Tin supply will not catch up with semiconductor demand before 2029, according to a Crux Investor analysis that warns of a mounting solder squeeze for advanced chip packaging.

The investor research note frames a structural deficit in refined tin. Stocks are draining. New mine output remains years away. Chipmakers that depend on tin-based solders for bumping, interconnects, and three-dimensional packaging face a tighter market than the copper, cobalt, or nickel watchers acknowledge. The report does not name individual buyers, contract values, or specific mine projects.

The hardest fact sits in the timeline. No major new tin supply will arrive before 2029. The wait stretches longer than most foundry roadmaps, longer than the typical high-bandwidth memory capacity expansion cycle, and well past the 2026 inflection point many analysts expect for AI accelerator volumes.

What is driving tin demand from chips?

Semiconductor consumption of tin has risen quietly alongside advanced packaging. Solder balls on flip-chip BGAs, microbumps for three-dimensional stacked memory, and lead-free tin-silver-copper alloys for HBM all draw on refined tin. The Crux Investor analysis argues the cumulative solder surface area per package keeps climbing as chiplet designs multiply die-to-die links.

Each HBM4 stack reportedly uses hundreds of additional solder joints compared with HBM3. The report does not publish a precise per-stack figure. Readers should treat vendor estimates with caution until independent teardowns confirm them.

Why does tin supply lag so badly?

Tin mining concentrates in five countries. China, Indonesia, Myanmar, Bolivia, and Peru account for the bulk of global output. Permitting cycles for new tin mines routinely exceed a decade. The Crux Investor note observes that no greenfield project on the public development pipeline will reach first metal before 2029.

Recycling adds some tonnage. Secondary tin from solder recovery and e-scrap will grow. The report does not quantify how much. That omission matters for any serious deficit calculation. The note also does not name the greenfield projects it considers.

How exposed are chipmakers and OSATs?

The risk falls hardest on assembly and test subcontractors. Outsourced semiconductor assembly and test providers and the advanced packaging units of foundries consume the most solder per wafer-out. Integrated device manufacturers with captive back-end capacity carry similar exposure. Fabless companies sit further from the metal but inherit the cost through packaging quotes.

The note stops short of estimating how much tin one 300 mm wafer of HBM base die requires. Solder volume scales with bump count, not wafer count. That distinction often escapes lay summaries, and Crux Investor does not correct it.

What is the report not telling readers?

The headline carries weight. The publicly available summary, however, is thin. Crux Investor offers no price forecast, no London Metal Exchange warehouse stock figure, no list of delayed mines, and no estimate of how many months of consumption current inventories cover. The reader must take the 2029 cliff on faith.

A skeptical investigator asks: what counts as new supply? Restarted idled capacity counts. Expanded tailings reprocessing counts. Each can add tonnage inside three years if prices clear operating costs. The note does not address restart economics or substitution risk in SAC alloys.

What happens next?

• Tin prices on the London Metal Exchange have already moved higher in recent sessions, per public exchange data.
• Whether the move sustains depends on whether exchange-warehouse stocks fall further.
• Chinese export controls on refined tin are the second variable to watch.
• Neither variable appears in the Crux Investor summary, a notable gap for a supply-claim report.
• Chipmakers with long-dated solder contracts will feel the squeeze last.
• Spot buyers will feel it first.
• The 2029 horizon gives procurement teams time to lock supply or substitute, but only if they start now.

— Sophie Lindqvist