The AI buildout is accelerating in 2026, and we see it is getting difficult to track what is being deployed, by whom, for what usage, with what financing, and on what underlying business model. As GTC approaches, we address all these questions in a series of one-sliders. Today we look at how rising AI infrastructure capex translates into sustained cloud revenues growth, and who will consume how much cloud services by the end of the decade.
The AI buildout is accelerating in 2026, and we see it is getting difficult to track what is being deployed, by whom, for what usage, with what financing, and on what underlying business model. As GTC approaches, we address all these questions in a series of one-sliders. Today, we look at what returns hyperscalers generate on AI investments today and how these could improve over time once the "tail effect" kicks in.
The AI buildout is accelerating in 2026, and we see it is getting difficult to track what is being deployed, by whom, for what usage, with what financing, and on what underlying business model. As GTC approaches, we address all these questions in a series of one-sliders. Today, we look at how the "tail effect" boosts ROIC for AI infrastructure and traditional compute deployments.
The AI buildout is accelerating in 2026, and we see it is getting difficult to track what is being deployed, by whom, for what usage, with what financing, and on what underlying business model. As GTC approaches, we address all these questions in a series of one-sliders. Today, we look at what we call the tail effect: How ROIC underestimates the actual return on AI infrastructure over the full useful life of chips.
The AI buildout is accelerating in 2026, and we see it is getting difficult to track what is being deployed, by whom, for what usage, with what financing, and on what underlying business model. As GTC approaches, we address all these questions in a series of one-sliders. Today, we compare the returns of AI cloud and frontier-model businesses with those of traditional cloud and core hyperscalers.
The AI buildout is accelerating in 2026, and we see it is getting difficult to track what is being deployed, by whom, for what usage, with what financing, and on what underlying business model. As GTC approaches, we address all these questions in a series of one-sliders. Today we look into the implications of TSMC capex for XPU spending over the next few years.
The Citrini report is an interesting thought experiment. It qualifies fully as neo-marxist, calling for the tendency of profit rates to fall. Like Marx 165 years ago, the authors are nevertheless getting a couple of things wrong. Click the link below for our thoughts and our perspective on the matter.
Today, we are publishing the Foundry section of our 29th Tech Infrastructure Quarterly Bible. The Tech Bible is a must-read for any tech investor, as it summarizes the quarterly earnings reports from the over 140 companies we track, providing an update on our key perspectives and convictions. TSMC increased its 2025 growth outlook to 35%, as we anticipated. For other foundries, expectations were revised down 1pt to 9% growth in 2025, which implies a reasonable 2% sequential growth in 4Q. For ne...
We publish today our comprehensive quarterly bible: 242 pages of detailed analysis on what happened in the last 3 months, and how we interpret it, in light of our current convictions. The first section acts as a PM summary, outlining our key findings, and latest thoughts on the semi cycle, in 6 slides:
We hosted another session of the 5th annual BCG × NSR Gen-AI Big Idea Series this week. We will host a slate of exceptional speakers through Feb-26 (full line-up here). This session was a conversation with Hock Tan, CEO of Broadcom, along with Kirsten Spears, CFO, and Charlie Kawwas, President of the Semiconductor Solutions Group. Please see the link below for our takes.
Unfortunately, this report is not available for the investor type or country you selected.
Report is subscription only.
Thank you, your report is ready.
Thank you, your report is ready.