Taipei, January 26, 2026, 08:32 GMT+8 — Premarket
Global Unichip Corp (3443.TW) will draw attention Monday as trading begins, after Morgan Stanley bumped its price target to NT$3,288. The upgrade comes on the back of rising demand linked to Google’s in-house chip initiatives, according to Taiwanese media. The stock has already climbed nearly 30% this year, the report added.
Global Unichip’s stock closed at NT$2,725, gaining 5.83% on the day. Earlier in the session, it touched an intraday high of NT$2,790, marking a new record, according to Yahoo Finance data.
The Morgan Stanley report bumped its 2026 shipment forecast for Google’s in-house chips to 1.5 million units, up from 800,000. It also projected the program could generate around $1.2 billion in revenue for Global Unichip that year — more than half the company’s total. According to Nownews, the note highlighted a possible 1–2 percentage point boost in gross margin as foundry orders scale up and Arm-based CPUs gain traction in AI inference and general server markets.
Global Unichip offers application-specific integrated circuit (ASIC) services, delivering custom chips tailored for specific tasks—from design right through to manufacturing support. The company has been ramping up efforts in advanced packaging and high-bandwidth memory-related intellectual property. TSMC calls it a market leader in advanced ASIC services within its value-chain program.
Global Unichip is slated to hold a board meeting on Jan. 29 to approve its Q4 financial report, according to a filing reported by local media.
The company flagged a steep sales surge. Global Unichip reported earlier this month that December net sales soared 75.7% year-on-year to NT$4.748 billion, while fourth-quarter sales more than doubled compared to the same period last year.
Still, the rally carries risks. With the stock priced at a triple-digit earnings multiple, any slip in project deadlines or a weaker order outlook could hit hard.
The bigger wager is that custom silicon will remain in demand as cloud companies ramp up spending on data centres and develop more chips internally. Inference—the stage when a trained AI model handles live data—is crucial since it operates nonstop after deployment, linking demand directly to daily usage instead of occasional training runs.
Traders tracking Global Unichip want to see if the demand translates into steady revenue growth and a more obvious boost in margins. Alternatively, the recent jump might have already priced in too much optimism, especially given the stock’s high valuation.
On Jan. 29, the board is set to approve the fourth-quarter report, offering a fresh look at how much of the Google-related boost is reflected in the figures.
