Good afternoon, everyone. Welcome to Innolux second-half investor conference.

Participants will be listen-only mode during the presentation. You may download the Chinese and English conference materials at Innolux website.

Please be reminded, the presentation includes forward-looking information based on our current expectations. Actual results may differ materially due to various market factors.

To understand these risks, please take a moment to read the disclaimer notice line. Today's meeting, we will have Mr.

Jim Hung, Innolux Chairman and CEO, to provide us information on market overview, corporate strategies, and financial results. After that, we will conduct a question and answer session.

Now I will hand it over to Jim.

Thanks, Phoebe. Again, thank you for joining us today on Innolux second-half investor conference.

I will start today's presentation with a quick summary. Of course, we divide the company's overall business into two major domain, the display and the non-display.

On the display side, our core business, continue to be impact by the consumer sentiment and also the component price hike. I think as everyone knows that the memory price has been skyrocket, so that also impact into the overall buying behavior.

That we will explain later on on how that impact to our results. On the non-display side, which is much more steady, we has continued to see a improvement on our shipment and also overall earnings performance on the non-display side.

Also we have seen a positive momentum on the market consensus. We have seen some of the upward revision on the company's forward-looking earnings estimate.

Our EPS from '26 to 2028 has been seeing the greatly improvement. We thank you for the investors' confidence on us, and so far this is align with our continue transformation strategy.

We try to lower our cyclicality on the earnings while improve continually on our overall performance. Later on we will explain more into the details.

On the panel's core business, I divide by the products. On the TV side, the overall market shipment was expect to be down about 3% year-on-year, while the area or the size will probably down slightly 1%.

This is mainly due to a lot of the Chinese manufacturers using the high generation, G10 or 10.5 line, for the large-size TV. While Innolux differentiate ourselves on focusing at the 50-inch or below market segmentation, which we believe we have a much stronger cost advantage and also a much better customer engagement.

Overall our shipment on the TV we believe will be flat or even slightly gross year-on-year. On the monitor side, the market overall still focus on the high-end gaming and also the OLED model.

Far I think our high-end monitor has already account more than 50% of the overall monitor shipment. This is also a positive feedback from both of the TV and the monitor products.

On the notebook side, which is impact significantly by the market momentum. I think the major impact from the high memory price has been triggered the customer to have a defensive inventory.

They pull in a lot of the demand from second half to first half, that's also why first half we have been seeing the more positive buying behavior than expected. While in return, I think the second half, probably even to the early next year, we are expect to see some slowing down of this notebook demand.

This is for the customer to try to avoid the cost rising, and also the extended lead time. The momentum on the notebook side, in the next probably 6 to 9 months, we expect to be slowing down.

However, in the longer term, we do believe that AI PC wave will continue to trigger the new refresh cycle in the longer term, so that is a long-term positive. On the other hand, on the non-commodity, i.e.

the industrial demand, which continue to see a very steady growth thanks to the booming of the Edge AI. Also our own ecosystem, like InnoGallery, has seen the positive feedback from the customers.

The industrial demand, we continue to invest and continue to see the great improvement by more orders. Okay.

Next page is on the company strategy. I'm sure that people who have been following us has very familiar with this page.

These four major pillars are the ones we use to identify our own transformation strategy. We review ourselves all the assets from facility to capacity, and also eventually want to yield a better profitability and the return.

Our overall transformation strategy remain unchanged at the light assets, and also maximize our investors' return. In 2026, we have announced to dispose three plants, include both Fab 2 and Fab 5.

Also we have a smaller plant, a module plant, also announced the earlier to dispose. Far just want to emphasize again that we are not dispose this plant for a one-time gain.

More important, by unload this assets, we believe this will help to ease our long-term maintenance expense or cost. This actually help us to improve on our margin side, and we have been seeing the positive feedback on this from both the clients and the internal feedback.

We believe that this positive momentum should continue to go on. This will definitely help with our shareholders' return.

To further elaborate more, in order to maximize our ROE, within a relatively short to mid term, our focus is to improve our overall revenue and earnings breakdown by focusing more into the high-margin business. That is on both non-commodity and on non-panel side.

While our bigger contribution now has continue to help to support and improve our business. Also earlier we mentioned about the divest of our idle plants has also reduce our maintenance cost and the capex.

This for the short-term improvement. While in the long term, we are still definitely focusing at the next major earnings growth driver.

As many people already mentioned, one of our key focus is on advanced packaging, i.e. the FOPLP.

We will use the next few page also explain more on this earnings driver. Other than this, one of our independent company, CarUX, which focus on the panel side or the auto panel.

After last year we announced to acquire the long-term branded company, Pioneer, has been create this positive momentum for the overall smart cockpit environment. We believe that CarUX combination with Pioneer to create a synergy as the Tier 1 position in the auto business will also be the major earnings driver in the longer term.

Again on the semiconductor side, I think people who are familiar with the FOPLP knows that Innolux has invest on all the three technologies, from chip-first to RDL, and also TGV. On the chip-first, it's already in the commercial run, in the mass production.

Our total shipment from very beginning on the testing to right now already more than 10 times growth, and we continue to see a double-digit growth this year. This is a continued business.

Although the technology barrier on the chip-first is lower, however, it is a steady revenue driver and also help with our continued support to the RDL and the TGV business. As you know that chip-first it's only a temporary carrier, you will unload our carrier before the shipment.

While on the other hand, RDL and the TGV, which is much higher entry barrier, and we've been very positive to see our strategy partner to working together with us to provide this TGV and the metallization. As you can see on the slides that our core of the organic substrate was being replaced from original organic to the glass substrate, because we believe this provides a much better interconnect density, and also a much better warpage control and the heat dissipation effect.

Overall we believe that this technology can help with the overall semiconductor industry to move to the next step. We are very honored to working closely with our partners to support to the global supply customers.

Further elaborate on the RDL and the comparison to the chip-first, I think people aware that RDL and TGV are both aim for the high end because the heat and the warpage control are the key for the high-performance computing. On the RDL, because we do the backside modification, which remove or grind away the backside, it actually allow a much thinner IC and also will help significantly with a better cooling effect.

Also because RDL is a compound material, which allows for the finer wiring space, and also because itself, glass, is a more rigid, the warpage effect is also a significant improvement from original flip-chip technology. On the back of both advantage, we continue to see a strong engagement from the customer side to leverage with RDL.

This page shows you that the RDL obviously is a type of the packaging, and we're using that into the glass substrate and as a upstream foundation material. This much better CTE has able to provide a mechanical benefit and also the electrical benefit.

Our ultimate goal is to help to enhance the power delivery performance on the per-unit area. This what we believe is in the AI or HPC era what all the customers eventually looking for.

That's also why we believe the TGV technology can help GCS, the glass core substrate, to fulfill what the customer needs. According to the market research, the FOPLP and the glass substrate packaging, the market expect to have a significant growth, a CAGR of more than 50% over years.

This is on the back of the AI and the HPC's applications. We are seeing this, and definitely witness for the market size tremendous growth.

We expect as the first-moving advantage to help the company to realize a significant share of this part of the business. Traditional packaging, of course, have its limits.

However, we believe it's complementary to each other, it's a feasible solution. We are very open and have a great potential to work with all the OSAT partners.

Lastly, in terms of our financial results. Our revenue in the first half on the non-display and the non-commodity has reached about 55% in the first half, and both domain has a double-digit gross margin.

This is actually what we say earlier our transformation. The whole purpose is to minimize the cyclical impact of the commodity.

We do see this improvement or the growing percentage of the non-commodity and non-display have helped the company to stabilize our margins. Next page shows you our second quarter result.

Our gross profit has improved from first quarter's 14.4 to 14.6%, and also our operating profit also has seen the positive improvement. Of course, this is also thanks to the great team efforts on the cost control.

Company have been working extremely hard on this side. Our depreciation and amortization has marked about $7.5 billion and also our capex about $2.3 billion.

As you can see that in second quarter, our significant increase of the net profit after tax also due to our deferred tax assets for our loss carryforward, roughly about $3 billion. That makes to our total EPS in the first half about 77 cents.

In terms of our balance sheet, Innolux has continued to run on the net cash base. Actually our net debt to equity is -10%.

The company is sitting on the net cash position, and we have trying to further leverage our healthy financial situation to build into the support for our development on both the FOPLP and the other new areas which need investments. However, we are also want to emphasize again that we are very cautious in terms of our future assets investment.

We want to stay more on the asset-light. Our current ratio and, of course, our quick ratio are also at the very healthy stage.

I think I will stop here to take any questions.

現在是QA時間，管理團隊會先回復預先收集的問題，然後再回答線上提問。Now is the QA session. We will have two parts today.

In the first part, we will answer the questions collected from the investor in earlier time. After that, we will take online questions.

預排問題第一題：目前很多企業都投入FOPLP跟玻璃基板研發，群創的優勢跟領先會不會很快被取代？ 好，嗯，我用中文回答好了吼。嗯，當然，這個問題的確蠻多人有呃提到過了。那呃在我們來看，就是面板跟半導體它有相似之處，但是整體業務的思維還是不一樣的。面板是比較接近一個終端的成品，所以它到呃最終的時候它的規格變化比較少，它最終它可能因為這個有Turnkey Solution，設備一旦起來，資本支出投入了之後比較容易變成這個commoditize，就會被呃一般化。但是半導體的IC因為它的整個的這個供應鏈、產品鏈非常的長，從設計到封裝，幾乎都是高度的客製化。那從設計的時候才需要有客戶的engagement，那有很明確的這個出海口，那客戶的驗證過程也長，尤其是針對這種高階的應用的一些半導體，所以除了技術發展還有各種中間的這個過程都需要有很好這個夥伴一起去磨合。那也因為這樣子，所以剛剛我想我特別也提到就是群創非常感謝在這個我們在投入FOPLP的這個過程裡面有非常好的這個合作夥伴，那不管是設備商、材料商，一起把它從一個實驗室的一個產品能夠走向一個量產的過程。所以我想回答您剛剛的問題，就是我覺得這個有點像行業的前輩可能講的吧，它不是像去超商買一個牛奶這麼簡單，就說只是因為它成本低我就要換，它比較更多的時候是因為基於很多的一個互信，然後互相的那個磨合，那邊的這個可以可以調整多少，才能夠達到最終的一個理想，符合客戶的這個需求。所以這個部分我覺得比較難是說換就換，那也因此就說我覺得可能不會是那麼快的被被能夠因為資本投入就被取代。

預排問題第二題：群創TGV技術除了玻璃基板，還有什麼應用方向？ 嗯，TGV的技術。嗯，應該這樣說吧，就是到目前為止，當然呃之所以玻璃能夠運用到作為一個substrate是有它的等於是材質上先天的優勢嘛，因為它的剛性還有它的平整度，CTE又更好，所以它運用在半導體裡面，還有當然它因為它是方形的關係，所以它的utilization rate也比較好。那這些就是剛剛在投影片裡面都提到，對於整個power

integrity啊，對於散熱啊，對於平整度都是它原先優的先天優勢。那但是我覺得不是光侷限在這個，對未來其實還有蠻大的呃應用的一個範疇。像是大家現在很多提到的就是不管是光電光進銅退啦或者在光通的上面，其實最終如何運用到這個這個glass substrate上面，不管是用雷射啦，用怎麼樣去改質啦，或者是用不同的這個waveguide來去做調整，其實還是可以落在這個上面。那因此我覺得它是一個基本功啦，那其實未來可以有非常寬廣的應用空間。那也因此我想很多的夥伴一直在不斷的投入，那我覺得這個是一個正向的，因為唯有更多的人進來就表示說產品啊、這個設備啊、技術啊、材料啊都會更成熟，能夠讓這個越越進步越快。 預收問題已經回復完畢，現在進行線上提問。

謝謝。現在進行線上QA時間。發問時您可以自由選擇中、英文發問。歡迎利用電話或直播網頁下方訊息框提問。發問前請先說出您的機構名以及大名。如果您需要提問，請在電話機上按米字鍵以及數字1，當我讀出您的名字後，請開始發問。如果要取消發問，請按米字鍵以及數字2。謝謝。Now

那我們這邊先回答第一個線上提問，是Morgan Stanley的Daniel Yang，他的問題是玻璃基板的capex跟產能規劃如何？ 好。我這邊先回答這個問題吼。其實這個還在那個討論的過程當中啦，那的確因為非常敏感，也跟客戶之間有互相的保密協定，所以很抱歉我們現在還不能夠把這個數字那個，那但是就是應該這樣講，我們相信在這個發展當中，因為這個技術還在不斷的調整，所以這個金額可能會是我們會盡量維持capex在原來的一個水準，但是會有一次性比較大的投入，那我們會希望能夠藉由更多的這個合作夥伴的這個投入來把這個一起做分攤。那但是確切的這個金額還有時間點可能還要再等待後續更確切的時候我們透過董事會的這個程序才來公告。

以下還有一個線上提問，是元厚投資的Jason，他的問題是TGV process中遇到的主要困難跟目前預計推出glass

core substrate的時程為何？ 呃，TGV process的困難跟什麼？推出glass core substrate的時程。Okay。嗯，具體的時間，again，這當然是要看客戶啦，但是我們相信如果以很多的需求market research上面的資料來來看的話，我們覺得應該呃不會晚過28年吧。那呃這個是目前比較主流的市場看法都看到的應該是比較偏向是在28年。那目前當然就是也如同剛剛前一個問題在提到的，因為這是一個非常新的技術，那這個中間需要呃不是光單一的一個廠商決定規格我可以做到多少，我的深寬比啊、我鑽孔可以做到多少而已，它需要每一層，因為你做完之後你下一層的不管是metallization啊，不管你要增層上去用ABF或其他材質，一層一層的鋪上去，其實都需要很多來回的磨合。那這個影響到設備的這個不管是tact time啦、稼動的時間啊，那當然這也最終因為客戶的呃也會影響到客戶的需求嘛，那到底需要有多少的投資能夠多快搶佔市佔，等等的這些。那對於群創來講，我們就是基本上呃符合客戶的需求，那把我們這個這個部分能夠做到最好。所以呃我們沒有一個100%肯定的一個時間表啦吼，但是我們就是一直在跟客戶來回的這個過程當中，那一關一關的這個問題去去突破。那目前為止的看起來都還非常的一個正面。

接下來另外個線上提問是金雪茄經理儀婷，她的問題有兩個部分：第一個是FOPLP的出貨進度跟明年的狀況如何？問題的第二部分是請問下半年跟明年的整體營運展望？ 好。第一個是FOPLP的出貨進度嘛，對不對？跟明年的狀況。我想您指的應該是chip-first的這個部分啦。那這個部分從原先的一個月大概4,000多萬顆，那到現在已經是超過這個水準，有大概更往上大概兩位數以上的一個增長。那這個部分會在持續的一個增加。那明年的狀況就目前看起來的也是非常的正面，這個部分在chip-first。然後第二個是關於明年度全年的營收啊、展望這些嘛吼。因為我們沒有做那個財測啦，那我只能說就是又以應該說以面板來講，當然notebook上面剛剛提到還是相對比較保守，但是TV上面我們覺得是有一個穩定的一個成長，不管在量或者是價錢的成本的控制上面還可以。那除了這個剛剛的這一個就是純粹的面板以外，另外我們在non-commodity上面一些industrial的應用也有一些不錯的成績，所以這個部分也會持續的這個成長。那當然大家應該最關心還是FOPLP的部分，這個部分還要就是剛剛前面講的要看我們資本投資的這個這個速度，那會在等進一步更確定的時候來跟大家公佈。

現在請有意提問的來賓朋友，在您的電話機上按米1。If

目前還有線上提問。QIC Jason問：請問CarUX未來的上市計劃跟時程如何？謝謝。

嗯，好。呃，那個具體的上市計劃當然還要等待這個董事會啊這些的通過之後我們才會對外公告。不過從CarUX原先的這個定位點，就是希望能夠朝向獨立的這個運營的方向去去走，這個方向是肯定的。那因為去年9月在併購之後，有比較多需要重整的工作，那目前都呃這個部分就要看呃兩方的這個磨合的這個synergy，到有一些成效的時候，會重新再開始這樣子的一個資本市場的一個運作的計劃。不過到以目前來講的話，雙方的合作還算是非常的一個平順。我想再次強調就是因為CarUX原先在於面板上面的這個專長，那配合上這個Pioneer不管在head unit乃至於在speaker上面都有其強項，所以雙方的這個互動非常的這個頻繁，那很多的客戶也都非常正向的一個回復，對於未來在車用空間裡面如何能夠提供到一個更好的一個user experience有很很多的這個互互動。那這個部分我覺得應該是一個非常positive的synergy。 另外還有線上提問，統一投顧的張書銘，他主要的問題是說：公司還有計劃要持續輕化資產或減資去實現封裝帶來的獲利嗎？

嗯。呃，到目前為止其實就是從23年我們有一個市廠的處分，然後主要還有二廠跟五廠今年的這個宣告。我覺得主要的廠大概是這幾個啦，那當然有小的這個模組廠，有一個南科廠的這個處分。那應該會到此是一個目前看起來是一個optimal stage。那這個當中我覺得有兩層意義：第一個就是公司在持續精進的過程裡面這個腳步是不會改變的，就是應該好像大家喜歡講叫騰籠換鳥吧，就是原先的面板裡面的這些成分啊、組成啊這個到底哪個這個會一直持續的去做調整。但是呢，可能不會再以這種出售的方式對外去處分，但是在廠裡面的這個調整，譬如說剛剛也提到像是FOPLP的一個資本支出如果開始做的時候，可能會在其他的這個廠裡面就會有一個重新的一個調整的這個作用會出來，所以這個部分大概會持續在進行，但是不會暫時沒有在對外去處分的一個計劃。我剛剛我講那個四廠的處分應該是2024年，對不起，不是2023。 我們來到今天最後一個提問是來自線上的。有發問者提到說：公司目前手上現金有840億，請問未來有什麼規劃嗎？

有啊，其實有蠻多的規劃啦。就是一方面這個idle cash我覺得是很可惜的嘛，所以一定是除了原先的這些股息的這個配發以外，這個是群創作為對於股東的一個承諾，這個公司賬上有絕對充足的這個資金，不管是原有的現金的這個水位或者是每年所產生的這個現金的收益，作為股息來發放以外，另外就是我們也一直在看各種資本市場的規劃。所以賬上的現金就是除了當然現金股息啊，然後這個capex的這個主要的這個投資以外，那另外任何的一些資本市場的操作我們也都會去探索。 感謝各位的參與。今天會議到此結束。網路重播將於會議結束後一小時內放到公司網站www.innolux.com投資人服務頁面。謝謝您的參與。Thank

you for joining us today. We are going to conclude the call now.

Conference replay will be available in an hour at the company's website, www.innolux.com. Goodbye and have a great day.