Thanks, Jason. Welcome everyone to Parade Technologies 2026 Q2 webcast investor conference.

Parade Technologies' second quarter 2026 consolidated revenue was $134.48 million, and net income was $18.78 million. Its both basic and fully diluted after-tax earnings per share were 24 cents US dollar and 24 cents US dollar respectively.

These results compared to consolidated revenue $133.50 million and net income of $22.79 million or 29 cents US dollar and 29 cents US dollar per basic and fully diluted share in the year-ago quarter. In US dollars, the second quarter revenue increased 6.61% sequentially and was up 0.73% year-over-year.

The gross profit in the second quarter of 2026 was $54.06 million, an increase of 6.35% from the previous quarter and a decrease of 6.06% compared to the same quarter of last year. Based on the current business outlook, Parade is providing the following guidance for the third quarter of 2026: Revenue is between $139 to $153 million US dollar.

Gross margin is between 40% to 44%. Operating expense is between $33 to $36 million US dollar.

各位投資界的朋友大家午安，歡迎大家來參加譜瑞公司2026年第二季線上法人說明會。首先由我跟各位報告譜瑞公司2026年第二季財務數字。本公司2026年第二季合併營收為美金1億3,448萬元，稅後淨利為美金1,878萬元，除權後基本與稀釋每股稅後盈餘為美金0.24元，約當新台幣7.66元，及美金0.24元，約當新台幣7.64元。而去年同期合併營收為美金1億3,350萬元，稅後淨利為美金2,279萬元，除權後基本與稀釋每股稅後盈餘為美金0.29元，約當新台幣8.95元，及美金0.29元，約當新台幣8.92元。2026年第二季合併營收較前一季增加6.61%，較去年同期則增加0.73%。在營業毛利方面，本公司第二季營業毛利為美金5,406萬元，較上一季增加6.35%，較去年同期則減少6.06%。基於目前對2026年第三季營運的展望，我們預期合併營收為美金139到153百萬元，合併毛利率為40%到44%，合併營業費用為美金33到36百萬元。 It is my presentation for the 2026 Q2 financial results.

Now, I transfer to CEO Dr. Zhao to answer your questions.

Jason, you may begin.

Thank you for taking my question. Just first question probably just a little bit housekeeping.

Can you give us a breakdown for the second quarter revenue mix?

Thank you for everyone. For the second quarter, the DP [unclear]

Hello?

Hello? Yes.

Yeah. Okay.

I I'm not sure you you heard my Jerry, you heard heard my answer?

No, I didn't. Didn't hear anything.

A lot of noise.

Okay. The DP was 30%.

PS, our high-speed, is about 45%. TC, our source driver, slightly above 10%.

TT, now include our TTED device, is slightly below 15%.

Got it. Thank you.

Then based on your third quarter guidance, it seems like you're suggesting a little bit stronger than expected demand for PC and notebook market. I think that is a little bit different from the general industry's view.

Can you give us a little bit color on the PC notebook demand outlook? Or maybe elaborate a little bit about the revenue outlook for each of the categories into the third quarter?

Thank you.

In the last conference call, I mentioned that our TTED device, which is integrate the T-con, source driver, touch together for notebook panel, will have a good growth in the Q2. It did happen, and it's actually grows faster than what I had expected.

You see that our TT, which include that device, now close to 15%. We had a thought is over 10%, now it's about close to 15%.

That's definitely reflect we into the new market which will win a lot of models and ship quite a lot of device. That's really the story for Q2.

We are into the TT segment is stronger than what we had planned it.

Into the third quarter, should we still think that the TT is going to be the main driver, or DP high-speed also going to see sequential growth?

Into the Q3, the Q3 is traditional notebook PC hot season, and we continue see this year has the similar seasonality even though customer worry about the memory price increase so on, forth. In our specific case, our TT, which include our the TTED device, will continue grow.

In the Q3, the percentage also will increase. That's kind of more specific theme which the new device take over a lot of the notebook panel segment.

This is the way happen to see because simply notebook panel just one device and for Parade itself, will increase silicon content. Now has a touch integrated to our device and not only supported the normal touch, and also supported the active stylus.

The interesting thing is the notebook panel with touch function, a touch rate is increase and quarter-by-quarter in the general market. We are enjoy those kind of trend with our device, which is quite complicated and without competition so far.

Thank you. Very clear.

Then follow-up question on this year's outlook. I think previously, Jack or Yo-Ming has expecting that this year revenue probably can see I don't know, mid-single-digit or 10% growth year-over-year.

I'm just wondering now you have provided third quarter guidance, and then 4Q probably you should have some idea. How should we think about the revenue outlook for this year based on the current forecast?

We had a more positive view in terms of our business. That was the early this year.

At that time the DRAM the price impact is not that significant. Now we saw a lot of our customer being impacted because high price of the DRAM and the storage, and which really suppress the demand side.

You particularly see the gaming side of notebook. Demand decrease a lot.

Far the Q3 the trend is a good, but we're waiting for the see the Q4 whether we can make a few percentage increase versus last year. That's definitely what we hope.

That's just continue the Q3 and the Q4. The Q4 so far, the PO receiving looks good, but we don't want to commit too early.

Okay. Got it.

Fair enough. Then last question, and I'll go back in the queue.

Just for the ASIC-like business, can you provide some update, particularly maybe a little bit more detail on timeline, when will you start a tape-out and mass production? Because I think the customer you are targeting on probably has its own schedule for a second half 2027 launch.

I'm just wondering, based on your current timeline, are you confident that you can meet your customer's product launch schedule? How should we think about the revenue contribution as well as gross margin?

Thank you.

I think we are progress well. It's a complicated device, and we have a two device there.

It's pretty complicated device and the big device. The project, and we put a lot of R&D resourcing to.

The project progress well, and we did a lot of system design. With our customer together to early look at the performance and the software compatibility and the system-level testing.

Even on the FPGA side, we're already doing that. We're looking forward on time to tape out our chip.

As we shared, sometimes the late of this quarter, the first chip will tape out, and continue second chip will be quickly tape out as well. We think those chips will contribute to our revenue in the late of 2027, as we shared last quarter conference call.

During the quarter, we have a more engage with executive of our customers, and I think the confidence level is getting higher. By working with the engineering team of our customer, and yeah, hopefully once we tape out, we will have a more clear picture, and especially when chip come back, how healthy the chip is.

Those are we certainly looking forward.

All right. Thank you, Jack and Yo-Ming.

Thank you.

Thanks, Jack and Yo-Ming, for taking my questions. My first question just want to ask about your second-quarter margin.

I think it came a little bit shy of Street estimates. May I ask the impact behind this?

Is it because of the material price hike or the impact from the foundry side? I have a follow-up.

Thank you.

Thank you very much for the notice on the those gross margin. I think as we shared last quarter, and the the especially the OSAT price increase, and caught us suddenly.

The OSAT, the back-end side, it's not only increased once, has been increased multiple times. As we are and we are addressing those the the the cost increase.

By two direction, one is share communicate with our customers and like to have a share some cost increase with our customers to reflect the the the product price increase. As for ourselves, we have to looking for more cost-efficient, cheap the the OSAT price to make our gross margin improve our gross margin.

We work on the both side very hard try to recover our gross margin as we we are we're have been doing, and hopefully the next few quarters we can see the improvement.

Got it. On the pricing adjustment side, when do you expect this to take effect?

Next year or maybe in fourth quarter? Thank you.

Some of parts we are segment by segment to do it. Some of parts already in place, price already reflect in place.

Some of them are being this quarter we're talking to customer. It's segment by segment going and selective product to adjusting.

Certainly your OEM customer will not be happy. You have a good communication with them to reflect what current situation, I think they all understand.

It's a negotiation, but most of them willing to take some of price adjustment, yes.

Got it. May I ask the magnitude of this price hike?

Is it like 10%, 20%, or even higher? Cuz as you mentioned, we see several times of price hike from our suppliers, YTP,

yeah. The OSAT price increase is like what you said, and it's widely reported and published by the media on the cost increase.

For us, we are working with our customer to reflect those increase, and yeah, it's different product has a different increase, but has been executed, and some of them are quite meaningful to reflect our cost increase.

Got it. One more follow-up from me.

Do you expect second-quarter gross margin to be the trough of the year?

What I would say, yeah, it's kind of hopefully this kind of bottomed there. That's certainly is our hope.

We will move up as we go, and because at the beginning of quarter, we didn't realize that magnitude change so big. As now we know, and we are doing multiple sides to address this issues.

Got it. My second question, just want to follow up on the standard plus customers' overall demand.

We saw that their demand is quite strong, and do you have a preliminary outlook for their demand into 2027? Thank you.

I think they are pretty healthy, right? It's the demand remain pretty good, I would think.

Based on the forecast they give us, so-called long-term forecast, I think 2027 will be the reasonable year. We are developing multiple chip with them, and some of them will start production late of 2027 or 2028.

They will make a contribution to us. For them, I think the business is a lot more predictable than compare with others, because the impact to them seems not as big as other customer.

Got it. My final question, sorry.

What's your view on the PC shipment into 2027? The overall market.

That's a good question. I had a chat with the OEM industry leaders, and I think 2026 has been caught by the DRAM price increase and suppressed the demand side.

If AI still remain the big theme for the high-tech or the company. AI demands more computing power.

Notebook PC remains as a fundamental the needs for the productivity and for our daily life. With this the subdue of 2026 the shipment, I think people all know it's a big drop.

We think the 2027 will have a hope. To be better than 2026, to grow.

Fundamentally is a demand a more better the notebook and those the AI agent calculation might need a more powerful one there. Hopefully there another segment might be interested.

Just like you have a edge of computing, small computing the box. It's getting popular as well.

Those are the thing added together, and we keep our view positive for the 2027.

Got it. Thank you, Jack.

Thank you, Yo-Ming. Congrats on the great third-quarter revenue guidance.

Thank you.

Hi, thank you for taking my questions. I have a few follow-up.

Just one for the ASIC-like. Could you share some details on the potential ASP and the margins, is that the ASIC-like products its gross margin probably better than the current product lines?

I think the single device price in Parade's portfolio is high, right? We're talking about $10, $20, those kind of range.

Gross margin we are in check with our corporate gross margin. However, those are the assumption that current yield thing.

If yield have an improvement, we might have a do better. That's available there.

If yield not doing as good as we calculated, it might yield not as good as the corporate margin, but should be in the neighborhood of our corporate margin. That's what I can share with you.

Thank you. Yeah.

Sorry, I just joined a little bit late, it might be already discussed. Just want to know more about how fast the ASIC-like contribution or tape-out would be.

We said it will be the late of 2027 on the revenue contribution side. The project is moving forward and doing well.

We are exciting is we working together with our customer system together with our emulation system, their emulation system, and doing the software adjustment. To make sure we already once chip come back and everything's ready.

It's a big chip. We have a two chips to go.

We're all exciting why we work very, very hard.

Got it. Pretty clear.

I also have a follow-up on the data center related. Previously you've been discussing about the PCIe or ACC cable.

Could you give us some updates on updates in design wins and the data center related product lines?

In the Q2, we have been using our acquisition asset, which is we have the silicon germanium cable driver. For 112-gig and 224-gig.

The particular 224-gig, our customer made 800-gigabit per second Ethernet cable, so as 1.6T Ethernet cable. We using that cable participate plugfest in New Hampshire, in US, New Hampshire.

We sent a big team over there and doing the system-level testing. Result looks pretty good.

Surprised, that's the first time we participate, and device looks good. Those for the ACC cable for data center on those Ethernet.

We're talking about 3, 4, 5 meters, those cables. On the PCIe Gen 6 side, those the cable using what was Parade-developed PCIe solution.

Those cables are working with the server guys for the data center, and the world leading server guy needs those cables, and we working pretty close. Those are a little bit shorter cable, but still they will need those cable for the system performance.

Those are the two front we work those case, and we share with you guys. ACC cable on the data center looks pretty promising because the performance are decent.

A lot of cable already been built. Hopefully, we continue to see whether those one will convert to our revenue.

Thank you. My last question would be about the product mix change.

Saw there was an increase in the integrated display solution revenue contribution in second quarter this year. Just wondering how about the product mix all within the second half?

Are you still seeing the rising contribution or faster growth from the integrated display solution over high-speed in second half, and how would that impact the margin trend into by the end of the year? Thank you.

You expect to see the integrated solution will continue growth at least the second half and 2027 because we have a multiple new device will come out to address the more broader market. We are doing very well on those.

Those revenue, unit shipment will grow pretty quickly, and we're happy on this because we invest many years. In terms of gross margin on this segment, and you also will expect we're getting close to our corporate margin.

As device getting mature, the yield getting better. The PS remain is our big focus.

We continue will see our high-speed, the PS product line will grow, but percentage-wise will be different because this the integrated solution, the dollar turns big and the unit shipment getting bigger, and the percentage will grow very quickly. It looks like PS got suppressed.

It's not really, it's really the another side of may increase quickly.

Understood. It's pretty clear.

Thank you for taking my questions. I'll get back to the queue.

Thank you.

Hello, Jack. Can you hear me?

Yes. Thank you.

Okay. Thank you, Jack, and thank you for taking my question.

My first question is kind of a follow-up for the fourth quarter outlook. Previously you mentioned that hopefully fourth quarter revenue should still grow by several percentage point on a year-on-year basis.

I'm just wondering does that follow the seasonality, or is there any other factors including the memory price to have some impact on your fourth quarter outlook so far? Thank you.

I think it will follow largely the seasonality. Particularly, we have a TTED device which win a lot of model, will continue ship a units, and that's additional.

The notebook business has been suppressed for a couple quarters, and hopefully Q4 will have a some improvement. We had experienced not only the memory price increase, we had a shortage of CPU, if you remember, in the Q2 and the early Q1.

Now, I think CPU from AMD, from Intel, you don't hear anymore. Besides NVIDIA's Qualcomm all joining to that segment, and we keep our view positive, and yeah, later Q3 the NVIDIA system will start to the shipping, right?

That's a lot of people pay a lot of hope on this. Let's see how they go.

Thank you, Jack. That's very clear.

My second question is regarding to the foundry price hike. Heading into second half this year and even 2027.

Is that the case that you are seeing so far? Previously we were kind of suffered from the OSAT, the packaging price or cost.

Now looking ahead to your gross margin outlook and also maybe the guidance, does this kind of foundry price hike already being reflected? Thank you.

Foundry all talking about 2027 will increase the price, right? Certainly so far the guidance will have not been included.

We heard the story as well, the foundry might increase the price. It's a tough environment.

We have to deal with it, and I think we are moving quickly to address the back-end side of price increase. Let's see how that play out.

Our OEM customer also hear the story, also hear the cost increase, and they may have to go out just what Apple announced they will have to increase the system price, right? Let's see how that play out.

For the consumer PC notebook business, and 2026, might be 2027, it's become a challenge, quite challenge year. Because squeezed by the AI applications.

Got it. That's very helpful.

Thank you, Jack. I'll go back to the queue.

Thank you.

Thank you for taking my questions. Hi, Jack and Yo-Ming.

Just a follow-up on the TT business. I think you mentioned that TTED demand looks sustainable into fourth quarter and next year as well.

Do you have a rough estimate on its penetration rate, or do you have any target?

We have some model there. It's can be quite exciting.

It's a complicated model. You might want to talk to Yo-Ming to get those models.

Far, on this year we had expected the above 5 million shipment, but now it's much higher than the above 5 million ship units. It's better than what we had expected.

It's fundamentally the device and has provide a lot of value to the customer. We did not expect the adoption so quick.

However, we also have a new device to address even broader market, which the previous LPTS type of panel, now we address the oxide type of panel, which the capacity is a lot more bigger and also went to more massive the notebook segment. We also will have a device come out, the TTED to address the high-end with the much higher resolution, 120Hz the refresh rate, the touch, and active stylus.

It's pretty powerful, and the device. We work with the Wacom.

I think Wacom also surprised, and the single device can do this. You look at the touch.

The Japanese Wacom is pretty standard, right? Let's saw our device integrate solution is so good, and they're also surprised.

Can we have a rough idea about how many TTED chips are used in one single notebook?

Each notebook panel just one, right? Whatever the number we ship is how many notebook were using.

It's quite a big number, and it will getting much bigger.

Thank you. I think just a follow-up on previous question on your standard plus customer.

I think we expect a smaller DRAM impact to the customer. Are we seeing they are having some pricing pressure into their suppliers to mean release some pressure from memory?

Ha, ha.

I think they're openly media discussion they are going to search for the source and the new memory source. I think they will.

We see the other OEM already start to using the China DRAM. For the outside the US market.

I think they got a choke for the DRAM price. Certainly they have to defense to find the memory to use.

I think they will address that. Hopefully they can come through.

I know the Micron is arguing with saying you cannot, but that's what I think. I saw that other OEM already start to use.

Thanks. My last question is regarding your third-quarter guidance, because as in previous questions, I think you mentioned that we are trying to share some cost with customer as well.

In terms of third-quarter guidance, can we know roughly how much would it come from unit and how much come from ASP?

That's pretty difficulty to provide analysis, because ASP side, we're not increase every device. As we shared, we increase selectively the product.

It's pretty difficult. We also move some of manufacturing side to other place to keep a price in check.

Just basically try to dodge along the cost increase. Overall, as we pay a lot of attention, put effort there, I would think in the Q3 the margin will slightly improve.

Thank you. Very clear.

I'll get back to the queue.

Thanks for taking my question. Jack and Yo-Ming, can you hear me?

Yes.

Okay. Just a follow-up from the TTED shipment this year.

You said that TTED shipment could be more than 5 million units this year and...

That was our plan. It's a five

Yeah. Now it's more than that number, right?

Yeah.

I think that's around 3 to 5% of all notebooks per year or roughly maybe 15% of the total touch notebooks. You previously in the first quarter earnings call, you had said if you have 10% share, and what's the difference behind that 10% of market share?

What's your view about that market share growth in the future?

I think 10% should be not that difficult to grow to. You mean 10% of total notebook, right?

There are two reason. One is the attach rate for notebook panel have a touch is increase quarter-by-quarter, right?

That's the first thing. The another thing is yeah, it's really no competition.

I think the reach to 10% it's quite reachable. I think that's the good target for us to cracking.

I hope will be more. That's the goal.

Sure, very clear. My follow-up question is regarding to the competition the TTED market.

Can you share about your if you can share more on your view on the competition like your comp your competitors such as Novatek or Himax?

Far, I don't see they're doing that. We don't see the competition.

That device is so complicated. Far we have not seen any of this.

Especially when you support a stylus, that's really complicated.

Since you think the designs are more complicated, can you share the incremental ASP between TTED and just a single T-con between the ASP difference?

The TTED has a different category which will integrate PSR, which is a frame buffer there or without, right? Also different resolution and things are all supported the LTPS or supported the oxide type of panels.

Price if you look at the whole set of solution, will be compatible as you do the discrete, okay? The set price, because we are only one device.

You need a two source driver, three source driver plus one T-con plus one or two touch chip, right? We are just single device there and to add that together, right?

Those are the equation there. For customer, may see a little bit the price break is from you don't need a T-con board anymore.

The manufacturing is a lot more simpler. The more important, you achieve your industry design that the border is very narrow, and the panel is much thinner, because when you do the T-con board, you have to flip your T-con board on the back side of panel.

Those are the thing add a lot of value to customer.

Got it. That's very clear, very helpful.

My next question is on the data center side. I assume that there's around $10 million of data center related revenue in 2025.

What did the first half in 2026 actually contributed from data center revenue, and your future target in 2027?

We continue ship our the PCIe Gen 5 retimer, even PCIe Gen 4 retimer, we are continuing ship for the data center on the storage side. Those are the stable, and the single device price is high.

We also ship on the motherboard on the system side for the redriver for the storage. Those are the into, and I think the 2026 so far the half year it did not grow that much, but it's equivalent as the 2025.

Those are without considering for the cables. Those are the existing what we had done previously for the design.

Got it. For PCIe Gen 6 active cables and assume that a 2 to 3 meter 2 224G ACC, how many Parade IC are used, and maybe what's the content?

It depend. Some of them need a four, some of them need a two.

It's a depend, and depend on the different cable and the different solutions.

It's a wide range, right?

It's quite wide range. Those single device price is high as well.

Got it. My last question is regarding to the fourth quarter 2026 seasonality.

I think your third quarter midpoint is around $146 million, and it's flattish against same quarter last year. I think in the first quarter, management has think of maybe plus 5% year-on-year growth this year.

That means you have to do maybe over $150 million in fourth quarter, which is up sequentially. How do you think about the fourth quarter at this moment?

I think you have to remember that the unit shipment of notebook PC, we address market, unit shipment on 2026 versus 2025, widely people believe had a 10% reduction, right? If we remain the revenue similar, which equivalent we grow about 10% there on the if the unit price will be the same as 2025.

On the 2026, on the Q4, as previous the people asked, and I also shared, I think that will have a normal seasonality. For us, more specific, is that TTED will increase.

We certainly hope the unit-wise for the notebook PC shipment will increase, because 2026 we ran into too many things. It was DRAM, the price increase, and then the CPU shortage.

Whatever the DRAM price increase, we hope it's already settled. We hope the those CPU and no longer be shortage, and you will have a new CPU joining, as NVIDIA's were joining on the Q3.

Certainly Q4, we have a two device there or multiple device there on associated with that system, and those will help.

Got it. That's all of my questions.

I'll be back to the queue, and thanks again for taking my question.

Yes, thank you.

Again, I would like to follow up on the copper cable front. I would like to ask what's the current target revenue contribution magnitude, and also the timeline for this product line?

That's a good question. We work with the cable guy very closely, and so far cable guy has not given us the clear MP time, and we push very hard doing the system test and those kind of thing.

Hopefully, we can later on we'll share with you and exactly the revenue guidance and those kind of thing.

All right. On this product, I would like to ask what do you think are our advantages lying in compared with our competitors?

Our cable guys actually told us that our solution are pretty decent. I believe they already look at the other people's, our competitor solutions.

We know PCIe better, and we know those competitor better. I think the PCIe will definitely take a lead.

On the ACC side, and since that was acquisition from the Spectra7 last year, and we are still getting familiar with those the ACC cable. The competitor and but so far, our key customer, the cable guy, probably you guys all know who are the major player, all using our solution to make a cables.

The feedback to us is pretty good, and we also participate as I said to the New Hampshire plugfest and for one or two weeks there. Result being shared by the committee, and the result was good.

Got it. Very clear.

My second question is regarding the auto front. I'd like to ask what's the current revenue exposure to auto for Parade?

It seems that auto market has some recovery, or you can say inventory restocking. Are you monitoring that right now, and what's the outlook for the coming quarters?

Thank you.

Thank you for that question. We are a little bit enjoy the auto growth.

I think it's not like what you say the restocking. It's because our solution are associated with newer EV cars, right?

Whether they're in the US, whether in China. Because we have the number one US EV car as our customer, so as we have a number one EV car vendor in China.

Both their increase because the price went up a lot, and then certainly EV car shipped a lot. We even had experienced some shortage there.

EV in the Q3 or Q4 the revenue associated with this was slightly increase, and just simply because the end customer start to buy more the EV cars. That's good.

We have our DP hub to design into the EVs or they will start to MP in the Q4 as well. We are shipping quite our high-speed the so-called USB, auto USB box solution, and we also quite ship some of this.

It looks encouraged, I would think, especially our converter and the DP hub side, and looks good.

Got it. I would like to ask what's the current Parade content per car right now, and you just mentioned you will have several new products in the next few quarters, what's the expected increased content in the car in maybe next year or later this year?

Thanks.

Depend on customer, and How many model do they use? For the US guys and I think each car might be in the neighborhood $3 type of thing.

Some of them people using a DP hub will be more. It's about $5 and those kind of thing.

Unfortunately, the more expensive one may not every car they will use it. Maybe the only high-end and they use it.

Because they have a multiple panel there, and for the China side of EV guys, and they deploy our solution model by model to deploy. Now you're getting more the volume increase.

Those kind of story as different customer act differently. The US customer is once they do, they cross-border change it, but others is model by model deploy.

Got it. My last question is what's the current auto revenue exposure, and what's the targeted revenue that you're looking at in the next few years?

We this year definitely will exceed the $10 million. Hopefully, we targeted to move to the $50 million, those kind of range in couple years, and that's our goal.

Got it.

Very clear. Thank you.

yeah, 我剛才也有回答這個related這個這些問題哦。那我們在Q2

work very hard，這個這個跟我們的customer也work together，在system-level做一些testing，做做一些這個system-level design。那project move 這個very quickly。Yeah，就如as we 我們說的這個這個這個into the Q3 the product就會tape out。我們因為我們有兩顆，這個big one會tape out，那小小的也會quickly follow up tape out。呃我不覺得我們有太多的這個這個project move as 這個我們expected，那那那 就是 the next the second half late of '27會會有revenue contribution。那那我想要大概就這些吧，因為 chip is very big，這個這個我們put a lot of resource there。那chip也是蠻complicated，需要system-level的這些design。 嗯嗯，了解，好。那，如果這個ASIC在明年下半年出來之後，就是還有加上TTED，是不是其實對我們的毛利率是會有幫助是upward trend？

I hope so。這個不會去做一件事情是這個，我想 是不好的。 I hope so。這個我們的corporate margin對我們是很important。

對，了解。所以，因為其實今年是非常辛苦的嘛，就所有東西都漲。那，但是明年如果說明年成長的這兩個都是好的，其實就會壓力就會小一些。 呃就看剛才這個前面的已經share了這個，也許foundry還會漲價啊。這個這個我們要去address，在在這個market裡面 呃一定要去address，去去。但是今年的the back-end漲了算很多很多，這個是surprise到了。

所以所以我們只有跟customer去要求一起去share the cost，對吧？ 嗯嗯，對。

對。 了解。所以，因為我看你們第三季其實毛利率也還是維持啦，所以應該就是今年大概就是就是這樣。因為已經漲很多，應該不至於一直漲，哈哈哈。就是應該是可以維持上半年的水準。

這樣子我們是很很不滿足我們的

這個覺得毛利率是沒有做好。這個這個我們應該continue去improve，這個這個是對的，因為我們的product有這個potential去做這件事啦。

那是說改設計嗎？還是說改foundry？還是就是有什麼方式可以做？ 我我想，第一件事address的就是找一些比較 呃比較relative比較便宜的這個back-end，這個diversify out，這個我們已經在做，對吧？已經在做。那foundry也是一個部分，因為確實 也有蠻多的 customer並不需要我們用這麼leading，不是，用一定要用這個UMC、TSMC，那有一些就會比較鬆一點，因為他們對cost也是很在乎。所以 所以要要要去diversify out，這個這個這個是true了，這個是true了。比如說就像我們share的，我們有一些的TTED啊，我們繼續找一些，因為 一開始的技術已經成熟了，我們可以diversify out去跟一些foundry做一些比較cost比較低一點的，對吧？這個這個現在的這個cost變成 這個very significant to us，我是覺得這個gross margin對我們是非常非常重要。

懂，了解。好，那我想問另外一個問題，就是因為其實大家好像也不怎麼提AI PC了嘛，那我不知道Jack在美國看就是有沒有什麼別的 想法，或者是 就是對AI PC還可以期待嗎？或者是說 可能在agentic force進來之後，就是也許在明年可以去想就是high-end也許會開始賣得好？ Edge computing肯定會越來越大，而且high-end的一定會。

因為我們現在用的都在cloud上面run model嘛，對吧？ 對對，沒錯。

那這樣的方式會是 越來越貴。 這個 啊，會越來越貴，是嗎？

對，包括我們自己，我們公司也有AI哦。 這個一開始的AI都比較便宜，cloud，覺得多少。

但是用多了以後，這個是exponential increase。 嗯。

那越來越多的會在in-house去build所謂的

open-weight的model，包括Parade我們自己也會在做這種事情。 嗯嗯，嗯嗯。所以成本是自己大？

這一類的是呢，你就會要local 自己的computer要有一些這個computing power了。 那 這個就是edge computing這個逐漸逐漸起來了。那我們希望這種model會 會在enterprise裡面更加popular，對吧？那 希望有比較powerful的notebook，或者甚至有dedicated box。

dedicated box，那這也是一個 我們pay a lot of attention的，就是dedicated那些這個這個AI的box，這個這個就像 呃前幾個quarter這個這個比較hot的這個 這個用Apple的這個 這個 Mac mini。

Yeah，做做這些做這些東西嘛。

我我想這個trend應該會在那邊了，因為especially這個這個AI大家發覺發覺到，哦，這個really expensive這個這個 afterwards。我自己我們公司也experienced，一開始的時候很便宜，到後來 很多越用越多，這個這個是exponential cost increase。 嗯，了解。那那如果說這個趨勢成開始成形的話，呃譜瑞在這個地方

呃當然notebook就我們本來的生就是notebook有點好，當然我們好嘛，那有沒有什麼就是可以就可能有更多機會，在high-speed嗎？還是 Yeah，我們有在談蠻多的，希望在正上面 跟

CPU vendor、system vendor有更多的 這個合作的機會去做這些 事了。 嗯，嗯嗯。

現在的不管是AMD或者Intel或者什麼， 對。 都還在focus on這個data center的這些 computing的過程。

我覺得逐漸逐漸他們會轉，比如說 這個 有些customer因為他沒有data center的東西，他就開始轉了。就focus on這個more on the 這個edge side的這個calculation。

懂。那如果是這種enterprise的這個，就是 推行的對象會是誰？就是現在的NB的品牌業者嗎？ 還是它的整個ecosystem會是什麼？

一定是這些

這個enterprise people會用得更多。我們已經有一些 包括我們公司裡面也看到這些needs，對吧？我們自己有一些build一些model就發覺，哦，我要不同的東西去run了，對吧？這個這個我們已經看到了。 我們有一些design，包括我們的 呃 TTED有一部分的design， 我們都在用AI去做這些design。這個這個 嗯，對。

逐漸逐漸就發達到，哇，這個very expensive，let's using our own,哈哈哈。 對啊，我們已經run into這種expense，在美國已經run into這種expense。我想不會是我們 一家啦，有很多的企業 enterprise都會碰到這個問題。

但你們有比較過就是如果自己蓋、用自己local的這個 就會比較便宜嗎？ 因為好像也要算TCO什麼的。

Very sure，it's a lot more cheaper。 嗯，了解。

懂。 至少不用被賺。

哈哈哈，那個費用。 但是我覺得

有到一定的時候，你unaffordable，那個number是exponential

growth。 嗯嗯，了解，懂。所以這個明明年有可能就可以看到一些端倪嗎，明年？

我相信這個industry大家都很smart，會去想這不同的辦法。 嗯嗯，嗯嗯，懂。好，那我知道了，謝謝Jack。

謝謝。

Everyone, thank you.