legacy manufacturing nodes. Most chip production will possibly shift to next-generation process nodes, in our assessment.
For 2030, INTC has provided a rough financial roadmap. The firm is planning for $100 billion in revenues, and a 60% and 40% split between gross margins and operating margins. In addition, INTC anticipates that it’s CCG and DCAI revenues will expand at a CAGR of between 3% to 5%, and that it will add between $8 billion to $10 billion in accelerated compute revenues. Finally, it expects that MobilEye and Altera’s revenues will expand at a CAGR of 10%, each.
We believe, based on more competitive products, spending cuts, and operating leverage, as well as industry dynamics, we believe INTC’s business will begin to rebound over the next couple of years. Consequently, the 2030 estimates provided by the company appear conservative.
Incorporating the above described qualitative narrative into our 10-year Discounted Cash Flow model, we arrive at a Price Target of $33/share for INTC. Our valuation assumes a normalized 10-year revenue growth rate of 8%. In addition, we derive our net income for 10-years using a net profit margin of 13%. Based on our analysis of INTC’s historic financial reports, we model normalized 10-year operating cash flows as 30% of revenues/year, and straight lined 10-year capital expenditure as 22% of revenues/year. Furthermore, we deploy a perpetual growth rate of 3% and weighted average cost of capital of 8% to reach our terminal value and present value of free cash flow figures. We utilize the current diluted outstanding share count of 4,267 million to arrive at our Price Target for INTC.
Risk
Undoubtedly, the long-term outlook for the semiconductor industry is lock-stepped with outcomes for AI. Sure, there are camps within the technology ecosystem that continue to posit that potential AI implications are overblown. However, we don’t buy the thesis that AI is simply a trend likely to run its course and disappear. Two factors are driving our conviction on the issue.
First, it is noteworthy that 2030 semiconductor industry revenue projections of ~$1 trillion attribute $850 billion to integrated circuits, up from $410 billion generated in 2023. The solid growth estimate is despite trends that indicate a continued decline in sales of memory chips, and CPUs and GPUs associated with smartphones. Therefore, it appears that industry powers that determine revenue projections anticipate that AI chip requirements will meet and exceed the slack created by the shortfall in demand for chips linked to legacy PCs and smartphones.
Second, clearly machine learning is driving transformations across industries, from helping optimize shipping and delivery routes within the logistics category, to personalize shopping experiences across the retail segment, and automating manufacturing tasks in factories. Broadly, it appears that LLMs represent a tectonic shift in human evolution, much like the advent of electric power. In our judgment, use cases for AI are likely to expand as practitioners get comfortable with the technology.Overall, given the anticipated significant demand for chips to power the AI era, the risk that INTC’s turnaround efforts will flounder is minimal.
Bottom Line
Jeff Bezos famously indicated to his employees that because U.S. public companies typically have a lifespan of 30 years, Amazon (AMZN) would one day fail and file for Chapter 11. INTC is 56 years old. In addition, although the company has successfully navigated through several turnarounds, the struggles have weakened its fundamentals.
Delving into INTC’s history, clearly quality of leadership has strongly impacted its fortunes. Solid CEOs, such as Andy Grove, Craig Barrett, and Paul Otellini, drove prosperity at INTC. In our opinion, the current CEO, Patrick Gelsinger, is a chip from the block of INTC’s famous former CEOs. He joined INTC when he was 18, spent 30 years at the firm, before being ousted from his role as Chief Technology Officer, in 2009. He then went on to become CEO of DELL subsidiary VMware, where he turned around the company, expanding revenues 3x to $12 billion, and diversified the business from its core hypervisor, to networking, cloud security, containers, and 5G. In 2019, based on a survey performed by Glassdoor, Patrick Gelsinger was recognized as America’s Best CEO.
Considering Patrick Gelsinger’s prior elongated tenure at INTC, and that he lead the firm’s key business segments during the period, he understands INTC well. In addition, Patrick Gelsinger appears dedicated to return the company to its leadership of the semiconductor industry. Until now, he has delivered on the promises he made to shareholders. Further, given his frenetic efforts to execute the turnaround strategy, most often leading from the front, coupled with his success at VMware, it appears that Patrick Gelsinger’s endeavor to deliver a solid rebound in INTC’s financial performance is highly likely to bear fruit.
Consequently, in our opinion, at current levels, INTC’s shares represent a low-risk high reward opportunity. Buy, Buy, Buy.
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.
Read the full article here
