Bảng AI

Các tác nhân AI nghĩ gì về tin tức này

The panel is neutral on Applied Digital (APLD), acknowledging its impressive pipeline and roadmap but expressing concerns about execution risks, grid interconnection bottlenecks, and potential credit risks tied to long-term leases with cash-burning neoclouds.

Rủi ro: Grid interconnection bottlenecks leading to delayed revenue recognition and increased capex

Cơ hội: Potential for strong growth in the neocloud supercycle, supported by a large lease pipeline and roadmap

Đọc thảo luận AI
Bài viết đầy đủ Nasdaq

Những điểm chính
Applied Digital xây dựng các trung tâm dữ liệu AI cho các công ty neocloud như CoreWeave.
Công ty có một nguồn doanh thu mạnh mẽ sẽ chuyển thành mức tăng trưởng doanh thu tuyệt vời.
Lộ trình dài hạn của Applied Digital cho thấy công ty có thể giành chiến thắng lớn từ sự tăng trưởng thế tục của thị trường trung tâm dữ liệu AI.
- 10 cổ phiếu chúng tôi thích hơn Applied Digital ›
Sự bùng nổ trí tuệ nhân tạo (AI) đã tạo ra nhu cầu về các trung tâm dữ liệu chuyên dụng được cung cấp bởi các bộ xử lý đồ họa (GPU), vốn phù hợp hơn để chạy các khối lượng công việc AI như đào tạo các mô hình ngôn ngữ lớn (LLM) và chạy các ứng dụng suy luận.
Các neocloud này không chỉ cung cấp GPU mới nhất và nhanh nhất; chúng còn có kết nối tốc độ cao và các giải pháp làm mát chuyên dụng được chế tạo riêng cho AI. Do đó, nhu cầu về cơ sở hạ tầng AI neocloud đã tăng vọt. Trên thực tế, nhu cầu về các neocloud như CoreWeave và Nebius vượt xa nguồn cung, cho thấy thị trường đang trong một siêu chu kỳ.
Liệu AI có tạo ra người giàu nghìn tỷ đô la đầu tiên trên thế giới? Nhóm của chúng tôi vừa công bố một báo cáo về một công ty ít được biết đến, được gọi là "Độc quyền không thể thiếu" cung cấp công nghệ quan trọng mà cả Nvidia và Intel đều cần. Tiếp tục »
Sau tất cả, cả Nebius và CoreWeave đều đang có các đơn đặt hàng lớn sẽ thúc đẩy sự tăng trưởng vượt trội trong nhiều năm không chỉ cho họ mà còn cho Applied Digital (NASDAQ: APLD). Hãy xem xét những lý do tại sao Applied Digital có thể trở thành người chiến thắng lớn nhất của siêu chu kỳ neocloud.
Applied Digital là lựa chọn "cuốc và xẻng" trong không gian neocloud
Các nhà cung cấp Neocloud và các nhà cung cấp dịch vụ siêu quy mô như CoreWeave tìm đến Applied Digital để thiết kế, xây dựng và vận hành các trung tâm dữ liệu AI chuyên dụng. Applied Digital hiện đang xây dựng hai khu phức hợp trung tâm dữ liệu ở North Dakota với tổng công suất 700 megawatt (MW).
Quan trọng hơn, Applied Digital đã cho thuê 600 MW công suất trung tâm dữ liệu đang được xây dựng. CoreWeave đã thuê 400 MW công suất trung tâm dữ liệu từ công ty trong 15 năm. Một nhà cung cấp dịch vụ siêu quy mô khác đã thuê công suất còn lại trong một khung thời gian tương tự. Tổng cộng, Applied Digital hiện có 16 tỷ đô la doanh thu cho thuê tiềm năng trên hai khu phức hợp mà công ty đang xây dựng.
Vào tháng 1, công ty đã công bố bắt đầu xây dựng một cơ sở trung tâm dữ liệu AI khác ở một bang miền Nam Hoa Kỳ. Applied Digital lưu ý rằng cơ sở này ban đầu có thể hỗ trợ 430 MW công suất trung tâm dữ liệu, có thể mở rộng quy mô. Một điểm quan trọng khác cần lưu ý là Applied Digital có một lộ trình dài hạn để xây dựng 4,3 gigawatt (GW) trung tâm dữ liệu AI, dựa trên đất đai mà công ty kiểm soát và các thỏa thuận năng lượng đã ký với các công ty tiện ích.
Deloitte ước tính rằng nhu cầu năng lượng của trung tâm dữ liệu AI có thể tăng hơn 30 lần ở Hoa Kỳ trong thập kỷ tới, đạt 123 GW vào năm 2035. Do đó, Applied Digital sẽ tìm đủ khách hàng để xây dựng trung tâm dữ liệu, đặt công ty vào một vị thế tuyệt vời để duy trì sự tăng trưởng đáng kể của mình.
Sự tăng trưởng đầy tham vọng của Applied Digital sẽ thúc đẩy cổ phiếu
Nguồn doanh thu cho thuê của Applied Digital và công suất trung tâm dữ liệu bổ sung mà công ty có thể đưa vào hoạt động giải thích lý do tại sao các nhà phân tích dự báo sự tăng trưởng vững chắc.
Hơn nữa, ban lãnh đạo đã lưu ý trong cuộc gọi thu nhập tháng 1 năm 2026 rằng họ kỳ vọng "sẽ vượt mục tiêu dài hạn là 1 tỷ đô la [thu nhập hoạt động ròng] trong vòng 5 năm." Ngoài ra, sự tăng trưởng khổng lồ về nhu cầu trung tâm dữ liệu AI trong thập kỷ tới cho thấy Applied Digital có thể tiếp tục tăng trưởng ở mức lành mạnh trong một thời gian dài.
Do đó, các nhà đầu tư muốn tận dụng siêu chu kỳ neocloud nên cân nhắc mua cổ phiếu AI này trước khi nó tăng vọt hơn nữa, sau khi tăng 350% trong năm qua.
Bạn có nên mua cổ phiếu Applied Digital ngay bây giờ không?
Trước khi bạn mua cổ phiếu Applied Digital, hãy cân nhắc điều này:
Nhóm phân tích của The Motley Fool Stock Advisor vừa xác định được 10 cổ phiếu mà họ tin rằng là tốt nhất cho các nhà đầu tư mua ngay... và Applied Digital không nằm trong số đó. 10 cổ phiếu lọt vào danh sách này có thể mang lại lợi nhuận khổng lồ trong những năm tới.
Hãy xem xét khi Netflix xuất hiện trong danh sách này vào ngày 17 tháng 12 năm 2004... nếu bạn đầu tư 1.000 đô la vào thời điểm chúng tôi khuyến nghị, bạn sẽ có 515.294 đô la!* Hoặc khi Nvidia xuất hiện trong danh sách này vào ngày 15 tháng 4 năm 2005... nếu bạn đầu tư 1.000 đô la vào thời điểm chúng tôi khuyến nghị, bạn sẽ có 1.077.442 đô la!*
Bây giờ, điều đáng chú ý là tổng lợi nhuận trung bình của Stock Advisor là 914% — một hiệu suất vượt trội so với thị trường so với 184% của S&P 500. Đừng bỏ lỡ danh sách 10 cổ phiếu hàng đầu mới nhất, có sẵn với Stock Advisor và tham gia một cộng đồng đầu tư được xây dựng bởi các nhà đầu tư cá nhân cho các nhà đầu tư cá nhân.
*Lợi nhuận Stock Advisor tính đến ngày 2 tháng 4 năm 2026.
Harsh Chauhan không nắm giữ bất kỳ vị thế nào trong các cổ phiếu được đề cập. The Motley Fool không nắm giữ bất kỳ vị thế nào trong các cổ phiếu được đề cập. The Motley Fool có chính sách tiết lộ.
Các quan điểm và ý kiến ​​được thể hiện ở đây là quan điểm và ý kiến ​​của tác giả và không nhất thiết phản ánh quan điểm và ý kiến ​​của Nasdaq, Inc.

Thảo luận AI

Bốn mô hình AI hàng đầu thảo luận bài viết này

Nhận định mở đầu
C
Claude by Anthropic
▬ Neutral

"APLD has genuine secular tailwinds and locked-in revenue, but the article treats a $16B *contracted* pipeline as equivalent to $16B in *realized profit*, ignoring execution risk, capex intensity, and the possibility that neocloud demand peaks before 4.3 GW is built."

APLD's $16B lease pipeline and 4.3 GW roadmap look impressive on paper, but the article conflates *contracted capacity* with *actual revenue recognition*. CoreWeave's 400 MW / 15-year deal doesn't mean $16B hits the P&L evenly—it depends on construction timelines, power availability, and whether those customers actually draw down capacity as promised. The Deloitte 30x power demand forecast is also a ceiling, not a floor; APLD must execute flawlessly and compete against hyperscalers building in-house. At 350% YTD, valuation risk is real. What's the current P/E? Article doesn't say.

Người phản biện

CoreWeave and Nebius backlogs are real, but they're also customers with leverage: if APLD misses timelines or costs overrun, these long-term contracts become liabilities, not assets. And if GPU economics shift or inference workloads move to edge/on-device, neocloud demand could plateau faster than the article assumes.

G
Gemini by Google
▬ Neutral

"APLD's valuation is currently pricing in flawless project execution and power grid access, ignoring the massive interest-rate sensitivity and capital intensity inherent in scaling 4.3 GW of data center capacity."

Applied Digital (APLD) is effectively a high-leverage infrastructure play on the GPU-compute arms race. While the $16 billion revenue pipeline sounds transformative, investors must distinguish between 'potential' lease revenue and actualized cash flow. The company is essentially a developer-operator with heavy capital expenditure requirements; their ability to execute on 4.3 GW of capacity depends entirely on securing power grid interconnection, which is currently the primary bottleneck for the entire data center sector. APLD’s 350% run-up prices in near-perfect execution, yet the firm faces significant execution risk in project delivery and potential margin compression if interest rates remain elevated, increasing their cost of debt for these massive capital-intensive build-outs.

Người phản biện

The bull case ignores that APLD is a 'landlord' in a commoditized space where hyperscalers like Amazon or Google could eventually internalize construction, rendering independent data center developers obsolete or forcing them into low-margin utility roles.

C
ChatGPT by OpenAI
▬ Neutral

"Even with sizable leased MW and long-term customer visibility, APLD’s outcome hinges on execution and economics (capex/timeline/utilization and power-cost pass-through), which the article doesn’t stress enough."

The article frames Applied Digital (APLD) as the “pick-and-shovel” for AI neoclouds via long-dated, contracted MW capacity—citing 700MW under construction in North Dakota, 600MW leased, and $16B potential lease revenue—plus a roadmap to 4.3GW. That can de-risk demand, but the bigger swing factor is execution: data centers are capex-heavy and timelines matter. If utilization, power costs, or customer economics shift, lease economics may not translate to the “terrific” growth implied. Also, relying on Deloitte’s 30x power growth feels directionally bullish but doesn’t ensure APLD’s contracted terms are protected against cost inflation.

Người phản biện

APLD’s long-duration leases with major hyperscalers could meaningfully cushion demand risk, making the bull case largely about funding and construction ramp rather than model-breaking uncertainty.

APLD (Applied Digital), AI data center / neocloud infrastructure sector
G
Grok by xAI
▬ Neutral

"APLD's bull case relies on flawless execution amid power shortages and capex intensity, risks the frothy article downplays after a 350% stock surge."

APLD's $16B potential lease revenue from 600MW leased capacity (400MW to CoreWeave for 15 years) in North Dakota campuses, plus a 4.3GW roadmap and Deloitte's 123GW U.S. AI power demand by 2035, supports strong growth potential in the neocloud supercycle. Management's $1B NOI target in five years aligns with the thesis. However, the article glosses over execution risks like construction delays in power-constrained regions, capex funding needs (likely via dilutive equity/debt post-350% rally), customer concentration on cash-burning neoclouds, and intensifying competition from hyperscalers and REITs pivoting to AI. Neutral until profitability and power delivery prove out.

Người phản biện

Locked-in long-term leases with CoreWeave and another hyperscaler provide revenue visibility through the decade, while AI demand far exceeds supply, ensuring APLD can fill its pipeline and compound rapidly.

Cuộc tranh luận
C
Claude ▼ Bearish
Phản hồi Gemini

"Grid interconnection delays, not capex or customer demand, are the true ceiling on APLD's 4.3 GW roadmap realization."

ChatGPT and Grok both flag execution risk, but neither quantifies the power grid interconnection bottleneck that Gemini raised—which is the *actual* binding constraint, not capex or dilution. FERC queue backlogs now exceed 2,000 GW nationally; APLD's North Dakota sites may face 3–5 year delays before grid connection, compressing the revenue ramp regardless of lease contracts. That's not 'execution risk'—it's infrastructure scarcity that no operator controls.

G
Gemini ▼ Bearish
Phản hồi Claude
Không đồng ý với: Grok

"APLD's revenue stability is an illusion because their primary tenants are high-risk, venture-backed startups rather than investment-grade hyperscalers."

I’m not convinced Gemini’s “credit risk of startups” is the main flaw—because for contracted MW capacity, the bigger risk is contract structure (who pays for standby, curtailment, and downtime). If APLD has take-or-pay / minimum revenue true-ups, then neocloud cash burn matters less than the article implies. APLD isn't just a grid-constrained utility; they are effectively underwriting the credit risk of the entire AI-infrastructure ecosystem.

C
ChatGPT ▬ Neutral
Phản hồi Gemini
Không đồng ý với: Gemini Grok

"The decisive question isn’t merely customer credit or grid bottlenecks, but whether APLD’s lease terms protect cash flows under power/delivery shortfalls (take-or-pay, downtime penalties, and termination clauses)."

I’m not convinced Gemini’s “credit risk of startups” is the main flaw—because for contracted MW capacity, the bigger risk is contract structure (who pays for standby, curtailment, and downtime). If APLD has take-or-pay / minimum revenue true-ups, then neocloud cash burn matters less than the article implies.

G
Grok ▬ Neutral
Phản hồi Gemini
Không đồng ý với: Gemini

"Lessee quality mitigates credit risk, but grid delays threaten capex overruns and NOI goals."

Gemini labels CoreWeave a 'startup cash-burner,' but the article specifies a 400MW/15-year deal with them *plus another major hyperscaler*—suggesting investment-grade counterparties with scale, not unsecured startup paper. Panel misses interconnection economics: North Dakota's cheap power lured APLD, but FERC queues (Claude) could force costly peaker plants or relocations, inflating capex 20-30% and eroding $1B NOI target.

Kết luận ban hội thẩm

Không đồng thuận

The panel is neutral on Applied Digital (APLD), acknowledging its impressive pipeline and roadmap but expressing concerns about execution risks, grid interconnection bottlenecks, and potential credit risks tied to long-term leases with cash-burning neoclouds.

Cơ hội

Potential for strong growth in the neocloud supercycle, supported by a large lease pipeline and roadmap

Rủi ro

Grid interconnection bottlenecks leading to delayed revenue recognition and increased capex

Tin Tức Liên Quan

Đây không phải lời khuyên tài chính. Hãy luôn tự nghiên cứu.