AI and IoT: The Driving Forces Behind Tech Infrastructure Investments

Astera Labs Stock: Powering The Future Of AI (NASDAQ:ALAB)

My price target for Astera Labs is $795 by 2031, implying 1322% upside from current levels. With the rise of generative AI applications and the increasing number of hyperscale data centers, the semiconductor industry is seeing all-time interest from investors due to chips' critical role in data centers. In turn, companies serving the semiconductor and data center ecosystems have the potential to witness substantial growth in the long term. One of those companies is Astera Labs, Inc. (NASDAQ:ALAB) which manufactures semiconductors addressing network bottlenecks to improve the speed and efficiency of data flow within data centers. With its stock down 41% from the highs it reached following its IPO last March, I believe Astera Labs is fairly valued at its current valuation, making it an ideal entry point for long term investors. As is, the company is estimated to dominate 80% or more of the PCIe retimer market, per JPMorgan (JPM) analysts. More so, the company works closely with leading GPU manufacturers such as Nvidia (NVDA), AMD (AMD), and Intel (INTC), per its S-1 filing, to ensure that its chips are compatible with all GPUs used in hyperscale data centers. This is especially beneficial for the company as this might be the reason why Astera Labs ships retimers to all major hyperscalers and AI platform providers. In my opinion, these relationships, as well as the company's potential technology edge could see its reportedly 80% market share in the retimers space grow in the coming years. In light of these factors, I'm rating Astera Labs as a buy with a price target of $795 by 2031, implying 1322% upside from its current valuation. Astera Labs designs, manufactures, and sells retimers that connect different parts of a data center together with a main focus on AI and cloud computing needs. These retimers are extremely critical for hyperscale data centers since they allow data centers to handle the growing demands of AI and cloud computing by facilitating efficient data transfer. While Astera Labs is mainly competing with semiconductor giants Broadcom (AVGO) and Marvell (MRVL) in the retimer space, it differentiates itself by being entirely focused on high speed interconnect solutions. Meanwhile, Broadcom and Marvell are more diversified with broader product portfolios. In my opinion, this focus allows Astera Labs to be more innovative in the retimer space. At the same time, Astera Labs' focus on high speed interconnect solutions is probably why it is seeing rapid topline growth. The company's revenues grew 45% YoY in 2023 from $79.8 million to $115.8 million, and reported revenues of $65.3 million in Q1 2024. The company expects Q2 2024 revenues to grow 10-12% sequentially, implying $72.4 million at the midpoint, as shared in the Q1 earnings call. In terms of its financial health, Astera Labs has total liquidity of $801.4 million, with $696 million in cash on hand and $105.3 million in marketable securities, and has zero debt on its balance sheet. This strong financial position allows the company to explore growth opportunities such as acquiring startups with breakthrough technologies, which could help it maintain its rapid topline growth. The core of my bullish thesis on Astera Labs is its growth potential along with the broader AI market. According to Statista, the number of hyperscale data centers has grown exponentially from 259 in 2015 to 992 in 2023. At the same time, Dell'Oro Group forecasts global data center capex to rebound this year to 11% growth after only growing 4% YoY in 2023 from $241 billion in 2022. This indicates that global data center capex is expected to top $278 billion this year, driven by select hyperscalers returning to an expansion cycle. The forecasted growth in data center investments bodes well for Astera Labs since retimers are a critical part of data centers. As is, the retimer market was valued at $167.7 million in 2022 and is forecasted to reach $5.3 billion by 2031, growing at a CAGR of 46.77% during this period. The retimer market is expected to exhibit such growth due to PCIe retimers becoming essential in overcoming signal degradation and extending signal reach in data centers, which is crucial for the efficiency of data centers. In my opinion, Astera Labs stands to benefit substantially from the expected growth in the broader AI industry as well as the retimer industry thanks to its deep relationships with key players in the data center ecosystem such as Nvidia, AMD, and Intel. According to the company's S-1 filing, its products are integrated into reference and commercial designs provided by Nvidia, AMD, and Intel, showcasing its position as a reliable supplier to these giants. Additionally, Astera Labs has a deep relationship with the world's largest hyperscaler, Amazon (AMZN), which is reported to be one of Astera Labs' largest customers. The tech giant is one of the hyperscalers expected to increase their data center investments in the coming years as it intends to invest $150 billion over the coming 15 years in data centers to meet the rising demand for AI applications and other digital services. Amazon's plans to increase its investments in data centers is good news for Astera Labs thanks to both companies' existing warrant agreement. Per Astera Labs' S-1 filing, the company issued warrants to Amazon to purchase up to 1,484,230 shares at $20.34 per share in 2022. This agreement was amended in October 2023 and the company issued additional warrants to purchase up to 831,945 shares at $20.34 per share. Overall, Amazon holds warrants to purchase up to 2,316,175 shares at $20.34 per share, of which only 232,608 shares were vested. For the rest of the shares to be vested, Amazon must purchase up to $650 million worth of Astera Labs products. As such, Astera Labs stands to benefit from Amazon's plans to increase its data center investments in the form of revenues as well as up to $42.4 million in cash from the vested shares, in case Amazon exercises the warrants. Thanks to Astera Labs' first mover advantage in the retimer space, it was able to dominate this market with a market share of 80% or more, per JPMorgan analysts. Considering the forecasted growth in the retimer market and the company's relationships with industry giants, I expect its market share to grow over the coming years. These relationships might give Astera Labs earlier access to upcoming chip architectures and technology roadmaps. In this way, Astera Labs could design retimers compatible with next-gen technologies from those giants and launch them into the market before its competitors. Another factor that could see Astera Labs' market share grow in the coming years is its technological edge over its competitors, Broadcom and Marvell. The company's Aries 6 retimers, currently being sampled, was showcased at Nvidia's GTC 2024 where it was linked at PCIe Gen6 x16 speeds at 10-11W of power consumption. This shows that Astera Labs' next-gen retimers have lower power consumption than Broadcom's next-gen retimers, Vantage 6. In my opinion, the power efficiency of Astera Labs' retimers may be a major attraction point for its customers. Data centers consume massive amounts of energy and electricity costs represent a significant expense for hyperscalers. At the same time, power limitations can restrict data centers' ability to scale and add new computing resources. Therefore, using power efficient solutions, such as Astera Labs' retimers, can allow hyperscalers to pack more processing power into their facilities while also reducing their electricity bills. Based on these factors, I expect Astera Labs' market share in the retimer market to grow to 82% in 2025 and 2026, 83% in 2027 and 2028, 84% in 2029 and 2030, and 85% in 2031. According to the aforementioned retimer market forecast, my revenue projections for Astera Labs until 2031 are as follows. According to my revenue estimates, Astera Labs would be trading at the following EV/sales multiples. While Astera Labs might seem expensive at its current EV/sales multiple compared to Broadcom's 16.66 and Marvell's 12.21, it doesn't look as such when compared to other companies with direct exposure to the AI sector such as Nvidia, AMD, and ARM Holdings (ARM). As is, these companies are trading at the following EV/sales multiples. I believe Astera Labs should be valued relative to these giants not Broadcom and Marvell due to its higher exposure to the AI sector. The company's core products directly address data transfer bottlenecks within a data center which is crucial for data centers. Meanwhile, Broadcom and Marvell's AI-related products only represent a portion of their broader business focus. As such, investing in Astera Labs would provide investors more exposure to the AI sector compared to Broadcom and Marvell, in my opinion. Accordingly, my price targets for Astera Labs until 2031 are as follows. Please note that I didn't use EV/EBITDA to reach my price target for Astera Labs since it might have to increase its R&D costs substantially to keep up with the rapid technological advancements in AI applications. These developments would require the company to invest in developing new products to maintain its leading position in the retimer industry. The daily chart shows that Astera Labs has been in a downward trend since late April. That said, the stock appears to be forming a double bottom chart pattern as it is holding its current support level near $55. With the company reporting its Q2 earnings on August 6, a solid earnings release could help the stock break out of its downward trend to confirm the double bottom pattern. In that case, the stock could reach $68 which would represent 22% upside from the current share price of $55.89. The main risk to my bullish thesis is Astera Labs' revenue concentration. Per the company's S-1 filing, its top 3 customers were responsible for 89% and 70% of its revenues in 2022 and 2023, respectively. Therefore, if the company loses one of those customers or if one of those customers reduces their demand for Astera Labs' products, its revenue growth would be at risk of turning negative. In summary, I'm bullish on Astera Labs due to its growth potential along the broader AI industry. The company's products are crucial for data centers as they address network bottlenecks, allowing for faster and more efficient data flow within data centers. Astera Labs is already a dominant force in the retimer market with an 80% market share and has deep relationships with key data center infrastructure players such as Nvidia, AMD, and Intel allowing it to manufacture retimers compatible with these giants' products. At the same time, the company has a deep relationship with the world's largest hyperscaler, Amazon, which intends to invest $150 billion in data centers in the coming 15 years. In light of these factors as well as the company's potential to gain market share due to its advanced technology, I'm rating Astera Labs as a buy with a price target of $795 by 2031, representing 1322% upside from current levels.

Sterling Infrastructure: The Picks And Shovels To The Future Of AI (NASDAQ:STRL)

Despite shares having doubled in the last 12 months, the valuation is still compelling enough because of the growth ahead. Sterling Infrastructure (NASDAQ:STRL) is an infrastructure company in North America that's benefitting from long-term tailwinds in AI. Over the last few years, the company has grown its top-line double digits while experiencing some margin expansion over time, as the business mix shifts to more e-infrastructure. Despite the company's shares having more than doubled over the past year, I still see good value here. Sterling Infrastructure is a specialized construction and engineering company that's involved in three verticals: e-infrastructure, transportation, and building solutions. In e-infrastructure, Sterling Infrastructure is involved in building out large-scale projects for data centers, e-commerce distribution centers, warehouses, and energy companies. This might sound simple at first glance, but many of its customers often need specific requirements that necessitate custom solutions tailored to their operational needs. Essentially, what this means is that considerations like high-capacity data storage, robust network connectivity, scalable computing power, and energy-efficient solutions to minimize operational costs and environmental impact, need to be taken into account from beginning to end. As such, the company is involved with help its customers, many of which are large, blue-chip customers, from site selection all the way through completion. As the company's largest and fastest growing segment, the e-infrastructure business generates 48% of the company's total sales. In the transportation segment, Sterling Infrastructure is involved in the construction of projects like roads, bridges, highways, airports, and ports -- all things that are needing in helping people get from one place to the other. This is generally a lower margin segment that does 6.6% operating margins (compared to 15% for the e-infrastructure business) and represents 32% of total sales. Finally, in the building solutions segment, the company serves North America's top builders in Texas and Arizona, which are high growth markets for construction for single-family and multi-family homes. While the building solutions is the company's smallest segment at 20% of revenues, it's still growing at has good operating margins of 11.4%. Shares of Sterling Infrastructure have been on a tear as of late, with the company's shares more than doubling over the past twelve months. We can see the divergence in total return compared to the S&P500, where Sterling's shares started ripping in early 2023. Over the last five years, the company's shares have delivered a total return of 792%, compared to the S&P 500's return of 102%. While a valuation re-rating certainly played a large role in this, as e-infrastructure for data centers (with the emergence of AI) have become increasingly in the spotlight for investors. However, Sterling Infrastructure has continued to deliver impressive returns for shareholders, growing both the top and bottom line. Over the last twenty years, the company has delivered CAGRs in revenue and EBITDA of 13.8% and 15.4%. More recently, in the last five years, the company has compounding revenues and EBITDA at 13.7% and 33.9%, respectively (source: S&P Capital IQ). Clearly, Sterling Infrastructure is a company whose growth rates are not slowing down, as evidenced by double-digit top line growth which translates to the bottom line with some margin expansion. Sterling Infrastructure's financial performance (revenue and EBITDA growth) has a huge impact on free cash flow conversion as well. As illustrated by the chart below, the company's operating cash flow has been nearly doubling every 2 years, providing flexibility as well as capital for growth investments. So what's been driving this growth? One of the biggest reasons is the company's e-infrastructure growth. While this makes up nearly half of the company's revenues today, this wasn't always the case. As illustrated below, the e-infrastructure business really only got started in 2019, and since then, it's seen massive success post-pandemic. When we look at the smart infrastructure market, the runway for growth is massive. As a $187 billion market, the market globally is poised to grow at a 23.1% CAGR to just under $1 trillion by 2032. Most of this growth was triggered out of the pandemic, where global lockdowns caused disruption to production and networks, highlighting the need for e-infrastructure to be robust. Within the AI vertical alone, a technology that has seen huge advancement in growth at an 18.5% CAGR. In my view, the spending from blue-chip companies is going to be massive Just a few months ago in May, Microsoft (MSFT) signed a deal with Brookfield (BN:CA), one of the world's largest infrastructure investors, to supply 10.5 gigawatts of renewable power capacity between 2026 and 2030. This was an important deal for Microsoft because the deal will help it meet AI demand as well as meet its emission commitments on ESG. As the largest deal of its kind (almost 8x larger than the largest single corporate PPA ever signed), I'd expect more investments like this one. While Microsoft and other blue chips companies lead the way for now, I'd also expect mid-sized companies to start investing in e-infrastructure as well. Companies like Sterling Infrastructure are almost certainly going to benefit from this trend. Particularly given their strong reputation in the industry working with other blue-chip companies and the fact that current capacity in data centers only represents a small fraction of what is needed to support artificial intelligence and other emerging technologies. When we look at the company's backlog of projects, management believes that the growing backlog should support the company's future plans as well as margin expansion over time. At the end of the most recent quarter, in spite of a small 10% revenue decline in the quarter, the company's infrastructure awards totaled $332 million, for a backlog of 961 million, a 32% increase compared to Q1'23. As for the mix of the backlog growth, management commented that data centers were again the largest contributor to awards growth as customers are feeling more comfortable building out their AI and technology infrastructure advancements. To date, data centers are now about 40% of the e-infrastructure backlog, highlighting just how important this growth is. In terms of the outlook, management is guiding for revenue of $2.125 to $2.215 billion with EBITDA of $285 to $300 million for FY'24. On EPS, management is guiding for earnings per share of between $5.00 and $5.15 for 2024, indicating that they see margins continuing to expand from here along with more awarded contracts. Given the backlog, these targets should be achievable for the year, in my view. Another reason to like Sterling Infrastructure is for its strong capital allocation framework. While the company doesn't pay a dividend, the company invests heavily into organic growth initiatives, M&A, as well as buybacks. Of the company's capex, most of its investments are tilted towards the e-infrastructure business, an area the company has been investing in so that it can better meet the demand from AI in the years to come. In addition to M&A, Sterling Infrastructure targets tuck-in and bolt-on acquisitions, particularly small and mid-sized deals where an acquired company would add new customers or complement the existing service offerings. With respect to M&A, on the earnings call, management commented on the types of deals they're looking for: Yeah. We are. We've seen several, we're looking at several. One of the core things for us is we're very picky. We want really good businesses with really good people that we think we can double over a five-year or six-year period. So it's getting that right match and that right opportunity put together. But they're out there, yes. And they're of various sizes. Clearly, management is very focused on finding deals with the right fit and not just doing deals for the sake of growth. In other investor materials, the company has also signaled that they have a preference for doing margin-accretive deals with end-markets that are attractive. Paying a fair price and finding the right fit means a high hurdle rate for the company, but this is what you want to see in M&A, given that it reduces the risk of integration of a deal being dilutive with inadequate synergy potential. When it comes to share repurchases, management is taking an opportunistic approach to buying back shares. In 2023, the company repurchased $9.6 million and the year before it purchased $9.4 million worth of shares (source: S&P Capital IQ). Given that shares have run up a lot lately, I don't expect management to be aggressive in buying back shares as the company's shares aren't exactly cheap nor is it likely a great place to allocate capital. As such, even with a $200 million authorization under the share repurchase program, I don't foresee significant buybacks being a major part of the company's capital allocation framework. From a balance sheet perspective, the company has a bit of debt, which includes $337 million in term loan borrowings. It also has a $75 million revolving credit facility, which is currently undrawn. With cash of $480 million at quarter end, cash exceeds total debt. Scheduled term loan debt repayments are for $26.3 million in 2024, $26.3 million in 2025, and $6.6 million in 2026. With an EBITDA Debt Coverage Ratio of 1.1x, the company has minimal leverage, but is comfortable taking that to 2.5x, should opportunities present themselves. Currently, there's only one sellside analyst who covers Sterling Infrastructure. He has a price target of $130.00, which implies about 6.4% upside from the current price. Given the hold rating and limited upside potential, shares of Sterling are likely fairly valued. I would disagree with the analyst here. When we look at the historical EV/EBITDA multiple, it's clear that the company might look expensive at 11.2x EV/EBITDA, a near doubling of where it was trading before 2023 at around 6.0x EV/EBITDA (source: S&P Capital IQ). In my view, what I think most have valued to realize about Sterling Infrastructure is that the company's business mix has changed tremendously in just the last three years. Compared to 2021, when the e-infrastructure business made up only 32% of revenues, Sterling's margins were vastly inferior at around 10% EBITDA margins (source: S&P Capital IQ). Given the transition to better growth markets, the shift in business mix to higher margin businesses, and the fact that the company can likely continue growing in the mid-teens for the foreseeable future (in line with the industry growth rate), I don't see why paying 11.2x EV/EBITDA would be expensive. When looking at the forward multiples of the company, Sterling Infrastructure looks even cheaper. When it comes to the risks to the investment thesis, the main ones would be a slowdown in AI spending. Macroeconomic pressures like higher interests rates are generally poor for infrastructure spending, as the cost of borrowing/ financing makes funding projects more expensive. For e-infrastructure in particular, I'm not too worried given that there is a huge runway for growth with AI having long-term secular tailwinds, that may prove to be several decades, rather than just a few years. Another risk would be competition. More market entrants into the market would mean that Sterling Infrastructure would need to sacrifice price in order to win business, potentially providing a headwind to margins. With respect to the company's competitive advantage, CEO Joseph Cutillo had this to say on the company's Q4'23 earnings call: Our competitive advantage is really, again at the end of the day, I look at us as an insurance policy, and the bigger the project, the better the policy, and our ability to get very large complicated site development done in a very short period of time for our customers. As you can imagine, if our project, which was a big project, would be $100 million, let's say that's part of a $2.5 billion to $3.5 billion project. And if we say, we'll get it done in six months and we're two weeks late, that project will be six months to eight months late on the back end with the domino effect or snowball effect of the trades. And the cost of capital on the rest of that, and the loss of revenue and profitability associated with incremental six months is a very, very small price to pay a slight premium to us over the rest to ensure that they get it. Given the company's size and scale, I think this isn't a risk to be too concerned with. My estimation is that given the high demands of customers (particularly complex blue chip projects), customers would rather pay more for a trusted company like Sterling to work on their projects for them, end to end. So to summarize, Sterling Infrastructure really is a picks and shovels company to the AI industry, piggybacking off of the long-term tailwinds of e-infrastructure buildout, particularly in data centers. Over the years, the company has grown at a decent clip with margin expansion over time, that should continue as the business mix shifts to more e-infrastructure. Because of the higher margin and faster growing end-markets that the company is positioned in, I think the recent re-rating in the company's valuation is warranted. So in spite of 100%+ rise in the company's shares over the last year, I think this is a company that has only now starting to get discovered and getting the valuation it deserves. As such, I rate the company's shares as a 'buy'.

Altair Stock: Too Expensive For Far Too Little Growth (NASDAQ:ALTR)

With so little to justify its valuation, Altair is a stock to sell. As the stock market continues to look ahead for Fed rate cuts, the fact that major indices are hovering at all-time highs has me rotating my portfolio into "growth at a reasonable price" stocks. Now, there are two inviolable pieces of that formula: the companies must have a strong multi-year growth trajectory, and it must be trading at an out-of-favor price that can buck the broader market trend. Altair Engineering (NASDAQ:ALTR), unfortunately, has neither. This legacy software company, known best for its CAD (computer-aided design) software programs and now pitching itself as a "computational intelligence" platform that helps companies in various industries build computerized models, has risen ~20% year to date, roughly keeping pace with the broader markets. It is also one of the slowest-growing mid-cap software companies in the market: not declining, which is a relief for a company that was founded in the '80s, but certainly not growing anywhere close enough to justify its year to date rally. I last wrote a neutral article on Altair in 2021, at the time noting that there were few prospects to lift the stock materially higher. Since then, the stock has appreciated modestly (but vastly underperforming most tech stocks, particularly in the software sector), while growth rates have continued to diminish from the mid-teens in 2021 to the mid single digits now. As a result, I'm rating Altair as a sell. In my view, the company had the fortune of benefiting from a "rising tide lifts all boats" trend in the tech sector this year, and the markets gave it credit for AI demand drivers even though the company's own growth rates are quite paler than many software peers. The chief argument against Altair is its valuation. At current share prices near $99, Altair trades at a market cap of $8.24 billion. After we net off the $557.6 million of cash and $307.8 million of debt on the company's most recent balance sheet, the company's resulting enterprise value is $7.99 billion. Meanwhile, for the current fiscal year, Altair has guided to a revenue range of $652-662 million, or 6-8% y/y growth. Wall Street analysts also have a consensus pro forma EPS target of $1.24 for this year (+10% y/y) and $1.39 next year (+12% y/y). The Street is also expecting revenue growth of 9% y/y to $717.2 million - which we note represents a meaningful acceleration to current growth rates. This puts Altair's valuation multiples at: No matter how we slice it, we can't find a way to justify Altair's valuation multiples, either on a revenue basis or on an earnings basis. Though we note Altair has been profitable for years (not something to take for granted in the software sector), its relatively meager earnings are nowhere near enough to support the company's ~$8 billion enterprise value. We note that there is a wide array of peers in the software sector that have been punished for decelerating growth this year. Examples include Asana, Freshworks, Okta, and many others. And yet each of these companies are still posting growth rates in excess of 10%+ and are trading at much more modest revenue multiples: In other words, Altair has continued to rally this year even though the stock already had a bloated valuation relative to the revenue and earnings it can deliver. And without major growth catalysts that can propel this company back to double-digit growth rates, I don't see a path to this rally continuing for Altair. Sell this stock and steer clear here. Recall that consensus is calling for revenue growth to accelerate in FY25, well beyond the company's current growth rates. This is essentially a macro statement that better economic tidings may unspool unfulfilled demand next year, but so far we've not seen any evidence of growth improving. Take a look at the latest quarterly results below: Revenue grew only 4.1% y/y in Q1 to $172.9 million, which slightly beat Wall Street's expectations of $169.2 million. Constant currency normalization provides a bit of relief, but not much: on an FX neutral basis, the company's growths would have been one point stronger at 5.1%. The company even noted that part of the Q1 outperformance was driven purely by deal timing, with some deals closing earlier than the company initially expected. While this is certainly better than the opposite (many software peers have reported deal cycles elongating and closings pushing out, a result of tighter IT budget scrutiny), the pull-in may rob revenue from future quarters. The company's Q2 outlook calls for as-reported growth rates of 2.7% to 4.8% (with roughly two points of adverse FX impact) - which, at the midpoint, is still slower than Q1's growth. Another indicator of softness: Altair's billings in the quarter were only $154.1 million, which declined -6% y/y and trailed revenue on a nominal basis. As seasoned software investors are aware, billings is the best forward-looking indicator of a subscription software company's growth trajectory, and a year/year decline in billings is a leading predictor of growth slowing down (let alone consensus hopes for acceleration). The company is making a tuck-in acquisitions of a company called Cambridge Semantics, a "knowledge graph" technology (announced in April), and another company called Metrics Design Automation (announced in July). No financial metrics were disclosed, but these acquisitions appears to be more for technology development rather than for providing an immediate, accretive revenue boost. Neither is Altair achieving anything impressive on the profitability front. Adjusted EBITDA in the quarter clocked in at $45.8 million, up 6% y/y and representing a margin of 26.5%, only slightly better than 25.9% in the year-ago quarter. For such little growth, red flags in billings declining, and limited upcoming catalysts to expand its market size, Altair has a lot of enthusiasm built up into its share price, which is trading at revenue multiples that are significantly above faster-growing peers. To me, all of these indicators point to a sell for this stock. There are only a few notable "upside risks" for this stock. The company has noted that it has performed well in the aerospace and defense sector. These types of clients can often generate very large deals, particularly in the public sector: but are also highly cyclical and not to be counted on. The second factor is that lower interest rates may also spur more new construction, which may provide an uplift for its CAD products. All in all, however, I mostly see a company with poor growth rates trading at a very high price, with very few catalysts to the upside.

Silicon Labs Ready For The Next IoT-Fueled Run (NASDAQ:SLAB)

With expanding addressable market opportunities, share gains, and new product launches, SLAB could achieve mid-teens revenue growth and mid-to-high teens operating margins in three years, driving a fair value of $130-$150. Writing about Silicon Laboratories (NASDAQ:SLAB) almost 18 months ago, I said that while I liked the long-term potential of this connectivity-focused IoT semiconductor company, I preferred to wait for a pullback. I didn't expect a virtual collapse in the business (a 70% peak to trough move in revenue), nor a "pullback" of roughly 50% in the shares, as the company saw significant weakness among its customers, as well as increased competition throughout 2023. Business appears to be stabilizing, though, and I like the prospects for Silicon Labs to continue gaining share in Bluetooth and WiFi and leveraging enhanced compute ability in its Series 3 products to participate in AI edge deployments. I don't think my low double-digit long-term revenue growth target is anything like "conservative", but I can still argue for a fair value in the $130 to $150 range on improving demand and improving product capabilities. Building on a return to sequential growth in Q1'24 (revenue was up 22% from the Q4'23 bottom), I do expect further recovery in this upcoming second quarter report. Management is usually pretty good about its quarter-forward guidance (large beats/misses relative to sell-side expectations are rare), and the 32% qoq guide also jives with guidance given from fellow IoT connectivity player Nordic Semiconductor (OTCPK:NRSDY) when they last provided guidance. Not to oversimplify, but destocking in consumer markets (smart home, mostly) appears largely over, and SLAB has been actively working down channel inventories, with the last quarter seeing those inventories return to the 60-70 day target range. At the same time, the company is on the front end of leveraging past wins in areas like electronic store shelves, smart metering, and glucose monitors (DexCom (DXCM), I believe), as well as share gains in Bluetooth more generally. I'm less confident in the near-term outlook for the industrial/commercial business. Although process automation has been holding up well, discrete/factory automation activity has slowed significantly and I'm not a big believer in the second half rebound hypothesis. While I do expect an ongoing rebound in Industrial and Commercial from a low trough, I expect Home & Life to be relatively stronger. My expectations for margins are not that bold. While I think Silicon Labs management did a decent job of managing expenses relative to a much steeper downshift in the business than initially expected, gross margin is still likely to be lackluster (in the low 50%'s), and likewise with operating margin - I think positive operating earnings (non-GAAP) is a big ask unless and until revenue gets back to around $900M on an annualized basis (roughly a year from now). The basic bull argument for IoT hasn't changed despite this cyclical downturn. IoT enables wireless convenience as well as monitoring and processing capabilities that haven't previously been possible. Smart meters save utilities money on data collection and can alert both utilities and property owners to potential problems (leak detection, et al). Smart cameras can not only record, but provide real-time alerts and trigger other compute functions (like automatically reading a license plate or identifying someone). Smart thermostats can automatically adjust the indoor climate based on power demand (many utilities offer discounts to homeowners who install thermostats that can adjust based upon peak demand) and/or the presence/absence of residents. Smart glucose meters can provide not only real-time data and alarms, but trend analysis and other feedback. So on and so forth. Silicon Labs has long focused more on the connectivity side of IoT and has a wide portfolio of connectivity technologies. Bluetooth and WiFi have emerged as increasingly important technologies, though, and I expect newer offerings like Series 2 and Series 3 to grow SLAB's share in BT and WiFi over the coming years; I don't see SLAB seriously challenging Nordic in Bluetooth (it has over 40% share), but there are plenty of marginal players (a large block of "other" on Bluetooth market share charts) that Silicon Labs can take share from in the coming years. I'm also intrigued by the upcoming Series 3 portfolio, which Silicon Labs is currently sampling. The main attraction here for me is that the compute capacity for these chips is 100x that of Series 2, making them viable options for AI edge compute installations. This is a big area of focus for a lot of automation companies, with Emerson (EMR) among those who are very excited about what AI-enabled edge IoT devices can offer in areas like monitoring/safety, maintenance, and process analysis. Now for a "but". One factor that I do believe contributed to the especially steep decline in Silicon Labs' revenue, and could be a headwind during this recover, is the re-entry of larger semiconductor players into the IoT market. I believe that companies like NXP Semiconductors (NXPI), STMicro (STM), and Texas Instruments (TXN) ceded opportunities in IoT when they were working through chip shortages (so as to pursue higher-margin opportunities and/or important long-term strategic opportunities like auto MCUs) but started coming back into the market as those shortages eased and demand in other end-markets weakened. In other words, the last peak for SLAB may have been "artificially" boosted and the company's recovery and expansion phase may be more challenging than a simple retracement. Likewise, opportunities like AI edge are not exclusively for Silicon Labs, as I expect others like Lattice (LSCC) to fight hard for some of the same real estate, and Lattice has a lot of relevant experience targeting these markets and developing rich software toolsets to complement their FPGA offerings. Between more "classic" IoT opportunities like smart metering and smart home devices, real-time emerging opportunities like smart appliances, smart shelf labels, and glucose monitoring, and still-emerging opportunities like AI edge compute, I like the growth outlook for SLAB. I think management's target of 20%-plus growth is likely to be too high to reach, but I do think the business can grow at a mid-teens rate from the 2023 starting point and at a low double-digit rate over the long haul. Margins will need time to bounce back, but there should be significant pent-up incremental margin leverage as the business recovers to $900M-$1,000M in revenue and then grows beyond. I'm expecting a return to mid-to-high teens operating margin in three years and I'm not ruling out 20% in five years. Discounted cash flow suggests attractive upside into the $140's if low double-digit long-term revenue growth and long-term adjusted free cash flow margins in the low-20%'s are viable expectations. Near-term valuation approaches are trickier. As regular readers know, I favor an operating margin-driven EV/revenue approach, but it doesn't really work well when margins are at the extremes, and I don't believe the negative operating margin I expect for FY'24 is representative. For now, then, I'm using my three-year expectations for margins and revenue and discounting back at a high single-digit rate. That gets me to around $130, which is admittedly not the most exciting target, but also doesn't include any premium for scarcity value (as an IoT pureplay) or above-average growth, both of which the market has been willing to grant in the past. The results from Silicon Labs over the last 18 months clearly demonstrate that however much growth there may be in the IoT opportunity, that doesn't excuse it from the volatile cycles that have historically characterized the semiconductor industry. While I think there were some unusual factors at play that magnified the cycle, what matters now is that the business appears to be coming off the trough and setting up for another significant growth cycle. Of course there are still risks to the macro outlook that could impact SLAB, not to mention risks from competitors. Even so, I think this is a name for more aggressive investors to consider as a play for the next up-cycle in the space.