Monday, May 5, 2025

CoreWeave's Valuation in High-Performance Computing as a Service

Times of high volatility in equity markets tend to be even more volatile where it comes to assets in new or emerging fields where demand is hard to forecast. So it is with high-performance computing “as a service,” the business CoreWeave and others are in. 


So it might not come as a complete surprise that some observers are skeptical about CoreWeave’s valuation, at the very least. 


To evaluate, we would need to estimate the global market; the U.S. share of that market (CoreWeave will operate internationally as well as in the U.S. market); CoreWeave’s  enterprise value (debt plus equity) and then CoreWeave’s estimated revenue. 


So assume global 2023 HPCaaS revenues of about $36 billion; North America share is 40 percent and U.S. share of North America is 85 percent. That gives us a 2023 U.S. market of about $12.4 billion. 


U.S. High-Performance Computing "as a Service" Market

Year

Low Scenario (8% CAGR)

Baseline Scenario (13.4% CAGR)

High Scenario (18% CAGR)

2023

$12.42 billion

$12.42 billion

$12.42 billion

2024

$13.41 billion

$14.08 billion

$14.66 billion

2025

$14.49 billion

$15.97 billion

$17.30 billion

2026

$15.65 billion

$18.11 billion

$20.41 billion

2027

$16.90 billion

$20.54 billion

$24.09 billion

2028

$18.25 billion

$23.29 billion

$28.42 billion

2029

$19.71 billion

$26.41 billion

$33.54 billion

2030

$21.29 billion

$29.95 billion

$39.58 billion


So that establishes some expectations for the market opportunity CoreWeave and others are chasing. 


An immediate question is whether CoreWeave can hope to make a profit, given its debt and equity structure, both of which place claims on some of CoreWeave’s cash flow.


The company took on $7.5 billion in debt facility led by Blackstone and Magnetar in May 2024, with participation from Coatue, Carlyle, CDPQ, DigitalBridge Credit, BlackRock, Eldridge Industries, and Great Elm Capital Corp., following an earlier $2.3 billion debt facility led by Magnetar Capital and Blackstone in August 2023.


A $650 million credit facility was added in October 2024. So total debt is $10.45 billion.


Equity issued includes a $1.1 billion Series C round in May 2024, valuing CoreWeave at $19 billion.


A $642 million Series B round was completed in December 2023 with another $420 million Series B round in April 2023.


According to Crunchbase, CoreWeave raised $2.37 billion in equity across 13 rounds, including seed, Early-Stage, and Late-Stage rounds).


So assume $2.37 billion as the total equity raised. Add the $10.45 billion in debt and total investment  is $12.82 billion. 


Enterprise value includes the firm’s market capitalization plus debt plus cash. If CoreWeave’s market cap is about $18 billion, then enterprise value is about $28.55 billion. 


For the current EV of $28.55 billion and revenues of $1.92 billion, CoreWeave has an implied revenue multiple of 14.87 times. That might be considered conservative for a fast-growing technology firm.


At $1.9 billion in a U.S. market that might have generated $14 billion in annual revenues, CoreWeave might already have as much as 13,5 percent share in the high-performance computing as a service market. 


Can AI Replace Cloud Computing?

I saw an odd headline asking “when will AI replace cloud computing?” It’s a non-sequitur. Sort of like asking when a layer 7 function is replaced by a layer 1 function. In other words, AI exists at a different layer of the software stack.


OSI Model

AI/Cloud Layer

7. Application (user-facing apps)

AI Model Deployment & APIs4

6. Presentation (data formatting/encryption)

Data Preprocessing/Feature Engineering4

5. Session (connection management)

Cloud API Gateway Management2

4. Transport (data integrity)

Distributed Training Orchestration4

3. Network (routing)

Cloud Network Virtualization2

2. Data Link (node-to-node transfer)

GPU Cluster Interconnects2

1. Physical (hardware)

Cloud Infrastructure (GPUs/TPUs)


Saying AI replaces cloud computing is like saying "video games replace electricity." Cloud computing is infrastructure, supplying functions lower in the stack.


Sunday, May 4, 2025

2 Customers Account for 77% of CoreWeave Revenue

Financial analysts typically express concern when any firm’s customer base is too concentrated. Consider that, In 2024, CoreWeave’s top two customers (Microsoft and “Customer C”) accounted for 77 percent of revenue. 


Based on available information, CoreWeave’s largest customers by revenue contribution are primarily Microsoft, with other notable customers including OpenAI, Meta, NVIDIA, IBM, and Cohere. 


Customer

Revenue Contribution (2024)

Importance

Microsoft

62%

Largest customer in 2023 (35%) and 2024. Significant reliance noted in S-1 filing.

Customer C

15%

Unnamed second-largest customer, contributing to 77% of 2024 revenue with Microsoft.

OpenAI

Not specified (future revenue)

Signed a $11.9B five-year contract in March 2025, expected to reduce Microsoft’s share to <50%.

Meta

<10%

Confirmed customer, but no individual revenue share above 10% in 2024.

NVIDIA

Not specified

Customer and investor; $250M order during IPO, but no specific revenue share.

IBM

Not specified

Using CoreWeave for Granite AI training; no specific revenue share.

Cohere

Not specified

AI startup customer; no specific revenue share reported.


Saturday, May 3, 2025

Why Kuiper?

 Amazon has launched into orbit the first 27 of a planned Kuiper constellation of 3,236 low-earth-orbit satellites. Some will liken the effort to Google Fiber, in the sense of a major app provider getting into the access business. And obviously it is a challenge to Elon Musk’s SpaceX. 

But some might also wonder whether the initiative consumes capital that might be better spent elsewhere. 

Amazon itself, and others who think the plan has value, might point to the upside. For starters, 

Kuiper can help AWS offer low-latency services even in remote areas, as well as providing redundant or backup links for AWS data centers, especially in underserved regions.

Kuiper also can provide AWS clients space-based connectivity for internet of things apps, for example. 

Kuiper enhances Amazon’s logistics network by providing coverage in rural or hard-to-reach areas, improving route optimization, tracking, and communications.

As other app hyperscalers have found, without internet access, potential users cannot be served. So Amazon could reach new customers in remote or underserved areas using Kuiper, supporting content services (Amazon Prime Video), e-commerce and other services and apps. 

And once access is provided directly by Amazon, device and other bundles can be offered that might include Alexa, for example. 

Also, to some extent, Kuiper means Amazon can vertically integrate its stack, providing internet access and applications, reducing its value chain dependence on partners. 

Friday, May 2, 2025

The Optimal Shopping Problem and Online Dating

The “Secretary Problem,”  also known as the “Optimal Stopping Problem,” is an example of decision theory. It deals with the challenge of making the best choice (like hiring the best candidate or picking a life partner,) when options must be evaluated sequentially and decisions are irreversible.


Some of us would say it also applies to online dating. One should skip all of the first 37 percent of profiles. Then one picks the first candidate afterward that tops the best "score" of the first 37 percent of profiles.


Even then, the odds of getting a second date are said to range around 20 percent. It's a tough business!


The classic example is hiring a candidate for a secretary position. Interviews are sequential, and after each interview, you must either hire that candidate immediately or reject them permanently.


Your goal is to maximize the probability of hiring the best (highest-ranked) candidate overall.


The theorem suggests that the hiring team literally skip the first 37 percent of candidates without hiring anyone, while scoring them. So for a universe of 1,000 candidates, one skips the first 370, before picking the very next candidate that "scores" better than the top candidate of the prior 370!


Then, the team “should” hire the first candidate thereafter who is better than all the ones you've seen.



Mathematically, this strategy maximizes the probability of picking the very best candidate, and this success probability approaches 1/e (~37%) as n → ∞.


Applications might include online dating, choosing a life partner (the "marriage problem"), real estate decisions, investment decisions or possibly even parking space selection.


Apple Services Revenue Led by Google Search Default

Apple services revenue reached an all-time high of $26.6 billion, with double-digit growth in both developed and emerging markets, according to the firm’s latest quarterly financial report. By way of comparison, iPhone revenue was just shy of $47 billion.


Overall, services revenue is about 25 percent of Apple total revenue, while iPhone sales are about half of Apple total revenue. 


Estimated Apple Services Revenue Sources (FY 2024)

Source

Estimated Revenue ($B)

Percentage of Services Revenue

Google Search Deal

20–26.3

20–27%

App Store

22

22%

Apple Music

10

10%

iCloud

10

10%

Apple Pay

3

3%

AppleCare

2.5

2.50%

Apple TV, Other Subscriptions

3

3%


Most observers would agree that services revenue is important for Apple to smooth out revenue trends, as subscription revenue is predictable and recurring, compared to one-time hardware purchases.  


Services also tend to have significantly-higher profit margins. While iPhones have healthy gross margins (typically 30 percent to 40 percent), services can exceed 70 percent margins. Services revenue growth rates also are higher than device revenue, while services create more ecosystem lock-in for Apple as well. 


Thursday, May 1, 2025

Meta Quantifies AI Monetization

Though we might still note that artificial intelligence benefits often are somewhat indirect or hard to measure, Meta joins the ranks of firms pointing to financial benefits from AI. 


AI is transforming everything we do,” said Meta CEO Mark Zuckerberg. Specifically, he sees “improved advertising, more engaging experiences, business engaging experiences, business messaging, Meta AI and AI devices” as opportunities.


“In just the last quarter, we are testing a new ads recommendation model for Reels, which has already increased conversion rates by five percent, and we're seeing 30 percent more advertisers are using AI creative tools in the last quarter as well,” he noted. 


“In the last six months, improvements to our recommendation systems have led to a seven percent increase in time spent on Facebook, a six percent increase on Instagram, and 35 percent on Threads,” Zuckerberg added. 


So even if the impact is somewhat indirect, Meta's advertising revenue, which constitutes the vast majority of its total revenue, continues to benefit significantly from AI-driven enhancements. In the first quarter of 2025, Meta reported revenue of $42.31 billion, a 16 percent year-over-year increase, with advertising revenue estimated at around $40.57 billion. 


That does not necessarily mean all the upside was because of AI, but AI-powered tools, such as the Advantage+ marketing service, have improved ad targeting and relevance, leading to a 70 percent growth in Advantage+ adoption. That might be said to have driven higher ad impressions and an 11 percent increase in average price per ad. 


Likewise, AI-driven content recommendation systems have increased user retention and engagement, with an eight-percent increase on Facebook and a six-percent growth  on Instagram. 


While Meta AI is not yet a significant revenue driver, Zuckerberg has talked about plans to monetize it in the future, potentially through ads or a premium subscription model. 


Meta’s commentary complements that of Alphabet, which in its latest quarterly financial report talked about AI  monetization as well.

D-Day Plus 81

Some things should not be forgotten.  D-Day, 81 years ago  is among them.