Invesaro
Investment thesis · Live

The AI data paradox: a storage explosion revives hard drives and hands their makers real pricing power

AI is generating a flood of data that lands on cheap nearline hard drives, while a consolidated two-supplier market gains pricing power and fatter margins.

Published July 10, 2026 · 150-day horizon · supply and demand

Basket return
-15.1%
equal weight, since publication
The market over the same window
+1.0%
benchmark for this basket
Edge over the market
-16.2%
in percentage points
Causal chain: Explosion of AI-generated data → Cold storage economics: flash ~6x pricier per TB, data stays on hard drives → Industry consolidated to two players, capacity never rebuilt after the 2022-2023 trough → Demand for high-capacity nearline drives outstrips supply → Longer lead times, build-to-order production, upward price pressure → HAMR ramp (30TB+ drives) cuts cost per TB and lifts margins → Seagate, Western Digital, TDK: higher margins and cash flows

What this thesis rests on

Each one is a statement that has to be true. When a filing says otherwise, the thesis is in trouble, and this is where we say so.

  1. Seagate reports non-GAAP gross margin above 35% in both quarterly results covering the 150 days after 10 July 2026, with no sequential decline in either quarter.

    financial · Unverified

  2. Western Digital's disclosed nearline/data-center exabytes shipped in the quarter ending December 2026 are at least 10% above its June 2026 quarter level.

    operational · Unverified

  3. Neither Seagate nor Western Digital announces a nearline HDD capacity expansion, and each keeps quarterly capital expenditure below 8% of revenue through 7 December 2026.

    operational · Unverified

  4. Seagate discloses volume shipments of 30TB+ Mozaic HAMR drives and qualification at additional cloud customers in a 10-Q or 8-K filed before 7 December 2026.

    operational · Unverified

  5. Seagate's HDD revenue per exabyte shipped falls by no more than 5% sequentially in each of the two quarters after publication, showing no flash-driven price erosion.

    competitive · Unverified

The market equates artificial intelligence with GPUs, HBM memory and fast flash, and treats hard drives as a sunset technology. That is a mistake. Training and deploying AI models generates enormous volumes of data that have to be stored durably and cheaply: training sets, successive versions of model weights, checkpoints, inference logs, vector databases for RAG and multimodal data (image, video, audio). The vast majority of that data is cold or warm, rarely read, and economically it has to land on the cheapest medium per terabyte, meaning high-capacity nearline drives.

The core of the thesis is cost-per-terabyte economics. Flash remains several times (roughly 5-7x) more expensive per terabyte in the capacity class, and that gap is not closing quickly, because NAND makers are steering limited capacity toward high-margin enterprise SSDs instead. As a result, the marginal petabyte of AI data ends up on spinning platters. Counterintuitively, AI is not killing hard drives, it is expanding their market.

This thesis, over time

What has happened since publication. Every entry is computed as of its own date, from price history.

August 16, 2026

Day 38 of 150: basket -2.6%

The first quarter of the horizon is behind this thesis. The equal-weight basket stands at -2.6% since publication day. The benchmark (S&P 500) returned +2.8% over the same window, so the basket trails its market by 5.4 percentage points. The strongest contributor is STX (+9.4%), the weakest is WDC (-12.0%). 112 days remain until the verdict.

July 17, 2026

Basket fell below -10% since publication

This basket is at -14.5% since publication, 7 days after the thesis was posted. The benchmark (S&P 500) returned -1.6% over the same window, so the basket trails its market by 13.0 percentage points. The strongest contributor is STX (-11.9%), the weakest is WDC (-19.2%). This is not a change of status: the horizon keeps running and the verdict only lands once it expires.

Open the interactive thesis

Interactive causal map, the basket charted against its market, and the full timeline.

Open in Invesaro →

New theses straight to your inbox

Subscribe and we will send you every new thesis and every verdict. No spam, one-click unsubscribe.

This is analysis, not investment advice and not a recommendation to buy or sell anything. We publish it and track it in public, mistakes included. Any decision is yours and yours alone.