RESEARCH
Understanding the infrastructure
behind intelligence.
My research starts with memory: how it is built, where supply gets constrained, and how technological change reshapes the economics of AI.
MemoryCapex is where I develop and publish this work.
MY PRIMARY RESEARCH PROJECT
MemoryCapex.
Memory, technology
& the capital cycle.
A dedicated research platform exploring DRAM, HBM and NAND, and their role in computing infrastructure. I connect technology roadmaps with supply dynamics, industry economics and capital allocation.
THE RESEARCH LENS
From physical constraints
to economic consequences.
Selected research.
Browse researchA starting point for exploring the questions behind my work.
01Memory as a capital-allocation constraint
How HBM, DRAM and NAND are rewriting the economics of AI data centers.
Beyond HBM: the memory wall
How evolving architectures could reshape the memory cycle through 2029.
The HBM enablers
What Marvell reveals about the next layer of the memory cycle.
HOW I WORK
Evidence before conviction.
Start with the evidence.
Company disclosures, technical publications and industry data form the starting point. The source stays visible alongside the argument.
Explore sourcesMake the argument testable.
State the thesis, separate observations from assumptions, and identify the evidence that would change the conclusion.
Keep the view open.
Follow new developments, revisit the thesis and update the interpretation as the evidence evolves.
Read the JournalONGOING EXPLORATION
MFI RESEARCH IN PROGRESS
An ongoing research project investigating the memory cycle through a more systematic framework.
About this project
MFI remains a research project. Its measurement framework and testing are being developed; it is not presented here as a validated forecasting tool.