Author: Ada, Deep Tide TechFlow

On April 23, Texas Instruments' stock price achieved its best single-day performance since 2000. On the same day, ServiceNow recorded its largest single-day drop in history.

In the same earnings season, on the same trading day, two contrasting signals. The market is drawing a line with real money, and below this dividing line, AI infrastructure wins, while AI upper applications lose.

The chip makers are laughing, the subscription sellers are crying

On Thursday, Texas Instruments delivered an almost flawless report card. Q1 revenue was $4.83 billion, a year-on-year increase of 19%, with an EPS of $1.68, far exceeding the market expectation of $1.40. Revenue from data center increased by a staggering 90% year-on-year. The industrial and analog chip business fully recovered.

On that day, the stock price rose by 18%. Bank of America directly upgraded its rating from neutral to buy, raising the target price from $235 to $320.

The expectation for the second quarter is revenue of $5 billion to $5.4 billion, with the median exceeding Wall Street expectations by more than 10%. Management stated that the recovery in industrial and data centers is "continuing to accelerate."

After hours, Intel dropped another bombshell. Q1 revenue was $13.58 billion, far exceeding the expected $12.42 billion. Non-GAAP EPS was $0.29, while the market expectation was $0.01, a 29-fold increase over expectations. Revenue from data center business grew by 22%, reaching $5.1 billion.

Intel's stock price rose over 20% after hours, surpassing the historical high from the internet bubble period in 2000. Last year, the U.S. government acquired a 10% stake, and this year's stock price has already risen by more than 80%.

The frenzy in the chip industry has a fundamental logic. Because AI is not air; it requires electricity, chips, and occupies data centers. From Nvidia's GPUs to Texas Instruments' analog chips, and Intel's CPUs and advanced packaging, the entire food chain of AI infrastructure is being called to "get on board."

The semiconductor ETF (SMH) has risen nearly 28% this year, with a 22% increase in April alone. The S&P 500 has risen 4% during the same period.

On the flip side, the software sector is undergoing a massacre.

ServiceNow plunged 18% that day, its worst single-day performance ever. IBM fell nearly 10%. Then the contagion quickly spread: Salesforce, Workday, Oracle, Adobe, and Palantir all saw declines. iShares Expanded Technology Software ETF (IGV) dropped nearly 5% that day.

The most ironic part is that IBM and ServiceNow's financial reports themselves were not bad. IBM's revenue exceeded expectations, and ServiceNow also surpassed them. But the market does not care. The market is pricing a deeper fear: your moat is being eroded by AI.

The end of SaaS

This is not a one-day affair.

In recent months, a new phrase has circulated in technology, venture capital, and public markets: "SaaSpocalypse," the end of SaaS. Since February, the collapse of software stocks has not stopped. So far, about $2 trillion of enterprise software market value has evaporated.

Salesforce has dropped over 30% this year. Workday is down 33%. Adobe fell 27%. Even Microsoft is down 16%. The software ETF (IGV) has plunged from a historical high of $117 to around $82, entering a technical bear market. The forward P/E of the software sector has fallen below the overall level of the S&P 500 for the first time since the mid-2010s.

Why?

The core logic can be summarized in one sentence: AI allows enterprises to do it themselves.

The traditional SaaS business model charges per seat; if you have 100 employees using my software, you buy 100 licenses. However, with the arrival of AI agents, one agent can replace the work of 10 employees. When the number of seats decreases, the subscription fee also decreases.

More critically, some companies have started to build internal tools using AI directly, bypassing the SaaS middle layer. Previously, buying a CRM system would cost $300 per user per month, but now having AI write an internal system might cost only a fraction of that.

In other words, the moat of SaaS was previously seen as the near-term costs of migration and user stickiness. But AI has short-circuited both of these aspects. Migration costs are lower because AI can automatically handle data migration; stickiness is lower because users no longer need to learn new tools.

The shift in narrative

Look at two sets of data.

This year so far, the semiconductor index has risen by about 40%. The software index has fallen by over 13%. The gap between the two has exceeded 50 percentage points.

What does this mean? Capital has not left the tech industry. It has merely rotated precisely within the tech sector: from the application layer to the infrastructure layer.

The logic now is simple. If I want to bet on AI, I buy chips, because no matter which company's AI wins, chips must be used. But I may not necessarily buy SaaS.

This is the cruel aspect of the market. Chips are a bet with certainty; regardless of how AI develops, the demand for computing power only increases. But software is a conditional bet: it only has value if AI cannot fully replace software and software companies can successfully transform.

Both conditions are uncertain, and capital hates uncertainty.

However, attributing all of ServiceNow and IBM's declines to AI threats is not entirely fair.

ServiceNow's CFO, Gina Mastantuono, mentioned a very specific reason during the earnings call: the Middle East conflict caused order delays. Clients from the Iran direction pulled new orders, which directly dragged down the subscription revenue for the season.

IBM's issues also had specific explanations. Software business growth slowed from 14% in the previous quarter to 11.3%, primarily due to the drag from Red Hat's cloud business. Overall revenue growth decelerated from 12.2% to 9%. Subsequently, IBM maintained its full-year guidance without an upgrade.

But the market doesn't care about these details.

In an environment where everyone is worried that "software is dying," any less-than-perfect earnings report will be interpreted as "see, it's starting." Once this sentiment takes hold, data becomes irrelevant. The narrative becomes important.

And the current narrative is: AI is the hunter at the top of the food chain, SaaS is the prey at the bottom of the food chain.

Behind the surge

Behind Texas Instruments' 18% surge is a less attractive number: a P/E ratio exceeding 50 times. In the past three months, insiders have sold $26.5 million worth of stock and have not bought a single share.

Meanwhile, Intel's forward P/E is 120 times, more than four times that of the S&P 500. One valuation agency gave an intrinsic value of $27, while the stock price is around $67. A 147% premium.

What can be seen is that current chip stock prices have already discounted growth for the next three years. Buyers are not purchasing this quarter’s performance but are instead buying a belief in the unending capital expenditure on AI.

This year, the combined AI capital expenditure of the four major tech giants exceeds $500 billion. Google alone plans to burn $180 billion. As long as this capital expenditure cycle does not stop, chip stocks will have support. But what if one day the giants suddenly realize that the returns on their spending are not high enough?

Don’t forget, the last time Alphabet announced a $180 billion capital expenditure, its stock price plummeted 6% in after-hours trading. The market's reaction was: “We know you are building AI, but we are starting to worry about whether you can make it work.”

Looking at it from a broader perspective, the financial report divergence on April 23 reveals a larger structural shift.

The value capture of AI is migrating downward. From subscription fees at the software level, it is penetrating down to hardware level fees for chips, energy, and data centers. The profit distribution map of the entire tech industry is being redrawn.

When the stock price of chip makers rises to records not seen since 2000, while the subscription sellers fall to new lows for the year, the market is essentially saying one thing: I know AI is real, so I want to buy infrastructure. But is AI useful? I'm still not sure, so I'm not buying applications.

How long will this split last? No one knows. But one point of reference is the next round of earnings season. If the giants’ AI capital spending continues to increase, the cash flow of chips can continue to support valuations. But if one giant suddenly hits the brakes, this dividing line will reverse.

Before that, the chip's celebration continues, while SaaS’s funeral also continues.