GigaVis Follow-Up: Foreign and Institutional Money Tests the AI FC-BGA Yield Bottleneck

GigaVis is not an AI chip maker or a substrate producer. It is an inspection, review, repair, automation and software equipment supplier for high-end FC-BGA / ABF substrates used in AI GPUs and ASICs. Its 2026 disclosed orders total KRW 29.41B, or 56.1% of FY2025 revenue, while the FY2025 mix was AOI/VRS/AOR-heavy with FA and software attached.

Recommendations 0

AI Data Center CapEx at $5.3 Trillion: Korea Should Look Beyond GPUs to Power, Substrate and Storage Bottlenecks

Goldman lifted FY2025-2030 hyperscaler capex expectations to $5.3T, while Alphabet guided 2026 capex to $180-190B and a further increase in 2027. For Korean equities, the better read-through is not broad AI beta but power equipment, Samsung Electro-Mechanics, high-layer PCBs, optical links and eSSD/storage bottlenecks.

Broadcom Reconfirmed a $100B+ AI Semiconductor Run-Rate for 2027: How Korea Should Read the Value Chain

Broadcom's Q2 FY2026 result was not an AI-demand slowdown. It was an expectation reset. Q2 AI semiconductor revenue was $10.8B, Q3 guidance is $16.0B, and the FY2027 $100B+ AI semiconductor frame remains intact. For Korea, the read-through is HBM, FC-BGA, silicon capacitors, AI networking PCBs, packaging equipment and test.

Marvell Follow-Up: Why EML and CW-DFB Laser Sources Matter for AI Data Centers

A follow-up to the Marvell / Broadcom connectivity thesis: AI optical interconnect bottlenecks are moving deeper into EML and CW-DFB laser sources. Korea's most direct proxy is OE Solutions, but customer qualification and repeat orders matter more than product headlines.