scripod.com

From the Archive: Can Anyone Catch NVIDIA? | The Future of Chips and Infrastructure

The a16z Show

12 HOURS AGO
The a16z Show

The a16z Show

12 HOURS AGO

Shownote

As part of our summer replay series, we're revisiting one of our favorite conversations on the future of AI infrastructure. SemiAnalysis founder Dylan Patel joins Erin Price-Wright, Guido Appenzeller, and Erik Torenberg to examine the rapidly evolving eco...

Highlights

This podcast revisits a conversation on the future of AI infrastructure, exploring the economics of hardware from GPUs to data centers. The discussion examines the competitive landscape, the challenges for new entrants, and the strategic moves of major tech companies in the race for compute.
15:38
Value capture is broken—companies capture less than 10% of the value they create.
28:57
Challengers need a 5x hardware advantage and hope workloads don't shift.
43:46
Capital costs for GPUs dominate, not power.
48:59
Intel must split design and fab to survive.

Chapters

Is NVIDIA Unbeatable? The 5x Advantage and the Value Capture Problem
00:00
The Startup Struggle: Why Copying NVIDIA Isn't Enough and Custom Chips Are the Real Threat
18:07
Power, Supply Chains, and the Real Cost of Building AI Data Centers
33:32
Strategic Urgency: Intel's Restructuring, Google's Missed Opportunity, and Advice for the Tech Giants
46:23

Transcript

Dylan Patel: NVIDIA is going to have better networking than you. They're going to have better HBM, They're going to have better process node. They're going to come to market faster. They're going to be able to ramp faster. They're going to have better nego...