Tape still isn’t dead, but shipments slipped by 16 exabytes in 2025

LTO tape shipments fell to 160.3 exabytes in 2025, down 9% from 176.5 exabytes in 2024, with the LTO Program citing an evolving international trade environment and more cautious customer purchasing. Despite this slip, Q1 2026 shipments are 57% higher year-over-year and the group expects 2026 to set another record, supported by the debut of a new 40TB cartridge compatible with 30TB-capable drives. The article also flags data center power constraints as a tailwind for tape’s low-power profile, though hyperscaler-led storage alternatives and component supply tightness remain risks.

Analysis

This reads more like a procurement pause than a demand break. The market should focus on who actually monetizes consumables: the real economics sit with cartridge supply, where concentration creates more pricing leverage than the alliance members themselves. HPE and IBM get only indirect ecosystem benefit unless tape helps preserve storage attach or support services mix; the P&L sensitivity is likely too small to move estimates.

The near-term catalyst is the new 40TB media cycle, but the second-order effect is on archival-storage competition. If hyperscalers stay constrained by power and AI-driven HDD shortages persist, tape keeps a niche advantage on watts-per-terabyte and could take share in cold storage. If power constraints ease, the relative case weakens quickly because retrieval latency still limits adoption; that makes this a 1-3 month operating data story, not a secular re-rate for the platform names.

Contrarianly, consensus may be underestimating supply concentration and overestimating shipment volatility as a directional signal. With only a couple of cartridge manufacturers, upside could show up first in pricing rather than units, while the downside is that any bottleneck can make growth look choppy even when demand is fine. The key falsifier is not the alliance’s commentary but whether Q2/Q3 actual shipments and cartridge availability sustain the Q1 rebound.

Key Decisions for Investors

• Stay flat HPE and IBM on this tape read-through; the signal is too small to justify a position change. Revisit only if the next storage-segment print shows a measurable attach-rate or margin benefit, or if guidance moves by >2-3%.
• Small speculative long FUJIY on pullbacks, 3-6 month horizon, as the cleaner consumables exposure if 40TB adoption tightens cartridge supply. Risk/reward is modestly favorable, but cut the trade if Q2/Q3 shipment growth slips back below ~10% YoY.
• Use this as an alert for WDC/STX rather than a trade: if HDD lead times normalize, tape loses a relative tailwind; if shortages persist, archive substitution stays supportive for tape. Do not short HDD names purely on this headline.
• If you need a relative-value expression, a small long FUJIY / short HPE pair is the least bad way to isolate consumables vs. platform noise, but size it tightly because correlation is weak and the short leg is mostly a hedge against multiple compression, not earnings impact.