JCET Plans ~10 Billion Yuan Capex in 2026, Betting Big on AI Advanced Packaging
N.R. Finch
JCET (长电科技) plans to spend roughly RMB 10 billion (~$1.4 billion) in 2026 on advanced packaging and automotive chips — one of the largest single-year capex commitments in its history — staking near-term margins on the bet that AI packaging demand will fill those lines fast enough.
Where does the RMB 10 billion go?
The largest share targets advanced packaging: JCET is building 2.5D packaging capacity — a process that lays multiple chips side by side on a single substrate — while developing 3D wafer-level and 3D system-level packaging.
Automotive electronics is the second priority; production ramp is tracking to plan.
This means → JCET is concentrating its bet on two high-conviction growth lanes — AI compute packaging and smart-vehicle chips — while legacy packaging gets only a partial budget.
Why are Chinese fabs busy in a slow season?
Despite the industry's traditional low season, JCET's China factories are running at full order load, with new capacity already coming online.
Overseas, a business restructuring launched in 2025 is starting to pay off: new clients and orders are coming in, and telecom demand is recovering.
In plain terms = domestic orders hold the base, the overseas overhaul is adding incremental volume, and both lines are running at once.
Where does the pricing power come from?
In 2025 JCET introduced a new pricing mechanism that passes raw-material cost increases directly into quotes; more customers have now accepted it.
Some capacity is tight, and the company is screening clients and orders, tilting line resources toward higher-value products.
This means → JCET is shifting from "take every order" to "pick the order," aiming to lift average selling price rather than chase volume alone.
Will margins take a hit during the buildout?
The company is blunt: heavy investment will pressure utilization rates and profitability until customer qualifications and volume orders catch up.
JCET says it is willing to absorb that near-term cost, provided its balance sheet stays sound.
This reflects the central test — whether JCET can push new lines to mass production and fill them with qualified orders fast enough within the AI and automotive demand window, or risk turning that capex into sunk cost.
Content is for reference only, not financial advice.