BizLink Holding Inc

TWSE:3665

Q1-2025 Earnings Call

[Mike Wang, executive]

Good afternoon, everyone. Welcome to BizLink's First Quarter 2025 Earnings English Conference Call. This is Mike Wang, Senior IR Manager. I am joined by Roger Liang, our Chairman; Felix Teng, our CEO; Florian Hettich, our COO; and Charles Tsai, our CFO. Our results were just released and are available on our IR website where you can download the latest earnings materials as well as access them from MOPS. This 1-hour call will begin with Charles for financial highlights before we switch to Florian for operational highlights and then end with Felix for corporate highlights. We will then conclude with Q&A. You may type in your questions now in the public or in the private chat, and we will answer as many of them as we can. There will be no quantitative forward-looking comments.

Before we continue, please kindly be reminded that today's discussions may contain qualitative forward-looking statements based on our current expectations, which are subject to significant risks and uncertainties and may cause actual results to differ materially from those contained in these qualitative forward-looking statements. We are not obliged to update these statements, which are to be used for information purposes only. Please refer to the Safe Harbor notice in our earnings deck for more details.

I would like to remind everyone that today's call is being recorded. This recording and these prepared remarks will be uploaded onto our IR website within 24 hours after the conclusion of this call. We sincerely appreciate Fubon Securities for hosting today's call.

With that, I will turn the call over to our CFO, Charles.

[Tse-Shen Tsai, executive]

Thank you, Mike, and hello, everyone. Let me quickly walk you through our Q1 numbers.

Our scale continued to grow with sales reaching a new historical high. Our gross margin improved to 30.38% and is just within reach of historical high of 30% to 31% -- 31% to 32%. Our operating expenses increased with G&A rising more sequentially, while R&D were also up to support future growth. Our OpEx to sales ratio rose to 15.74%, resulting in an operating margin of 14.64%, making another new record. Non-OP was negative, mainly driven by finance costs, which continue to fall. Our effective tax rate rose to 28.97%. EPS was TWD 8.49. IT DataComm sales rose 27% quarter-on-quarter and 99% year-on-year and accounted for 33% of total sales during the first quarter of 2025. This marks the sixth consecutive quarter of sequential growth as our HPC sales expand faster than all other businesses, rising 51% quarter-on-quarter during the quarter. This was led by our data sales, which grew by 87% quarter-on-quarter, while our power sales were also strong, growing by 14% quarter-on-quarter.

Our Industrial segment remained the largest at 39% of total sales, and it grew 4% quarter-on-quarter and 19% year-on-year. This was led by our capital equipment sales, which grew by 11% quarter-on-quarter. Excluding capital equipment sales, our Industrial segment continued to bounce along the bottom. We are seeing the book-to-bill ratio for INBG to rise above 1.0x for the past few months as the sales decline slow and order intake recovered, hitting that sales or bottoming for us in Europe. Healthcare sales continue to be stable with growth being driven by treatment solutions within our tailor-made product unit. Our total HPC and capital equipment sales mix was 42%. Auto sales fell by 4% quarter-on-quarter and fell by 32% year-on-year, while electrical appliances sales fell by 1% quarter-on-quarter, but are still up 19% year-on-year. This last 2 segments accounted for the remaining 26% of total sales.

In our bigger picture, we're entering this new phase of tariff from positional strength. Back in 2018 to 2019, during the first trade war, 60% of our production was in China, yet we sustained gross margin of 22% to 23%. Through COVID-19 and full production shutdowns, we improved to 24% to 25%. In 2022 to 2023, amid global supply chain disruption and the INBG acquisition, margin rose again to 25% to 26%. Over the past 5 quarters, we averaged 28% to 29%, demonstrating the success of our transformation. While the tariff landscape remains fluid, our structural improvement are firmly in place. We're prepared to adapt and are closely monitoring global trade development. Our record of execution during past disruptions is clear.

Today, our balance sheet is stronger and following significant deleveraging in 2023 to 2024, we're strategically reinvesting internally generated cash into growth, especially in HPC and capital equipment, where demand for production capacity, global reach and engineering strengths continue to rise. This is only the beginning. We're securing new customer projects and expanding our opportunity set. Our ongoing excellence initiative has improved efficiency across sales, operations and finance. These efforts are clearly reflected in our financial results and will help us absorb tariff-related costs while maintaining competitive pricing.

In parallel, our diverse customer base, balanced geographic footprint and broad segment mix give us multilayer resilience against macro shocks. We're meeting today's uncertainty with focus and action. Instead of reacting, we're proactively reinforcing geographic resilience and deepening customer engagement to ensure business continuity. While uncertainty caused hesitation elsewhere, we see disruption as a catalyst for selective share gains, particularly with global customers who prioritize agility, responsiveness and supplier stability. By staying ahead of risk, we're becoming an indispensable long-term partner. This approach is already reflected in our performance.

Revenue growth remains solid. Earnings are growing faster, thanks to operating leverage and disciplined execution. Our balance sheet is healthy. Debt metrics are improving, and we have gained a level of capital allocation flexibility that didn't exist a few years ago. What does this mean in practice? It means that we're in control. We're not reacting to short-term pressures. We're leaning into long-term opportunities. One of the most important ways that we're doing this is by accelerating investment in technology and product development. We're increasing R&D intensity in 5 core growth areas.

The first being high-speed copper and optical interconnect and high-voltage solution for next-generation data centers. Second, complex system box fill and system integration rack for front-end semiconductor tools. Third, humanoid robot connectivity, including high flex high [indiscernible] solutions. Fourth, advanced platform for autonomous vehicles; and fifth, miniature solution for invasive medical applications. This platform demand ultra compact high reliability and low latency solution, our specialty. The connectivity backbone is becoming a key differentiator, and we're investing now to stay ahead. From day 1, our strategy has been partnered directly with Tier 1 customers and co-developed solution to their most complex challenges.

As our customers expand into new markets and technology, we grow with them. We're especially focused on HPC and semiconductor, where AI accelerated computing and edge infrastructure are fueling robust near-term demand. This is a generational shift in computing hardware, and we're scaling both technical and commercial engagement to cement our role in this ecosystem. At the same time, we're preparing for longer-term growth in humanoid robotics and autonomous driving, technology that will redefine architecture and require exactly the kind of compact high bandwidth and durable connectivity we provide. We are already partnering with early adopter and engineers to lead the next wave of innovation.

Our health care commitment is just as strong. We're building capability in high-growth surgical and interbody applications where advanced raw material, extreme miniaturization and precision manufacturing are critical to improving patient outcomes. These are not opportunistic moves. They're long-term secular growth step that we aim to lead. Importantly, we're funding this growth from a position of strength with self-funding initiative without a promising margin or increasing debt. Though we are prepared to explore additional funding as opportunities expand. This balance executing today while investing for tomorrow is a key differentiator.

Our strategy is clear. First, to sustain strong current performance while building future relevance; second, invest in innovation from a position of strength; third, fund growth responsibly while preserving financial flexibility. Fourth, move in lockstep with customers shaping the future of digital, industrial and mobility landscapes. We're not chasing headlines. We're building a foundation for compounding long-term value in market where real growth is just prepared for disruption, not reacting to it.

To summarize, strong earnings, resilient positioning, disciplined capital deployment and a long-term strategy aligned with where the world is going. We deeply appreciate the trust of our long-term stakeholders and look forward to continuing this journey with you.

Florian will now provide update on our latest quarterly operational takeaways.

[Florian Hettich, executive]

Yes. Thank you very much, Charles. So I would like to begin with the tariff and the situation here.

So customer activity remains broadly in line with our expectations, though visibility, as you can expect, is beyond the 90-day window is still limited due to evolving tariff negotiations and ongoing global trade uncertainties. The pressure on our global teams is real and is real high, but we are managing tariff-related risks effectively through selective supply chain relocations, operational flexibility and proactive planning. Importantly, most of our intercompany shipments are structured such that tariff obligations fall outside of our responsibility. This structural advantage has shielded us from most direct impacts.

Our solutions typically represent a small percentage of the customers' overall bill of materials, but they are mission-critical when it comes to performance, reliability, quality and delivery time lines. That positioning limits the commercial effectiveness of price reductions as customers prioritize continuity and technical assurance over marginal cost savings. Our globally diversified footprint continues to minimize operational disruption, allowing us to act swiftly when localized conditions change. As Charles noted earlier, we have faced similar macro challenges before and have proven playbooks to navigate them effectively.

In consumer-facing segments, particularly electric appliances, we are seeing near-term softness and pricing pressure. However, these trends appear more tied to broader consumer caution and lingering overstock than directly to tariffs. Automotive remains subdued, albeit already at a lower base. In contrast, we are refocusing our attention and resources on AI-related growth in IT DataComm. Meanwhile, industrial and infrastructure verticals, including HPC, capital equipment and health care continue to show resilience, backed by durable structural growth drivers like AI adoption, reshoring and supply chain diversification.

On the operations side, we have strengthened our supply chain flexibility across Asia, North America and Europe. While this has led to moderate increases in operating costs and selective capital expenditures, these investments support long-term adaptability and customer responsiveness. We have also kicked off targeted localization efforts for raw materials, not only to reduce risk from potential counter tariffs, but also to prepare for supply disruption tied to geopolitical instability or supplier volatility. Early and proactive engagement with customer remains our top priority. We are embedding business continuity into the fabric of our operating model, ensuring it is a design principle, not just a contingency plan.

Shipment holds have been isolated and short-lived so far. In HPC, we are seeing a clear acceleration of customer activity with data center build-outs driving demand pull-ins and additional volume commitments. We have experienced no material losses of businesses due to tariffs, though some new project launches have been pushed back as customers reassess supply chain strategies. We remain fully prepared to support customer-directed site transitions wherever needed. While near-term conditions remain fluid, we believe we are well positioned to navigate through the uncertainty. Our direct exposure to tariffs remain limited and our regional diversification insulates us from concentrated risks. The larger concern lies in the potential broader economic impact of tariffs, but even that may take time to manifest.

Business dynamics could differ materially during and after the current 90-day window. Encouragingly, the latest steps between the U.S. and China mark a constructive development offering short-term economic and market stabilization. That said, we remain prepared to respond swiftly to any shifts in policy or demand. Consumer segments may continue to exhibit volatility, but industrial and infrastructure categories remain solid and offer selective opportunities for share gain. We are carefully managing pricing and cost structures to preserve competitiveness and the financial impact from tariffs to date has been modest and well within our ability to absorb. We believe our strategy anchored in operational agility, customer proximity and investment discipline in long-term secular growth markets will allow us not only to navigate this volatility, but to emerge stronger, more essential and more embedded with our customers.

If we look at IT DataComm, so that we can see that amid global uncertainty, structural drivers in AI and HPC infrastructure are gaining momentum. We believe we are still in the early phases of AI infrastructure development. Most hyperscaler investments to date has focused on training large language models using text-based data sets, pushing CapEx to record highs and driving robust demand for high-speed data and high-power interconnects. Yet inference workloads are only beginning to scale and already we are seeing signs of bottleneck and outages.

AI agents are now expanding from text-only to visual and image-based responses. This evolution means the next wave of applications like autonomous driving, humanoid robots and immersive user interfaces will require models trained on enormous volumes of video and multimodal data. These workloads demand dramatically more compute, power and bandwidth, which will require next-generation data center architectures that can handle higher density, voltage and throughput. We are actively developing high-voltage platforms for future high-power deployments along with higher speed, longer reach active electrical cables, AECs to meet rising bandwidth requirements. We are also advancing our PCIe interconnect technologies to support scalable fault tolerant GPU clustering. These initiatives are tightly aligned with the shift we see coming and are designed to keep us at the forefront of this transition.

We expect GB200 supply to gradually and then visibly improve over the next few quarters, and we are focused on delivering our qualified solution accordingly. Leading ahead, we are already prepared for GB300, which will include more power content, though we do not expect material revenue contribution until 2026. While our position in GB300 is strong, we believe our ability to execute and remain a trusted partner is just as important. We continue to emphasize reliability, responsiveness and readiness. In parallel, we are actively building out our CPO solution and business model in collaboration with key supply chain partners. We anticipate this becoming a meaningful part of our HPC portfolio within the next 2 to 3 years. Several potential customers have already shown strong interest. Our global -- our goal is to become one-stop HPC platform provider, offering integrated solutions and collaborating closely with customers on their road maps and architectures.

We are also expanding our engagement with new hyperscaler customers and expect initial contributions from some of these relationships as early as in the second half of 2025. The broader shift from narrow AI applications to multimodal AI ecosystems is creating a sustained investment cycle in infrastructure, and we are positioned to lead not just to participate.

In summary, we see growing durable tailwinds in HPC and AI infrastructure. The transition from text to multimodal workloads will drive exponential increases in compute, energy and interconnect needs. Recent commentary from hyperscaler reinforces the view that we are now entering a second accelerated wave of AI-related CapEx. These investments confirm that leading platforms are doubling down, not pulling back on AI. We are investing ahead of the curve to ensure we remain a core supplier as the market evolves from today's high-speed networks to tomorrow's immersive data-rich AI ecosystems.

Now looking at Industrial, the strength we are seeing in AI-related spending is also benefiting the broader semiconductor ecosystem. While recent headlines have focused on ASML's soft EUV bookings, it's important to note that this is tied more to the timing of advanced lithography ramps than to underlying market weakness. Most cutting-edge AI chips like NVIDIA's GB200 are being manufactured using combination of major EUV and DUV processes often enabled by chiplet architectures and advanced packaging. As a result, demand remains healthy across a broader set of front-end tool makers; Lam Research, Applied Materials, KLA, Tokyo Electron and ASM International. These companies continue to see strong order trends in etch, deposition, inspection and packaging. Their latest earnings also reaffirm AI's role as a major driver of CapEx within the semiconductor supply chain.

The shift from large language models to vision and robotics-enabled AI models is creating an entirely new floor for compute and power demand, and we are aligned with this growth. The Easys integration continues to progress smoothly, and we are now accelerating capacity expansion to capture near-term market opportunities. At the same time, we are transferring Easys specialized box build expertise across our global integration teams in Singapore, Penang, Hainan and Fremont. This will enable us to offer a truly global order fulfillment and prototyping solution for our semiconductor equipment customers, derisking their supply chains while enhancing flexibility and time to market.

Demand for larger, more complex box build and system integration project is rising rapidly, and we are scaling our operations to stay ahead. We are also expanding beyond electrical distribution system into fluid distribution system, unlocking a significantly larger addressable market. By offering fully integrated solutions, including both electrical and fluid, we are building the next-generation backbone of AI infrastructure. We are also deepening our presence in humanoid robots where we see multiple content opportunities across different business units.

Customers increasingly value our broad portfolio, which reduces supplier complexity, shortens response times and improves integration. These advantages, coupled with our investment in talent, sites, capabilities allow us to outpace the broader market and further entrench ourselves as a mission-critical partner. In short, we are not only capitalizing on near-term AI momentum, we are investing to shape the long-term future of global infrastructure. As AI continues to transform the industrial landscape, we are scaling with speed, precision and customer alignment.

Felix will now provide updates on our latest quarterly corporate takeaways. Felix?

[Chien-Hua Teng, executive]

All right. Thank you, Florian. Our company is fundamentally different from traditional global peers, and that is by design. Rather than relying on legacy manufacturing models, we have deliberately built a business model that is evolving into a leading hybrid OEM/ODM platform that combines deep engineering expertise, decentralized agility and Silicon Valley-driven innovation. Our executive team is distributed across all 3 of our major region footprints, enabling faster decision-making through local presence, constant communication and close collaboration. Their work is grounded in mutual trust, teamwork and deep experience. We place strong emphasis on management development, actively identifying and nurturing young talent to build the next generation of leaders.

Let me walk you through the core competencies that make us unique and how they translate into real competitive advantages. First, our engineering depth and co-development agility make us a preferred design partner. We do not just manufacture to order. We work side-by-side with our customers to co-develop solutions, accelerating their innovation and embedding ourselves early in the design cycle. Second, our decentralized but strategically integrated operations allow us to execute fast with global consistency. Our teams can move quickly at the local level while staying aligned to global standards, enabling us to deliver tailored solutions with precision.

Third, our Silicon Valley DNA combined with a global footprint give us early design visibilities and the ability to deliver timely, cost-effective solutions. Being embedded at the center of innovation means we see future trends early and act on them quickly across our manufacturing network. Fourth, our disciplined M&A integrations allow us to accelerate product and market development. We know how to bring in new capabilities and scale them without disruptions, strengthening our platform as we grow. And fifth, our cross-unit collaboration drives smart scaling into higher-value areas. We leverage the strength of our different teams to move into higher margin, faster-growing parts of the market. These core competencies have not only allowed us to navigate through periods of volatility, they have enabled us to consistently reinvent ourselves to capture new waves of growth.

While we recognize that near-term macroeconomic and geopolitical uncertainties remain, we remain fully focused on our long-term potential. We believe that staying disciplined in our strategy and continuing to invest ahead of secular growth trends will position us to emerge stronger from this period of volatility just as we have in the past. Looking ahead, we are expanding our box build and system integration capabilities to capture a greater share of high-value assembly projects from our SPE customers. At the same time, we are investing in automation, AI-driven operational enhancements and a decentralized global footprint to deliver value to customers faster, more efficiently and at greater scale. These actions are positioning us to not only maneuver effectively through the current environment, but also to scale with accelerating demand in HPC and semiconductors in the near to midterm and to capitalize our emerging opportunities in autonomous driving and humanoid robotics over the long term. Simply put, even amid uncertainties, we are not just adapting to change. We are leading the next wave of it.

Now let me turn the call over to Mike.

[Mike Wang, executive]

Thank you, Felix, Charles, and Florian. This concludes our prepared statements section. Now let us begin the Q&A section. Please type in your questions and then we will answer as many of them as possible in the time remaining.

For the first question, there's still a lot of questions about the AEC side of our HPC business. Let me read the 2 questions I see right now. Do we see the market competition of AEC market? How do BizLink maintain its advantage in the AEC market? What's the competitive advantage of BizLink now as a supplier versus new entrants?

For this particular question, I'd like to hand it off to Roger.

[Hwa-Tse Liang, executive]

Okay. Mike, can you hear me?

[Mike Wang, executive]

Yes, please.

[Hwa-Tse Liang, executive]

Okay. This question is very good. So there continue to be a widespread misconception around what is involved in being able to supply high-speed AEC in volume. AECs or electrical cables may look like relatively simple product from the outside. But delivering reliably at scale, especially at 100G or 200G per end is extraordinarily complex. There are a few key reasons why this market remains so difficult to penetrate. As Credo said before, it's not just about buying a chip and the cable and putting it all together. There is an entire business model with a hardware and software ecosystem to support it. This ecosystem was built from the ground up, and it continue to be adjusted in order to advance AEC technology even further. Credo has built their edge over a decade of certain innovation, system level partnership and with vertical integration.

New entrants face a multiyear [ half year ] battle in R&D, manufacturing and the customer qualification just to catch up. That's why there are virtually no second source high-volume AEC provider today. Credo has been and continue to be an incredible partner. It took us several years to develop the very first AEC product, and we have experienced many challenges and achieved many milestones together along the way. We have invested significant resource to help create the market for AEC and expect much more growth potential in the next few years. Even within the same customers and the same data center, there may be multiple designs requiring fast turnaround NPI capability. Each AEC has many components and are not easy to make as it's not your typical consumer electronics cable. They are constant and ad hoc design change, which will also change the process of producing and testing them.

Finally, even if you build a technically sound product, getting qualified by hyperscalers take time and trust. The qualification cycle are rigorous, and they often favor qualified and proven vendor with a track record of performance, interoperability and support. Our AEC product is difficult, is efficient in terms of both power and cost. And we have proven to our hyperscaler customers that we are flexible and reliable with in-house engineering and manufacturing expertise. In short, the barrier are highest and they spend across chip design, system architecture, manufacturing and customer engagement. We will spend years building up the know-how and the relationship needed to succeed in this space. And that's why today, there are still very few credible competitors in high-volume AEC deployment. Finally, we continue to develop new AEC products, aiming to provide higher speed and longer cable, which is becoming increasingly hard to do as they switch from 100G per end to 200G per end and eventually to 400G per end. We will be starting volume shipment of 800G next quarter.

This is my answer. Okay. Back to you, Mike.

[Mike Wang, executive]

Thank you, Roger. I know that was quite a bit. But hopefully, that answered the 2 questions that were posed in the chat. This is a question that we seem to be getting every quarter, but for the benefit of everyone dialed in today, let me read it. It's still AI related or HPC related.

I'm asking this on a quarterly basis. And I know you won't comment too much on specific customers, but from GTC, we gain more understanding regarding more power consumption for GPU-based server racks and potentially more AEC's content value opportunities in GPU-based racks. Could you give us updates on your AI business demand situation compared with 3 months ago?

For this question, I'd like to hand it over to Felix.

[Chien-Hua Teng, executive]

All right. Yes, thanks for the questions. And yes, as highlighted during the GTC and enforced by broader industry trend. We continue to see structural tailwinds across both power and high-speed interconnect content within GPU-based server architectures. The ongoing shift toward large-scale AI training and inference clusters is driving meaningful increases in rack-level power densities, where prior generations operate at 20 to 30 kilowatts per rack, we are now seeing deployments exceed 50 kilowatts. And in the case of next-generation systems like Blackwell and Rubin, 200 kilowatts per rack is increasingly common. This transition is accelerating the adoption of higher voltage power architectures, including 54-volt, 400-volt and even 800-volt DC.

These architectures, particularly when combined with liquid cooling and off-rack power deliveries significantly increased the complexity and value of the advanced power distribution and interconnect solutions, area where we were strong, strongly positioned. The interconnect side, AEC are becoming increasingly critical in this dense GPU racks. As lane speed moved to 100G and 200G and as connection distance stretch beyond DAC limits, but remain cost prohibitive for optics, AECs offer the optimal balance of reach, signal integrity and power efficiency.

Many of these deployments are hyperscale in nature, and we are seeing growing opportunities for multichannel AEC configurations, particularly for 1.6T and eventually 3.2T architectures, where both content value and technical differentiation increasing meaningfully. Compared to 3 months ago, we have seen increased design activities tied to AI platforms, specifically around GB200 and early-stage GB300. These systems are pushing power envelopes even higher and driving greater demand for long-reach high-performance AECs. Engagements have deepening both in terms of volume and technical scope. While we won't comment on specific customers, the trajectory is clear. GPU-based architectures are structurally increasing system level power and connectivity requirements, and we are well positioned to support that shift. We believe our content value per rack is rising, not just incrementally, but structurally as AI infrastructure becomes denser, harder and more bandwidth intensive.

Now back to you, Mike.

[Mike Wang, executive]

Thank you, Felix, for the color on that and also to the audience for posing that question. We see here's another one.

Semiconductor equipment demand remains the key growth drivers for the Industrial segment. Any other applications within industrial sectors, we can expect momentum to recover? And also part of this question, we see inventory across IDMs remains high. What's the current order visibility for auto demand? When should we expect to see a recovery in auto demand?

There's a lot to unpackage here. But for this particular question, I'd like to once again hand it over to Felix.

[Chien-Hua Teng, executive]

Yes, more industries. Yes, definitely more interesting. Okay. So we have seen dramatic investments in AI humanoids and AI drone markets in the industrial sectors. A majority of conglomerates have invested in the hypergrowth market. And as to AI agents, the market is still searching for tier apps to emerge. So we are actively building our presence in humanoid robotics from prospecting potential customers to engaging with target contacts for design and sampling to establish a business for delivery and winning of robotic subsidiaries. There are multiple areas for us to win content across our various business and the customers like that we are able to offer them a wider portfolio to choose from to reduce supplier complexity and response time.

These end customers also like that we have multiple capabilities in-house, including NPI, production, assembly and testing as well as specialized cables, complex harnesses and modular solutions and can support them in more locations. Auto demand remains weak with no clear signs of a bottom yet. However, despite this ongoing weakness, we continue to be involved in many customers' projects with a particular focus on supplying more module solutions requiring greater resources, including engineering expertise. We are preparing to help enable the future of autonomous vehicles in the near future. While our primary customers have been relatively weak in the near term, we are seeing strong sales growth from other U.S.-based EV start-ups, particularly following the launch of their SUV models. These companies are expected to introduce more affordable vehicle lines in the future.

One of them has been especially aggressive in its expansion efforts, including the acquisition of a factory from a bankrupt automakers, which position it well for potential larger scale growth ahead. Although this momentum is not yet sufficient to fully offset the decline from our main customers, we review this emerging player as a promising contributor within our EV business going forward.

Okay. Back to you, Mike.

[Mike Wang, executive]

Thank you, Felix, for the insight as well as the person that asked this question. Let me see there is something I can reply to.

Could you please give us HPC and semi-cap equipment percentage of total sales in the first quarter of 2025, respectively?

For that, first quarter of 2025, HPC was 28% of total sales. Capital equipment was 14% of total sales. We earlier said in the Charles' prepared remarks portion, the total 42%.

In terms of first quarter IT DataComm sales, could you please give us a breakdown by DataComm and PowerComm, respectively?

I assume that's what data accounts for and what power accounts for within HPC, right. I can give you that as well. Within HPC, roughly 62% was data and then the remaining 38% was power. So that would answer the second question.

Again, we are not providing any forward-looking comments. So could you please give us the sales outlook for second quarter 2025?

Again, we will not be providing forward-looking.

Do you have any humanoid robotic projects on hand and revenue contribution in year 2025 and beyond?

Yes. Let me give this part to Felix to help.

[Chien-Hua Teng, executive]

Yes. Yes, we do have a project going on with our customers. Although some of the point is not huge. However, it's like a very promising customers in North America, and we have 2 customers we are delivering, okay? So again, I mean, we are -- we don't expect any material volumes this year. And so we look forward to maybe next year if we can get higher -- to see higher volumes.

[Mike Wang, executive]

Thank you, Felix, for the following up on that part and then also to the investor that asked that question. Let me see what else is in the chat.

Okay. For the second question also from the same investor, are you allowed to work with other customers, not constrained by Credo only?

For this one, I'd like to hand it over to Felix as well.

[Chien-Hua Teng, executive]

Yes. So a simple answer. I mean, we are not -- well, we are not restricted by working with other customers. However, for now, I think, as Roger mentioned at the beginning, Credo is right now the most important player, AEC player in the market. And actually, we have been busy to fulfill their requirements, their demand. So I think for now, at least in the near future, we don't -- I think the partners between Credo and BizLink are still very -- not exclusively, but just very strong, and they work with us and work with them, right.

[Mike Wang, executive]

So thank you, Felix, for providing color as well as to the same investor that asked that question. We see what else is in the chat.

There are questions that are very specific. So we will be -- and then we did talk about avoiding anything that's specific to a particular customer. But we can answer a portion of this question.

Would you share what is the content upgrade in data and power products in GB300?

Again, we won't be disclosing anything specific, but there's some color we can provide.

What is the competition landscape in GB300 compared to GP200? Do we see more competitors come in?

For this one, I'd like to hand it over to Felix.

[Chien-Hua Teng, executive]

Yes. I think also we mentioned earlier that GB300, we don't see material shipments until next year, at least that's what we see right now. However, there, indeed, there are some challenges in design. I think people have heard that there were some design changes, big design changes recently. So anyway, so I think right now, we are working with the customers directly. So because there are quite high challenges. So not like -- I would say GB200 like already not more mature, okay? So you do see that some of the small component suppliers, they are making some less sensitive components. However, for the GB300, it's -- right now, we are still in the very challenging design stage. So not many suppliers, not many competitors.

[Mike Wang, executive]

Thank you, Felix, for regarding that part of that particular question. Please, investors, analysts, if you have any questions, we'll take a look at them and try to answer in the time left. So please type it in. There are no more questions. I'll leave it at this.

Thank you, Felix, Charles and Florian. This concludes our prepared statements section. Sorry, closing. Thank you, Roger, Felix, Florian and Charles. This concludes our Q&A section. A replay of the conference call today will be available on our IR website within 24 hours from now. If you have any further questions, please feel free to reach out to the BizLink Investor Relations team. We thank you very much for joining today's call. You may now disconnect.