AMD's Strategic Moves in the AI Chip Market: Challenging NVIDIA's Dominance

AMD Takes An Effective Bargain Route Against Nvidia

The additive growth outlook will also underpin sustained margin expansion as new products, especially within the more profitable data center segment, continue to scale. Catching up to rival Nvidia's (NVDA) AI prowess has long been a milestone in the works for Advanced Micro Devices, Inc. (NASDAQ:AMD) after it beat Intel (INTC) at its own game in server processors. To better equip itself as an adequate competitor of Nvidia's, AMD has been consistently taking a page from its rival's all-encompassing hardware-software full-stack business model through acquisitions like Xilinx and Pensando recently. And to further bolster its AI competitive advantage against the industry leader, AMD has also introduced its next-generation Instinct MI300 accelerators this year, which has proven success in winning market share. This was evident in AMD's robust 2Q24 results and reinforcing 2H24 outlook, complementing above-seasonal tailwinds expected in client segment sales driven by an emerging cyclical recovery and AI PC opportunities. Better-than-expected market share gains across its core data center and client segment sales have not only been key to offsetting ongoing pressure in AMD's declining gaming and slow-recovering embedded segments. It also has improved operating leverage and, inadvertently, margins for the company. Ensuing cash flow growth has also been supportive of AMD's increasing priority over AI investments to better its competitive positioning against Nvidia. The company has deployed more than $1 billion into AI company acquisitions and organic R&D activities over the last 12 months in an aim to bolster its reach into opportunities within the burgeoning AI ecosystem. They include the acquisition of Silo AI, an enterprise-focused large language model ("LLM") developer, which closed earlier this month. More recently, AMD has entered into an agreement to acquire server systems designer and manufacturer, ZT Systems, for $4.9 billion. Looking ahead, the continued ramp of new products - particularly those supporting emerging AI developments - alongside the integration of Silo AI and ZT Systems should enable further share gains and unlock greater growth synergies for AMD. The following analysis will provide an update on AMD's fundamental outlook based on its latest 2Q24 earnings update, and also gauge the implications from the impending ZT Systems acquisition. The stock remains in an opportunistic set-up at current levels, in our opinion, with AMD's prospective share of the AI market across hardware and software applications still underappreciated. AMD's 2Q24 outperformance provides structural support of its growing AI prowess. Its data center revenue mix is nearing 50%, doubling from a year ago after the segment's sales soared 115% y/y driven by strong uptake of both its server processors and AI accelerators. The increasing mix of higher-margin data center sales has also been a key margin accretive factor for AMD. While consolidated revenue grew a modest 9% y/y during the second quarter, GAAP net income surged more than eight-fold over the same period and more than doubled sequentially. AMD's latest Instinct MI300 accelerators were in the spotlight, as management disclosed a steep ramp for the company's newest product. Specifically, AMD's data center GPU revenue contributions continue to set records, with quarterly MI300 accelerator sales exceeding $1 billion for the first time during 2Q24. As a result, management has raised their estimate for full year 2024 MI300 revenue contributions to $4.5 billion, up from $4 billion guided in April and $3.5 billion guided in 2023. This implies further acceleration in MI300 sales through 2H24, with about $2.8 billion to go before hitting the newly raised target for completely additive growth within the data center segment. Recall that AMD's Instinct MI300 is a competitive alternative to Nvidia's best-selling H100 and H200 data center GPUs. AMD's latest accelerator builds on the CDNA 3 architecture, and enables up to 5x energy efficiency on training AI workloads. The MI300 consists of 153 billion transistors, which is almost double of the H100's, enabling superior performance efficiencies to peers, especially in processing complex AI workloads. The MI300 accelerator also boasts 192GB of HBM3 memory, offering more than double the memory of Nvidia's H100 and highly competitive with its rival's H200 GPUs. The competitive TCO metrics make AMD's MI300 a favorable choice for inferencing when stacked against industry leader Nvidia, which reinforces the product's prospective demand environment and market share gains. The MI325X shipping later this year will be further accretive to AMD's data center GPU sales growth prospects. The upgraded accelerator will consist of 288GB of next-generation HBM3E memory, punting it as a direct competitor to Nvidia's H200 and upcoming Blackwell B100 GPUs. The diminishing performance and cost efficiency gap between AMD's Instinct accelerators and its competition is likely to reinforce confidence in AI-related data center sales through the back half of 2024. This will be essential to driving incremental margin accretion for the segment towards corporate levels, giving further adjustments to operating leverage and scale benefits already observed through robust MI300 volumes shipped in 1H24. In addition to accelerators, AMD's EPYC server processor sales have also delivered a better-than-expected showing. This was critical to assuaging previous investors' fears over the potential obsolescence of general-purpose server CPUs in the ongoing AI-driven transition to accelerated computing. The sustained double-digit percentage increase in EPYC CPU sales during the second quarter continues to highlight AMD's strengths in addressing structural enterprise optimization trends. As discussed in our previous coverage on the stock, AMD's EPYC server processors have been key enablers of adding incremental "general purpose and AI compute capacity without the need to materially change the physical footprint and power needs of existing infrastructure." While the fourth generation Genoa EPYC server processors offer the same level of compute with "45% fewer servers compared to the competition," which enables up to 40% and 50% in initial opex and capex reduction for customers, the upcoming Zen 5 Turin EPYC CPU is expected to further optimize the product segment's "leadership performance and efficiency." The Zen 5 Turin EPYC processors, which is currently in public preview and will start volume shipments later this year, is compatible with existing fourth generation EPYC platforms. Coupled with up to 192 cores and 384 threads, the Turin EPYC processor extends the product segment's competitive TCO, with currently previewing customers highlighting the product's "significant performance advantages in multiple compute-intensive workloads". This is further corroborated by performance stats recently released by AMD on the 128-core configurated Turin EPYC processors in processing AI workloads, which outperformed Intel's latest Emerald Rapids Xeon processors by wide margins. The comparison also highlights the competitive technology advantage of the 192-core configurated Turin in the AI-first era. Taken together, we believe AMD remains well-positioned for further acceleration in its data center segment sales, which entails incremental margin expansion ahead. This will benefit cash flows underpinning the stock's valuation prospects. As discussed in a previous coverage, AMD continues to utilize the consolidation strategy effectively in bolstering its competitive advantage and technical capabilities. The effectiveness is evident in the combination of additive growth and cost synergies realized through ongoing M&A activities at AMD recently to better position itself against intensifying competition. In addition to the acquisition of Xilinx and Pensando, which has deepened AMD's hardware-software full stack capabilities, the company has also been investing heavily into furthering its AI ecosystem. This includes the recent acquisition of Silo AI, which provides AMD with direct access to expertise in enterprise-focused LLM developments. The acquisition of Silo AI is expected to bolster AMD's provision of AI development and deployment solutions against competing offerings such as Nvidia's "NeMo" platform and "NIM" inference microservices. The company has also recently entered into an agreement to acquire server designer and manufacturer ZT Systems for $4.9 billion, with a contingent consideration of $400 million subject to the completion of certain milestones. Similar to server makers like Super Micro Computer (SMCI) and Dell (DELL), which have benefitted from resilient AI infrastructure investments over the past year, ZT Systems also provides rack-scale solutions for rapid data center deployments. AMD's upcoming acquisition of ZT Systems, which it intends to settle through a combination of cash and stock, will directly extend the chipmaker's reach into the increasingly lucrative server industry, while also bolstering AI-related market share gains. Specifically, AMD intends to sell ZT Systems' manufacturing capabilities post-close, while keeping the subsidiary's server design expertise. This implies the impending launch of AMD AI supercomputer systems, which will potentially compete directly against Nvidia's HGX and DGX solutions for market share across hyperscaler and enterprise customers. The integration of ZT Systems' expertise in the server market is expected to further AMD's end-to-end AI capabilities, effectively extending its reach into the burgeoning AI TAM. Admittedly, the acquisition of ZT Systems also has its fair share of downside risks. The transaction is expected to be margin-dilutive initially due to the combination of integration costs, as well as lower profitability from ZT Systems' manufacturing-weighted business. Specifically, market research currently estimates LTM revenue of $10 billion at ZT Systems, with the majority generated from its manufacturing business. Server companies such as Dell, SMCI and Hewlett Packard Enterprise (HPE) currently trade at under 1x NTM sales, pressured by their capital-intensive manufacturing arms. This is in line with the ~0.5x sales that AMD has offered for the acquisition of ZT Systems based on the proposed consideration of $4.9 billion and estimated LTM revenue of $10 billion at the computer maker, generated primarily from lower-margin manufacturing sales. However, the transaction is expected to be fundamentally accretive for AMD over the longer-term. Specifically, AMD management intends for the acquisition of ZT Systems to be "accretive on a non-GAAP basis by the end of 2025," which implies immediate disposal of the computer maker's manufacturing arm post-close in 1H25. Management also disclosed that AMD will keep about 1,000 engineers upon the completion of ZT Systems' integration strategy, highlighting intentions to bolster its expertise in systems design through the acquisition. The transaction is expected to be net additive in the long-run for AMD's AI strategy by adding systems design to its resume of expansive AI capabilities that span networking (e.g., Ultra Accelerator Link and AMD Infinity Fabric), software (e.g., ROCm), hardware (e.g., Instinct and EPYC) and services. In addition to bettering AMD's stance against key rival Nvidia, the acquisition of ZT Systems' server design expertise will also be a complementary effort in driving uptake of AMD's data center processors, given its supportive end-to-end solutions ecosystem. This will accordingly reinforce the sustained long-term growth trajectory of AMD's higher-margin data center sales and, inadvertently, bolster cash flows underpinning its valuation outlook. Adjusting our previous forecast for AMD's actual 2Q24 results and forward outlook, we expect the company to grow revenue by 14% y/y to $25.8 billion in 2024. The updated estimates also reflect our continued confidence in an upcoming upgrade supercycle for AMD across the data center, client and embedded segments as discussed in our previous coverage. The continued ramp of new products - which include the Instinct MI300 and upcoming MI325X accelerators, Ryzen Pro 8000 and Ryzen 9000 Series CPUs, and second generation Versal Adaptive SoCs - is also expected to drive further accretion to margins. This is expected to extend the pace of sequential margin expansion observed in recent quarters, which has been primarily driven by scaling up more profitable data center sales and improving operating leverage across all operating segments. We maintain confidence that AMD continues to represent a favorable risk-reward set-up at current levels, given additive growth and margin accretive opportunities ahead. Our base case price target for the stock remains at $200 apiece. The price target is derived from the discounted cash flow approach, and considers cash flow projections taken with the base case fundamental analysis discussed in the earlier section. A 9.2% WACC in line with AMD's risk profile and capital structure is applied. The analysis also considers a perpetual growth rate of 6.0% applied to projected 2028E EBITDA in determining AMD's terminal value. The premium valuation assumption is consistent with the application of a 3.5% perpetual growth rate on projected 2033E EBITDA when AMD's growth profile is expected to normalize in line with the higher-range pace of estimated economic expansion. The assumption applied is also reflective of AMD's mission-critical role in enabling next-generation growth technologies critical to economic expansion across its core operating regions. AMD's upside potential is further corroborated by its modest valuation on a relative basis to peers with a similar growth profile. Under the multiple-based approach, AMD is currently trading at about 8x 2025E sales, which lacks the premium currently exhibited by comparable peers in the Philadelphia Semiconductor Index (SOX) despite continued market share gains in margin accretive segments. We believe the stock remains well-positioned for an impending upward valuation re-rate driven by several prospective catalysts that include a stronger-than-expected ramp-up of higher-margin data center revenues, and server market synergies with the integration of ZT Systems' design unit in 1H25. As discussed in the earlier section, AMD expects about $2.8 billion in data center GPU sales - a new revenue stream for the segment - in 2H24, representing growth of 65% from 1H24. This represents continued sequential acceleration in data center GPU shipments, supportive of incremental economies of scale benefits for the high-margin data center segment. Coupled with a potential pricing tailwind stemming from the persistently tight supply chain through 2025, and strong uptake for its Instinct MI Series accelerators bolstered by an expanding ecosystem of supportive solutions, AMD exhibits prospects for an upside surprise. Full integration of ZT Systems' design unit into AMD's Data Center Solutions Business Group in 1H25 and further scale of related solutions in 2H25 is likely to drive further margin accretion as well. This remains underappreciated at the stock's currently traded levels, in our opinion. Admittedly, AMD's share of AI opportunities and ensuing revenues continues to trail Nvidia's by wide margins. But AMD's ongoing efforts in bolstering its end-to-end ecosystem capabilities continues to showcase the company's commitment to narrowing its competitive gap with the industry leader. AMD also continues to exhibit an effective consolidation strategy, which represents a bargain route that optimizes capex deployments on a relative basis to its prospective growth and margin expansion trajectory. This accordingly assuages growing investors' concerns over potential risks of overspending on AI coming out of the latest earnings season. Taken together, we believe AMD's current market valuation continues to represent a conservative fundamental outlook that underappreciates the additive growth and margin accretive factors ahead of ongoing execution of the underlying business' AI strategy.

AMD: Buying The Bear Blitz (NASDAQ:AMD)

This idea was discussed in more depth with members of my private investing community, BAD BEAT Investing. Learn More " Advanced Micro Devices, Inc. (NASDAQ:AMD) stock has given up a ton of its value since peaking in March of this year at $227. On the last massive sell down, we guided our members to open buys at $135 and $125. Shares are up nicely in that short time frame, but we see more upside ahead longer-term. Make no mistake, the chip sector, and tech more broadly, has been pressured for many weeks, but has woken back up in the last 8 sessions. After a massive run higher, we are actually still in the middle of a correction, though one would argue and likely should that we are seeing some mean reversion. Now we have little doubt that the chip sector is anxiously awaiting NVIDIA Corporation's (NVDA) report next week, and the sector will take considerable cues from that report, the guide, and most importantly the commentary on the outlook for AI-chip related spend. So, it is likely we see another wave lower, and if so, you should scale in. While shares are down enough from the peak to be in a bear market situation, the valuation has reset, at least compared to where it was nearly 70 points higher. And this comes as AMD continues to perform well. We think you can buy now, but we embrace scaling in as the market allows, establishing a position. With that said, the earnings performance in Q2 supports a buy on this large pullback, in our opinion. The future is bright. The AMD Q2 earnings were overall strong. The headline results were above consensus. We think that operations continue to benefit from the AI revolution, which is only a boon to operations, even if it will take several quarters to really ramp up more from here, as competition is stiff. We are watching every chip report for signs that AI is not all it was cracked up to be (a risk), but thus far, capex from customers on AI needs continues to rise. That said, in Q2, revenue was $5.84 billion and was up 9% from last year, and surpassed estimates by $120 million. Now, we recognize that not all segments are firing. Leadership has emerged from data center and client revenue, while gaming continues to suffer, perhaps in part due to where we are in the gaming and PC demand cycle. Embedded segment revenue took a hit too. So, AI is driving a big chunk of sales as AMD enjoyed record Data Center segment revenue of $2.8 billion. Folks, this was up 115% year-over-year primarily driven by the steep ramp of GPU shipments, and strong growth in 4th Gen CPU sales. Revenue was also up 21% sequentially, which is one of the more bullish points we must highlight. We also like the bounce back in the Client segment revenue. This revenue was $1.5 billion, up 49% year-over-year and also up 9% sequentially, primarily driven by sales of its Ryzen processors. But as we alluded to, it was not all good news. Their Gaming segment revenue was just $648 million, down 59% year-over-year and 30% sequentially. Further, the Embedded segment revenue was $861 million. This was down 41% year-over-year. Management asserted that this was due to customers that are still in the process of trying to "normalize their inventory levels." However, one bright spot for Embedded is that revenue increased 2% sequentially. We closely watch gross margins, and these have been under some pressure depending on the quarter, but we are seeing margin expansion right now and that is a big bullish sign. AMD's GAAP gross margins were 49%, up 300 basis points year-over-year, and also rising from 47% in the sequential Q1. We saw a brief dip in margins in some 2023 quarters, but now we are clearly seeing progress here with this print. Adjusted gross margins also expanded, widening to 53%. This was up from the sequential quarter's 52%, and up from 50% a year ago. While Q2 2022 was 54% for margins, we are encouraged by the present widening of margins. As the revenues and adjusted margins were up from last year, AMD saw an unsurprising increase in gross profit. Operating expenses remain elevated. We note that the 15% increase in adjusted operating expenses outpaced the percentage revenue gains, but was not enough to offset the strong gross margins. In Q2, AMD's operating income, as adjusted, rose to $1.26 billion, or an increase of 18% from $1.07 billion a year ago. Now we want to be balanced. It is noteworthy that two years ago, adjusted operating income was $1.98 billion. So while this quarter saw positive trends and metrics are moving in the right direction, it is a far cry from 2 years ago. We would prefer more focus on controlling operating expenses in an environment where sales in key segments are mixed. Any decline in AI demand could be catastrophic near-term. That said, putting it all together, AMD's net income rose from $948 million from a year ago to $1.13 billion. Earnings per share increased to $0.69 from $0.58 last year and beat by $0.01. EPS was also up 11% from the sequential Q1. AMD Chair and CEO Dr. Lisa Su stated in the press release: "We delivered strong revenue and earnings growth in the second quarter driven by record Data Center segment revenue. Our AI business continued accelerating and we are well positioned to deliver strong revenue growth in the second half of the year led by demand for Instinct, EPYC, and Ryzen processors. The rapid advances in generative AI are driving demand for more computing in every market, creating significant growth opportunities as we deliver leadership AI solutions across our business." It was encouraging to see that the AI business is accelerating, setting AMD up for a strong H2, but we are cognizant, as we stated above, that all eyes will be on NVDA and its report. Regardless, AMD had a strong quarter and a positive outlook. The company also has a great balance sheet. The AMD balance sheet is quite healthy. At the end of Q2, cash was $4.1 billion. Cash from operations was $593 million compared to $379 a year ago, though these numbers still lag Q2 2022 which was over $1 billion on this front. Free cash flow was $439 million, up from $254 million a year ago and up from the sequential Q1's $379 million. Still, this also lags compared to $906 million two years ago. This is not a bearish take, we are just pointing out that shares have ballooned on hopes for more growth, while 2022 was a year of considerable growth and better segment performance outside the AI-involved segments. Total debt stood at $1.72 billion, so net cash is a strong $2.4 billion. AMD stock has corrected heavily on uncertainty for AI and a rotation that took place in late spring and early summer. We are also potentially about to hit a seasonal correction. As we look ahead, we still see ongoing weakness for PCs and client revenue. We continue to think that the data center leads the way, but we like the rebound in Client revenue. We think the Gaming and Embedded segments are in a lower trough of their cycle and would expect to see some rebound in these segments in 2025. That said, here today we view shares as a buy here, but if we get another selloff, do some more buying. For Q3, we believe Advanced Micro Devices, Inc. will start to see performance ramp up after the Q2 print. That is our thesis here. AMD guided for revenue to be approximately $6.4-$7.0 billion, while gross margin will be around 53.5% a respectable margin. Our view for the year on revenues is $25.5-$26.2 billion. We see revenue strength ramping up later this year and into 2025. Assuming at least a 53% margin, and even roughly comparable capex and opex, we see EPS of $3.43-$3.50 for 2024 as likely. Assuming the midpoint, AMD stock is not cheap, but is much cheaper than a few months ago. We also think it is worthwhile to point out that the market really liked the most recent acquisition news. The company is acquiring AI infrastructure provider ZT Systems in a cash and stock deal worth about $4.9 billion. After announcing this, the stock actually added some $10 billion in market cap and has tacked on recently. The acquisition will help AMD expand its data center and AI systems capabilities, and is another step forward for AMD in its AI strategy. Advanced Micro Devices, Inc. stock is technically in bear territory. We have the pending catalyst of the NVDA report, and we are also coming into a seasonally weak period for markets. We also have macro forces in a US election year that could add more uncertainty. Thus, we see AMD shares as a buy here, but if you get a selloff, do some more buying.

AMD Appears Primed To Return To Its Recent Highs (Technical Analysis) (NASDAQ:AMD)

AMD's recent surge back to mid $150s could lead to a potential breakout to $180 per share, supported by strong market fundamentals. Last month, Advanced Micro Devices, Inc. (NASDAQ:AMD) was on the verge of breaking out of an extended downtrend. Then, it and the vast majority of semiconductor and AI related equities sustained an extended period of decline that culminated in early August. As it turns out, AMD also touched the bottom of the trading range when the market touched its recent low. Since then, like many other peers, it has substantially appreciated. I believe it is entirely likely that AMD will again test the top of this recent trading range, and potentially break out to the upside this time. Earlier this week, AMD reported that it will be acquiring ZT Systems, an AI infrastructure provider, in a cash and stock deal worth about $4.9 billion. The transaction also includes a contingent payment that may be worth up to $400 million, based upon reaching certain post-closing milestones. The acquisition will help AMD expand its data center and AI systems capabilities, and will provide AMD with additional existing AI revenue from its operations. ZT systems also owns a domestic data center infrastructure manufacturing business that AMD intends to divest upon finding a strategic partner. The acquisition is expected to close in the first half of 2025. AMD CEO Lisa Su noted that "Combining our high-performance Instinct AI accelerator, EPYC CPU, and networking product portfolios with ZT Systems' industry-leading data center systems expertise will enable AMD to deliver end-to-end data center AI infrastructure at scale with our ecosystem of OEM and ODM partners." At the end of July, AMD reported second quarter 2024 earnings that beat estimates due to multiple factors, including record Data Center segment revenue. AMD's second quarter revenue came in at $5.84 billion, beating consensus estimates by about $120M, and non-GAAP earnings of $0.69 per share were slightly better than the $0.68 estimate. As part of its earnings, AMD CEO Lisa Su commented that "Our AI business continued accelerating, and we are well positioned to deliver strong revenue growth in the second half of the year, led by demand for Instinct, EPYC and Ryzen processors." Data Center revenue growth was about 115 percent higher than the same quarter one year ago. Second quarter Data Center revenue was a record $2.8 billion. Moreover, AMD's Client segment revenue was $1.5 billion, which was a 50 percent year-over-year increase. Not every segment was as strong. For example, AMD's Gaming segment was down 59 percent year-over-year to $648 million. This decline was primarily due to a decrease in semi-custom sales. Furthermore, AMD's Embedded segment realized $861 million in revenue, which was down 41 percent. AMD provided Q3 2024 guidance expecting total revenue of $6.7 billion, which was a guide higher than consensus analyst estimates of about $6.6 billion. There are several reasons to expect AMD to remain competitive in coming years, including the company's history of remaining so with in Intel (INTC) in the market for personal computing chips. It now appears that AMD took a sizable lead in that race since at least mid-2020, when Intel was compelled to announce a delay to its 7nm product due to a defect in the process. At that time, AMD had already brought a 7nm product to market, and has since brought to market a 4nm product. Despite previously generally lagging Intel in performance, it was able to catch up and eventually surpass that competitor. Now, AMD is likely the second horse in the A.I. chip race, where Nvidia (NVDA) maintains a sizable lead in both technology and market share. It is possible they will one day catch up to NVDA, or at least take a sizable market share by offering reasonably competitive products at a lower price point. AMD chips are also reasonably competitive for cryptocurrency mining, where its most recent version of the Ryzen 9 is often among the better performing choices at its price point for solving the related algorithms. The competitiveness of AMD's CPU chips towards such specific tasks is unusual, as it is typically handled by GPUs, such as graphics cards, and application-specific integrated circuits (or ASICs) that are essentially specialists designed to perform only one function. AMD's recent surge since the market bottomed in early August has taken shares back to the mid-point of the recent trading range, in the $150s. This price point is also roughly coinciding with AMD's 50 and 200 day moving averages. This pricing is likely to provide some level of resistance, but if AMD can break through here again, it should also act as a level of support. If it can break through the $150s, shares appear likely to again attempt to reach the top of the declining trading range it has been stuck in since around the start of spring of this year. The top of the range is roughly $180 per share, and AMD was there in the first half of July. At that point, it took AMD less than two weeks to make it from the $150s to above $180. The same could happen in the next two weeks, or at least in the two weeks following AMD's ability to break above its 50 and 200 day moving averages. Again, shares are basically at those averages, so any continued appreciation could trigger this move higher. One potential near term catalyst could come in the form of analyst revisions premised upon the acquisition of ZT Systems, as well as AMD's own recent performance in the second quarter. This acquisition brings with it existing revenue streams, which means AMD should almost certainly realize an increase to its revenue following the combining of ZT's business into AMD's. This must result in increases to AMD's data center and AI revenue after the completion of the acquisition, and this inorganic growth shall eventually be added to estimates for the second half of 2025 and beyond. The most apparent risks facing AMD bouncing off this resistance point in the mid $150s and falling back towards the bottom of its trading range, as well as a possible broad market selloff that takes most equities down. If AMD were to fall below the mid $140s, I would lose conviction in AMD making a near term test of the top of its trading range. Another risk is that analysts react poorly to the ZT Systems acquisition. This appears unlikely, as the market's immediate response was unquestionably positive. Nonetheless, a change in analyst sentiment would likely act like an anchor to share pricing, and make it more difficult for AMD to move considerably higher in advance of reporting Q3 earnings. AMD appears to be one of few equities that are likely to be a long-term beneficiary of the current and continuing tailwinds from A.I. and cryptocurrency. Further, AMD is likely to benefit from a potential PC refresh cycle. Despite these probable catalysts, AMD has underperformed since spiking to overbought this spring, where it briefly reached prices above $200 per share. Last month, AMD almost broke out of its multi-month downtrend, before breaking down and touching the bottom of its trading range when the market bottomed in early August. Since then, AMD has performed exceedingly well and is now back in the middle of its trading rage, at a price point that previously acted as strong support. Any move higher is likely to test the top of this range, at about $180 per share. A potential breakout to higher prices is entirely possible if AMD can report strong earnings and raise guidance when it reports Q3 2024 earnings this autumn.

Nvidia And The AI Arms Race (NASDAQ:NVDA)

The spring 2024 earnings season was one of the most interesting I can remember. During the various earnings calls and CEO interviews, a group of well-heeled tech companies all admitted to a fact we could already plainly see. They are engaged in an AI arms race, potentially over-investing in GPUs from Nvidia Corporation (NASDAQ:NVDA), which reports its earnings on August 28th. So there's this side of the equation, Nvidia's revenue: Those are the companies engaged in this figurative arms race. Like literal arms races, it is the result of what's known as a "Prisoner's Dilemma" in game theory. As we will look at below, Nvidia customers have decided that the risk of under-investing in GPU cloud capacity is greater than the risk of over-investing. Looking at the table from Company 1's perspective, this is how they are ranking their preferences. In this simplified game, regardless of what Company 2 decides to do, Company 1 thinks they are better off buying more GPUs. For Company 1, the right cells are always a better choice than the corresponding left cells. For company 2, the bottom cells are always preferred to the corresponding top cells. The point is that like in literal arms races, both sides are acting in what they perceive to be their self-interest, but wind up with a suboptimal outcome, nonetheless, an arms race that mostly benefits the arms dealer. In the case of the Cold War, it was nuclear arsenals so large that they eventually had to be negotiated away. Now imagine that there was effectively only one vendor of nuclear missiles for both the US and USSR. That is what we see now. Nvidia's customers are engaged in a GPU arms race, and Nvidia is effectively the only vendor. At Long View Capital, we've talked a lot about the supply and demand sides of AI We have already seen plenty of capital going into the supply side -- both hardware and models/services. We have also seen plenty of demand for cloud services from those new models/services running in the cloud. A lot of that cloud spend is subsidized by the cloud companies themselves via investments in AI startups. But what we have yet to see is high demand for AI services that translates into enough revenue to pay for all these GPUs. The unanswered questions I keep asking on that last point: What is the value proposition? From where is the revenue going to come? When will it come? David Cahn, a partner at VC Sequoia Capital, has called this the "$600 Billion Question." His very rough math is that for every dollar spent on Nvidia GPUs, there needs to by $4 in AI services revenue to back it up, with rough estimates of a 50% margin for the cloud, and a 50% margin for the software. That means with something like $150 billion in Nvidia GPU sales by Q4, $600 billion in AI revenue needs to flow into the ecosystem from customers. Does that seem likely to you? For reference, AI leader, OpenAI, was recently at a revenue run rate of $3.4 billion a year. The questions are starting to pop up during earnings calls, like this one on the Google (GOOGL) Q2 call. Ross Sandler (Barclays) So it looks like, from the outside at least, that the hyperscaler industry going from kind of an under-built situation this time last year to better meeting the demand with capacity right now to potentially being overbuilt next year if these CapEx growth rates keep up. So do you think that's a fair characterization? And how are we thinking about the return on invested capital with this AI CapEx cycle? Sundar Pichai (CEO) Look, obviously, we are at an early stage of what I view as a very transformative area. And in technology, when you are going through these transitions, aggressively investing upfront in a defining category, particularly in an area, which in a leveraged way, cuts across all our core areas, our products, including Search, YouTube and Other Services as well as fuels growth in Cloud and supports the innovative long-term bets and Other Bets. It is definitely something that for us makes sense to lean in. The one way I think about it is when you go through a curve like this, the risk of underinvesting is dramatically greater than the risk of overinvesting for us here. Even in scenarios where if it turns out we are overinvesting, these are infrastructure which are widely useful for us, they have long useful lives, and we can apply it across and we can work through that. But I think not investing to be at the front here, I think, definitely has much more significant downsides. Having said that, we obsess around every dollar we put in. Our teams work super hard, I'm proud of the efficiency work, be it optimization of hardware, software, model deployment across our fleet. All of that is something we spend a lot of time on, and that's how we think about it. This was an excellent question, and Pichai gave only a partial answer. He talked about: Nothing about from where the revenue is going to come, and when it will come. "How are we thinking about the return on invested capital with this AI CapEx cycle?" goes unanswered. That's because the answer is that the return on investment is a long way off. The GPUs purchased in 2023-2024 may be fully depreciated before that happens. In a Bloomberg interview, Facebook (META) CEO Mark Zuckerberg echoed that first point: Emily Chang (Bloomberg) But when does it start paying off? Like, is it a bubble? And if not, when do you start seeing the money? Mark Zuckerberg I think bubbles are interesting, because a lot of the bubbles ended up being things that were very valuable over time, and it's just more of a question of timing, like you're asking. Even the dot-com bubble, it's like there's all this fiber laid, and it ended up being super valuable, but it just wasn't as valuable as quickly as people thought. So, um, is that gonna happen here? I don't know. I mean, it's hard to predict what's gonna happen in the next few years. I think AI is gonna be very fundamental. I think that there's a meaningful chance that a lot of the companies are overbuilding now, and that you look back and you're like, "Oh, we maybe all spent some number of billions of dollars more than we had to." But on the flip side, I actually think all the companies that are investing are making a rational decision, because the downside of being behind is that you're out of position for, like, the most important technology for the next ten to 15 years. Like Pichai, Zuckerberg did not answer the important question Emily Chang asked: "But when does it start paying off?" What did we get? Keith Weiss (Morgan Stanley) Can you give us a little bit more help in understanding the timing between the CapEx investments and the yield on those investments? Satya Nadella (CEO) I would say we primarily start right now from the demand side. What I mean by that is what's the shape of the product portfolio, what we learned even from the cloud transition, which, as you know, Keith, was similar in the sense it was both a knowledge-intensive and a capital-intensive transition. We needed to have the product portfolio where there was the right mix of, I'll call it, infrastructure meters as well as SaaS applications. So that's the first thing that we are looking at. And how is that value landing with customers and what's the growth rate. So when I think about what's happening with M365 Copilot as perhaps the best Office 365 or M365 suite we have had, the fact that we're getting recurring customers, so our customers coming back buying more seats. So GitHub Copilot now being bigger than even GitHub when we bought it. What's happening in the contact center with Dynamics. So I would say - and obviously, the Azure AI growth, that's the first place we look at. That then drives bulk of the CapEx spend, basically, that's the demand signal because you got to remember, even in the capital spend, there is land and there is data center build, but 60-plus percent is the kit, that only will be bought for inferencing and everything else if there is demand signal, right? So that's, I think, the key way to think about capital cycle even. The asset, as Amy [Hood, CFO] said, is a long-term asset, which is land and the data center, which, by the way, we don't even construct things fully, we can even have things which are semi- constructive, we call coal shelves and so on. So we know how to manage our CapEx spend to build out a long-term asset and a lot of the hydration of the kit happens when we have the demand signal. There is definitely spend for training. Even there, of course, we will only be scaling training as we see the demand accrue in any given period in time. So I would say it's more important to manage to capture the opportunity with the right product portfolio that's driving value. And on that front, I feel good about the breadth of Microsoft offering, whether it's in consumer side, whether it's on commercial per seat side or on the consumption meters, that's, I think, the fundamental driver. That is a lot of words to not answer the question directly: "Can you give us a little bit more help in understanding the timing between the CapEx investments and the yield on those investments?" What did we hear? We still do not have a sense of when this will all pay off in revenue, besides cloud demand. The answer is likely that it is a long ramp, but the capital and R&D costs are here, today. It's still a faint whiff, but in the last month or two I'm sensing more skepticism over all this investment. It's the beginning of popping this bubble. Bubbles can last a long time before they pop. The Dutch tulip craze went on for almost 3 years, and then one day it was over. But make no mistake. These Nvidia Corporation customers are not slowing down. They are signaling to us, and to each other, that they are going to stockpile GPUs because it is the scarce resource, and because they can. They don't care what analysts, or you, or I think. Not only that, but they are not going to stop buying GPUs, sitting as they are at the intersection of FOMO and YOLO. Driving all this is the idea that AI will keep advancing at the rate it did in 2022-2023. The history of AI is one of technological leaps followed by years of crawl, and eventually disillusionment. Since the release of GPT-4 in March 2023, just 4 months after ChatGPT was released, progress has slowed. But what hasn't slowed is this arms race. Given the noises everyone is making, I think it will continue into next year, in the near term, mostly to the benefit of Nvidia.