The AI Infrastructure Wave Continues: HPE/CRDO Deliver Far Exceeding Expectations, Driven by Both Optical Modules and Copper Cables
AI summary card
The AI Infrastructure Wave Continues: HPE/CRDO Deliver Far Exceeding Expectations, Driven by Both Optical Modules and Copper Cables
HPE and CRDO both reported results far surpassing expectations and raised their full‑year guidance, confirming the structural boom in AI inference and interconnect demand; Jensen Huang strongly endorsed the long‑term value of copper cabling, SK Hynix is expanding production to address chip shortages, and Google has launched an $80 billion financing round to ramp up its AI investments.
- HPE’s revenue grew 40% year‑over‑year, EPS exceeded estimates by 50%, and FY26–FY27 profit guidance was raised; backlog hit a record high.
- CRDO posted a gross margin of 68.3%, raised FY27 optical business outlook to over $600 million, and extended its growth story from copper cables to optical communication platforms.
- Jensen Huang emphasized the reliability and cost advantages of copper cabling in AI clusters, expecting it to remain the preferred solution for the long term.
- SK Hynix plans to double its capacity within five years to address the storage chip shortage expected to persist through 2030.
- Google intends to raise $80 billion for AI capital expenditures, signaling that the industry’s CapEx cycle is still accelerating.
Report interpretation
Overview
This report is a morning flash update issued by TMTBreakout, focusing on the latest developments across the AI infrastructure supply chain. It provides an in‑depth analysis of the earnings reports from HPE (Hewlett Packard Enterprise) and CRDO (Coherent), both of which delivered results far exceeding expectations due to surging demand for AI inference workloads and subsequent guidance hikes. The report also reviews the actions of major global tech giants, including Jensen Huang’s steadfast endorsement of copper cabling technology, SK Hynix’s capacity expansion plans, and Google’s massive financing initiative. Overall, the report concludes that AI infrastructure development remains in its early, high‑growth phase, with exceptionally strong and sustained demand across servers, interconnects, and storage components.
Core views
HPE (Hewlett Packard Enterprise): Comprehensive Performance Surge, Reshaping Valuation Logic HPE delivered a stellar quarter, posting revenue of $10.678 billion, up 40% year‑over‑year (compared to 18% last quarter), well above the market consensus of $9.78 billion; adjusted EPS came in at $0.79, versus an expected $0.53. Management significantly raised its fiscal 2026 profit guidance to $3.35–$3.45 per share (previously $2.30–$2.50) and outlined an 8%–12% growth outlook for fiscal 2027 (up from the market’s 5%). The primary driver behind this outperformance is surging demand for AI‑accelerated servers. Traditional server orders recorded triple‑digit growth as customers modernize their compute infrastructure, actively deploying agentic AI and inference workloads. Additionally, synergies from the Juniper acquisition have markedly boosted network segment profitability, with nominal revenue up 148% and adjusted growth around 10%; order growth is outpacing revenue. Management confirmed that backlogged orders have reached record levels, with a pipeline several times larger than current bookings, and there are no signs of front‑loading or order cancellations. On the bullish side, proponents argue that HPE has transformed into an AI infrastructure company with high‑growth characteristics; as the network segment’s contribution rises and Juniper synergies materialize, its valuation multiple could expand from the traditional 10x–12x to near 20x. On the bearish side, critics caution that current growth remains reliant on hardware pricing and supply bottlenecks; if customers rush to secure supplies, a demand cliff may follow, and the company’s exposure to enterprise rather than hyperscale cloud providers could lead to earnings volatility. CRDO (Coherent): Optical Modules Gain Momentum as a Second Growth Engine, Mitigating Valuation Concerns CRDO also posted impressive results, reporting fourth‑quarter revenue of $437 million, up 157% year‑over‑year; non‑GAAP gross margin reached 68.3%, significantly better than the market’s 65.4% forecast. Management raised its fiscal 2027 revenue guidance to over $2.4 billion (up more than 80%), with the optical modules business now expected to contribute over $600 million (up from $500 million), accounting for roughly 25% of total revenue. This adjustment marks a fundamental shift in CRDO’s growth narrative. Previously, the market worried that its heavy reliance on copper cable (AEC) technology faced a terminal valuation ceiling. However, the new guidance indicates that optical modules will become the new core engine, with ZF Optics, optical DSPs, and silicon photonics PICs each contributing over $100 million. Management noted that the high average selling price and premium margins of the optical modules business will effectively offset R&D expenses, maintaining strong operating leverage. Bullish investors believe that the addition of optical modules has transformed CRDO from a pure copper cable player into a comprehensive AI interconnect platform, capable of addressing the full spectrum of needs from copper cables to optical communications, thereby unlocking long‑term valuation upside. Bearish voices counter that the stock has already priced in much of the positive news, that the ramp‑up of optical module shipments is heavily concentrated in the second half of the year, posing execution risks, and that a rapid shift toward CPO/NPO architectures could still threaten the copper cable business. Industry Trends and Major Player Developments 1. Jensen Huang Backs Copper Cabling: At Computex, NVIDIA CEO Jensen Huang explicitly stated that, whenever feasible, copper cabling should be used for as long as possible due to its high reliability and exceptional cost efficiency. This stance directly benefits copper cable interconnect vendors like CRDO, easing concerns about a potential shift in technological pathways. 2. Tight Supply of Storage Chips: SK Hynix announced plans to double its storage chip manufacturing capacity over the next five years to address a supply crunch that could persist through 2030. This underscores that AI compute deployment is driving not only short‑term spikes but also a long‑term structural shortfall in storage capacity. 3. Google’s Massive Financing: Alphabet plans to raise $80 billion in equity to fund AI spending. According to Wells Fargo, this move not only meets immediate needs but also sends a signal to the market that the scale of AI infrastructure investment will far exceed current expectations, potentially spurring a broader uptick in CapEx among hyperscale cloud providers. 4. Other Tech Stock Updates: At Computex, Intel showcased the resurgence of x86 architecture in the AI era, emphasizing a rebalancing of CPU–GPU ratios; AMD has gained favor with Evercore, which sees accelerating CPU demand; Arm expects its own chip sales to reach the $15 billion target ahead of schedule; and Tencent’s upcoming launch of a WeChat AI agent lifted early‑morning trading in Chinese internet stocks.
Analysis framework
This report employs a dual analytical framework—performance‑driven plus industry trend‑based. First, by dissecting the financials of key names (HPE, CRDO)—revenue, EPS, gross margin, and guidance changes—we identify the specific sources of outperformance (e.g., AI server demand, Juniper synergies, optical module ramp‑up) and quantify the magnitude of the beat relative to consensus estimates. Second, drawing on management guidance and on‑site remarks (such as Huang’s statements), we distill the core industry tensions and trends. For example, we leverage Huang’s endorsement of copper cabling to validate the durability of that technological path, and we use SK Hynix’s expansion plan to reinforce the duration of the storage chip cycle. Finally, we conduct a辩证 analysis, weighing opposing viewpoints. For HPE, we balance the risks of “successful transformation” against those of “cyclical volatility”; for CRDO, we assess how the “optical module boost” addresses lingering concerns about “copper cable dependency.” This approach aims to help investors understand the structural shifts underlying earnings, rather than fixating solely on short‑term fluctuations.
Methodology notes
Judging industry prosperity and price trends through supply‑demand dynamics
The report repeatedly highlights the “supply shortage” of storage chips (SK Hynix’s expansion), server orders “outpacing revenue” leading to record backlogs, and the rigid demand for interconnects driven by AI cluster scaling. These observations are framed within the supply‑demand paradigm, indicating that the industry is in an upward cycle supported by both prices and demand.
Evaluating the sustainability of a company’s core competitive strengths
In analyzing CRDO, the report discusses how the company’s strategic pivot from a single copper cable product to an optical communication platform resolves the “terminal value” debate and builds a broader competitive moat. Similarly, HPE’s Juniper synergies are viewed as critical factors enhancing its competitive barriers.
Assessing stock valuation合理性 through the price‑earnings ratio and its growth rate
The report explicitly addresses the market’s debate over HPE’s valuation multiple (ranging from 10x–12x to potentially 20x) and the discussion surrounding CRDO’s terminal valuation. Analysts compare historical valuation ranges with projected earnings growth rates (e.g., HPE’s 8%–12% growth) to determine whether current stock prices are fairly valued or exhibit undervaluation/overvaluation.
Asset mapping & comparison
Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).
- HPE (HPE.US)Benefiting from the surge in AI inference server demand and Juniper synergies, HPE has delivered results far exceeding expectations and raised its guidance, serving as a prime example of a low‑growth hardware stock transitioning into a high‑growth AI infrastructure name.
- Strengths
- Record‑breaking order volume, robust network segment growth, improved margins, and high cash flow visibility.
- Weaknesses
- Still needs to prove whether its AI business can sustain high growth and avoid reliance on price hikes alone.
- Comparison
- Compared to traditional server vendors, HPE enjoys greater advantages in networking and business diversification, positioning it for a reevaluation of its valuation multiple.
- Risks
- Customers may delay purchases due to budget constraints, or front‑load orders to lock in prices, creating a subsequent demand vacuum.
- CRDO (CRDO.US)As a leader in the AI interconnect space, CRDO has successfully broadened its growth trajectory through the rapid ramp‑up of its optical modules business, alleviating market concerns about overdependence on copper cable technology.
- Strengths
- Exceptionally high gross margins (68%+), increasing share of optical modules in revenue, and a product lineup covering both copper cables and optical communications.
- Weaknesses
- Optical module shipments are concentrated in the second half of the year, posing execution risks; customer concentration remains relatively high.
- Comparison
- Compared to pure copper cable manufacturers, CRDO’s optical capabilities make it more resilient amid shifts in industry technological pathways.
- Risks
- If the industry accelerates toward CPO/NPO architectures, copper cable demand could peak prematurely; intensifying competition in the optical module market.
- SK Hynix (005930.KS)As a leading storage chip maker, SK Hynix’s ambitious expansion plan to address the long‑term shortage driven by AI represents a core beneficiary of the rising storage chip cycle.
- Strengths
- Strong commitment to capacity expansion and high certainty of market demand.
- Weaknesses
- Expansion is constrained by land availability, equipment costs, and significant capital expenditure pressures.
- Comparison
- Among storage chip producers, SK Hynix leads in high‑end DRAM and NAND technologies.
- Risks
- If demand falls short of expectations after capacity increases, prices may decline.
- GOOGL (GOOGL.US)Its massive financing initiative underscores a long‑term commitment to AI infrastructure and could set the tone for a broader rise in industry CapEx.
- Strengths
- Robust balance sheet and a mature AI ecosystem.
- Weaknesses
- The large financing round may cause short‑term dilution; ROI returns warrant close monitoring.
- Comparison
- Compared to other cloud providers, Google excels in integrating AI models with its search and advertising platforms.
- Risks
- The long payback period of AI investments could weigh on the stock if commercialization progresses slowly.
Key data
- HPE Q1 FY26 Revenue10.678 billion USD40% yoy growth; market consensus was 9.78 billion USD
- HPE Q1 FY26 EPS0.79 USDMarket consensus was 0.53 USD
- HPE FY26 Profit Guidance3.35–3.45 USD per shareRaised from 2.30–2.50 USD per share
- CRDO Q4 FY26 Revenue437 million USD157% yoy growth
- CRDO Q4 FY26 Gross Margin68.3%Market consensus was 65.4%
- CRDO FY27 Optical Modules Revenue Outlook>600 million USDRaised from >500 million USD
- SK Hynix Capacity Expansion PlanDouble capacityTo be completed within five years
- Google Proposed Financing Amount8 billion USDFor AI capital expenditures
Impact & implications
The report contends that the wave of AI infrastructure construction has only just begun, with HPE and CRDO’s results validating its intensity. For investors, this implies a need to reassess the valuation logic of hardware companies, shifting from simple cyclical plays to growth‑oriented AI platform equities. Meanwhile, the prolonged storage chip shortage and Google’s massive financing suggest that the entire AI supply chain’s capital expenditure cycle will be longer and more aggressive than anticipated. Copper cabling technology has not been marginalized; on the contrary, it demonstrates irreplaceable cost and performance advantages in the short term, benefiting related supply chain firms going forward.
Risks
- AI infrastructure demand falls short of expectations, slowing sales of servers and interconnect devices.
- The storage chip supply shortage eases faster than anticipated, triggering price declines.
- Rapid shifts in technological pathways (e.g., CPO/NPO) prematurely render existing solutions like copper cabling obsolete.
- Geopolitical risks disrupt global supply chain stability and tech giants’ capital expenditures.
- Customer budgets tighten, delaying or canceling AI‑related procurement plans.
What to watch
- HPE and CRDO’s sequential order growth and inventory dynamics.
- SK Hynix’s capacity expansion progress and storage chip price trends.
- The actual implementation of AI capital expenditures at Google and other cloud providers.
- NVIDIA and major chipmakers’ latest guidance on copper cable vs. optical module strategies.
- The commercial rollout and user adoption of AI applications (e.g., agents).