Largan's Lin: Developing CPO to Avoid 'Elimination by AI,' FA Pilot Production by September
Largan Precision Chairman Lin En-ping announced on June 9th that the company is developing CPO (Co-Packaged Optics) technology to counter the risk of the lens industry being rendered obsolete by AI. He stated that a pilot automated production line for FA (Fiber Array) is planned for before September 2026, with a large potential buyer already set to visit. Lin noted that Largan's precision has reached below 0.3 microns, surpassing industry standards, and gives them an advantage in stacking more than 4 layers, adding they can now "bravely ramp up mass production" of FA products.
📋 Article Processing Timeline
- 📰 Published: June 9, 2026 at 12:04
- 🔍 Collected: June 9, 2026 at 12:17 (13 min after Published)
- 🤖 AI Analyzed: June 9, 2026 at 12:19 (2 min after Collected)
(CNA, Taichung, 9th) Regarding progress on CPO, Largan Precision Chairman Lin En-ping stated that there are plans for a first automated pilot production line by September, and potential buyers will be invited for a tour. He estimates the time to small-scale mass production will be about 6 months to a year. He frankly admitted that developing CPO "is to avoid being eliminated by AI," and that they can now "bravely ramp up mass production" for FA products.
Largan held its shareholders' meeting in Taichung today, chaired by Lin En-ping. In a media interview afterward, he addressed the market's focus on CPO (Co-Packaged Optics) progress, stating a plan to launch the first automated pilot line by September 2026, with one large potential buyer already identified.
Lin stated the pilot line's products will primarily be FA (fiber array), which can be coupled with MLA (micro lens array) to form an FAU (fiber array unit). The main approach is to achieve overall precision goals through intelligent process control and compensation, rather than relying on high-precision V-grooves. Largan's business model allows for supplying both FA and MLA, depending on customer specifications and the final cost structure.
He further added that a large potential buyer will visit the production line, and initial customer specifications are all for single-layer designs. "Single-layer will still dominate next year and the year after," he said, with multi-layer demand likely to emerge in 3 to 4 years. "Then it will move into 4-layer, 8-layer."
Regarding the current precision status, Lin said the best external precision is currently 0.5 to 0.8 microns; Largan's internal tests have already achieved below 0.3 microns, with high precision and stacking of 4 or more layers being Largan's advantages.
In response to media reports that the CPO parts production line would require 40,000 people, Lin refuted this, stating that a "human wave" strategy is impossible in Taiwan and automation is necessary for survival. On the topic of CPO progress being faster than expected, he said that during the April investors' conference, he felt a bit "uneasy," but seeing the progress since has made him more confident.
Lin revealed that when he was in the lens industry, the rise of AI created "a lot of pressure" because AI development would increasingly reduce the need for lens upgrades. "It felt like the lens industry was about to be eliminated, which sparks the motivation to take another path." He added, "I was thinking about how not to be eliminated by AI and join it, which is why we developed the CPO project." He now considers it to have preliminary results, though the industry currently still faces a "shortage of land and talent."
Looking ahead, he said that Largan's core business is lenses, and now FA can be "bravely ramped up for mass production," while SiC and electronic materials are not yet at a stage for mass production.
Largan held its shareholders' meeting in Taichung today, chaired by Lin En-ping. In a media interview afterward, he addressed the market's focus on CPO (Co-Packaged Optics) progress, stating a plan to launch the first automated pilot line by September 2026, with one large potential buyer already identified.
Lin stated the pilot line's products will primarily be FA (fiber array), which can be coupled with MLA (micro lens array) to form an FAU (fiber array unit). The main approach is to achieve overall precision goals through intelligent process control and compensation, rather than relying on high-precision V-grooves. Largan's business model allows for supplying both FA and MLA, depending on customer specifications and the final cost structure.
He further added that a large potential buyer will visit the production line, and initial customer specifications are all for single-layer designs. "Single-layer will still dominate next year and the year after," he said, with multi-layer demand likely to emerge in 3 to 4 years. "Then it will move into 4-layer, 8-layer."
Regarding the current precision status, Lin said the best external precision is currently 0.5 to 0.8 microns; Largan's internal tests have already achieved below 0.3 microns, with high precision and stacking of 4 or more layers being Largan's advantages.
In response to media reports that the CPO parts production line would require 40,000 people, Lin refuted this, stating that a "human wave" strategy is impossible in Taiwan and automation is necessary for survival. On the topic of CPO progress being faster than expected, he said that during the April investors' conference, he felt a bit "uneasy," but seeing the progress since has made him more confident.
Lin revealed that when he was in the lens industry, the rise of AI created "a lot of pressure" because AI development would increasingly reduce the need for lens upgrades. "It felt like the lens industry was about to be eliminated, which sparks the motivation to take another path." He added, "I was thinking about how not to be eliminated by AI and join it, which is why we developed the CPO project." He now considers it to have preliminary results, though the industry currently still faces a "shortage of land and talent."
Looking ahead, he said that Largan's core business is lenses, and now FA can be "bravely ramped up for mass production," while SiC and electronic materials are not yet at a stage for mass production.
FAQ
大立光為何要開發CPO技術?
董事長林恩平表示,開發CPO是為了「不被AI消滅」。他擔心AI的發展會降低對鏡頭升級的需求,進而淘汰鏡頭工業,因此決定投入CPO領域以尋求轉型與新機會。
大立光CPO產品的試產時程為何?
計畫在2026年9月前設立第一條以FA(光纖陣列)為主的自動化試產線,預計從小量產到量產的時間約需6個月到1年。
大立光在CPO技術上有何優勢?
大立光在精度方面具有優勢,其內部測試已達到0.3微米(micron)以下,優於目前業界最佳的0.5至0.8微米。此外,高精度與4層以上的堆疊能力也是其技術亮點。
大立光CPO初期的主要產品是什麼?
初期產品以FA(fiber array,光纖陣列)為主,它可以和MLA(micro lens array,微透鏡陣列)耦合形成FAU(光纖陣列模組)。
林恩平對目前CPO發展的信心如何?
他表示,相較於4月法說會時還有些不確定感,現在隨著進度越來越好,感覺更為踏實,並認為FA產品已可以「勇敢衝量產」。