← 返回

2026.10.07 | TB | Battery | Panasonic Energy: In-Rack BBU

原标题：Panasonic Energy – In-rack BBU for AI Data Centres

#battery #datacenter

Third Bridge · 2026-10-07

Analyst-led • 1 Oct 2026

Companies: A373220 KR, 1211 HK, 6752 JP, A006400 KR, 300750 CN, GM US, UPS US, TSLA US

Ravinder Singh (RS) Director, Business Development at QuantumScape Corp

Ken Morozumi (KM) — Third Bridge Sector Analyst

Agenda

Key Insights

Introduction

KM [00:00:05]: Welcome to Third Bridge's Interview entitled Panasonic Energy – In-rack BBU for AI Data Centres. I'm Ken Morozumi, and I'm facilitating today's Interview. Pleased to welcome Mr Ravinder Singh, a former General Manager at Panasonic Energy.

Before we get started with today's Interview, would you please state I agree or I disagree to the following statement: You understand the definition of material non-public information and agree not to disclose any such information, or any other information which is confidential, during this Interview.

RS [00:00:41]: I agree not to disclose anything which is non-public.

KM [00:00:53]: Could you briefly introduce yourself, relating to Panasonic Energy?

RS [00:01:01]: Yes. I worked at Panasonic from January '24 to November 2025 as General Manager of Operations, actually launching the Kansas plant, the electrode manufacturing process there. My work was principally as a first employee to build the team, set up the operations, build all the systems working with Japan and all the new employees we hired, went through the commissioning of the plant and then launch and ramp-up, so pretty much from a greenfield facility to a fully functional plant. That's what I've done for almost two years.

Questions

00:02:03] Could you explain the current architecture of power backup in data centres, including facility-level UPS [uninterruptible power supply], BBUs [battery backup unit], power shelves and capacitor backup units?

RS [00:02:30]: As this is growing very fast, the technology is evolving very quickly, from grid power, then you come to UPS with a generator capacity already installed. Then from UPS, you go to BBU racks, power shelves and then there are some capacitors also involved. This whole architecture is principally based on how much latency or how much delay you have in terms of powering the GPUs, where actually you need a constant power and also have to manage the surges. The UPS role is, short-to-mid-term, that if the grid failure happens, it can sustain the power supply for some good 10-15 minutes, which is time enough for the generator backup to come up in place of the grid. From that UPS onwards, the BBU racks and power shelves… Power shelves are principally which are distributing the power to the GPUs or the processors. The BBU rack is there principally which is used to be in different rooms in the data centres, which is now transferring right by the side of the GPUs as the battery safety goes on and also, since the data centres are moving from 48V architecture to 800V, which actually brings a different kind of issue.

As we go forward, the requirement is to have safer battery systems and higher voltage to reduce the amount of current flowing through the wiring and then the copper losses because cost also plays a big role here and lessening the heat generation because the cooling cost of data centres is also very high. That's how principally UPS, BBUs and the capacitors, they work. Pretty much everything is moving towards the rack these days in my understanding.

00:05:00] How do you expect AI data centre and power architectures to evolve from the current systems towards 800 VDC [800-volt direct current]? Where will the battery physically sit in the architecture?

RS [00:05:28]: There is a very high need of putting batteries right adjacent to the GPUs. It's a multimillion-dollar set-up. The key there is how to get the safest possible battery to avoid any thermal runaway. Solid-state batteries or something close to that is actually the trend. Yes, there have been a lot of claims in the market, but there are manufacturers who are coming up with solutions. In my understanding, especially Chinese are pretty close and so are a couple of manufacturers in US. The whole idea is since you are bringing these things closer to the processors, the compactness of the system and the operational bandwidth at which these batteries have to work, that has to go to the higher temperature levels. That's where non-liquid platforms or solid-state platforms become very, very helpful and even the format of the battery. Today, cylindrical batteries, which are very prevalent, are available, but tomorrow, when you bring to compactness because the space near the GPUs is obviously limited, and since the whole purpose of the system is the compute, you want to have as much space available for compute than batteries. Obviously, to support that big compute for a couple of minutes in case of a power failure or fluctuation, using these batteries or power capacitors, you need to have compactness and safety as prime factors.

With this 800V architecture, the whole idea is to reduce the wiring, which obviously takes a lot of space, and then the heat generation. The higher the voltage, you will use lesser current to draw the same amount of power, and since data centres are very prone to fluctuations, the surges actually generate a lot of heat. An 800V architecture would be very helpful because the current [sic] for power is I square R. The rate at which if you draw current, the amount of heat generation is exponential. Reducing the amount of current for the same requirement of power with the higher voltage is the trend. My understanding is that the data centre companies have figured that out rather well.

00:08:29] Given that AI racks need huge bursts of powers, how are US and China-based developers overcoming the high internal resistance in SSB [solid-state batteries]?

RS [00:08:50]: There are chemistries, the interfaces which have evolved. In my understanding, the gel-type batteries, which everybody knows exist today, and now the layers through which you need to get the current flowing, those interfacial chemistries are evolving. In my understanding, it's a matter of a couple of years or rather lesser than, obviously, even a couple of years when you would get them very close to what a normal lithium-ion battery with the electrolytes would be doing. I would say there's a limit to which I can discuss this subject being still in the industry, but in my understanding, it is all well-known how these chemistries are working. Most of the players what I have seen in the solid-state battery industry have shown promising results. Here, there is another thing which is important is when it comes to the data centre applications, the power requirements are for a very short period of time. It's principally the surges or the impulses of a couple of minutes. The resistance… When you have to have an electric vehicle, a power requirement where energy capacity is very important, here, the case is different. You can find a good compromise to overcome those challenges while the chemistries further evolve.

00:10:52] What share of new AI deployments do you expect to adopt 800 VDC over the next 3-5 years? What are the main barriers to wider adoption?

RS [00:11:25]: I would say the key here for that adoption is principally the battery quality, the safety of the batteries, as we just discussed, solid state being one option or batteries with higher C-rating, the power cells, because you also want to reduce the types of power systems you integrate into the data centres. Today, for every kind of problem, there is a solution. Many people bring capacitors along with the batteries to take care of the higher-level surges, which the batteries, because their C-rating, there's a limitation to it. For large surges, you want to have a capacitor while the battery can only operate at 4C or 5C or whatever the best lithium-ion batteries can do today. Power cells also, but then there is the limitation on how much energy you can draw from them.

Depending upon what the system designer people, the hyperscalers would ask for, the battery systems will be available. I would say having complex systems with all these different kinds of capacitors and the batteries and then the switching between them, these are all losses because your power electronics just keep increasing. The simpler solution is a high-density, high C-rated battery. As that evolves, the 800 architecture would take over, I would say, at a very high rate. Given the power requirements in the data centres, the people are talking about megawatts of data centre racks, to achieve that, you have to create the space with all these extra power electronics because everything electronic will generate heat and then you have to have bigger cooling systems. Simplified version of all this is, as the batteries evolve, the data centres will be very quickly adopting it to this architecture.

KM [00:14:02]: So what percentage of the newly built data centre capacity over the next 3-5 years do you expect to deploy 800 VDC vs stay on 54?

RS [00:14:31]: In my understanding, everybody is claiming to get the solid-state batteries into the market in 2029-30 timeframe. When you say five years, I would imagine, if not more, maybe one-third of the systems will already be on 800V by 2030. From that point onwards, I would say it's a matter of a couple of years. I would say about two-thirds of the data centres by 2031-32 should be into that type of architecture.

KM [00:15:20]: What about in 2-3 years? Could it be around 10% or 20% in 2028?

RS [00:15:40]: Yes, very much feasible. Since the concept has been there for now quite some time and people are working on this, the architecture has a lot of advantages. You may not get everything with that architecture what you anticipate, but I think the evolution will take its time. First thing is even if you start saving on, let's say, the copper losses or the delays, if you bring the systems closer or your cooling systems are becoming shorter or lesser power consuming, that itself is a benefit. Bringing that architecture faster, I would say, is beneficial in every possible way. Only thing is you will not draw its full effectiveness, which is where I was mentioning one-third to two-thirds. That will take a bit of time. Your estimate 10-15%, I would say, yes, by 2028 should be very much possible.

00:16:58] How would you evaluate the current Panasonic Energy DC battery product portfolio and the business strategy in the next 2-3 years?

RS [00:17:33]: I've seen, I think even this year, there's been a lot of announcement. They seemingly command a very significant market share today. I would say the pivot or the centre point of this is today… Panasonic is one of the most successful companies in the US, with the help of Tesla, who has spent more than a decade building cylindrical batteries. Making a cylindrical battery, the maturity of technology is very good at Panasonic, maybe better than most of the players here. Obviously, after 10-15 years of battery production here, the reliability obviously is proven in the market, may it be through EVs, but that's how it is. Those cells now are worthy of being used anywhere or anywhere new technology evolution is happening, and data centre happens to be a significantly large market, just like the EV buzz, which started a couple of years ago. Since they have a significant capacity in US, I would say, with the Kansas and the Nevada facilities, should be 60-70GW capacity available. With Tesla not necessarily growing at the pace what they expected, all that, what you have, I would imagine, in Kansas or a significant portion of it should be available for data centres, which is what the recent strategy announcements have indicated.

The only thing is there are two constraints with it. One is the cells are cylindrical, so obviously, the volumetric efficiency of the battery pack or the module which you will put in data centre may not be top of the line. Then secondly, the C-rating of the cells. The cells originally have been designed for cars and 3-4C rating is more than enough for an electric car. Here, we are talking about the short pulses of a few minutes, which means the same cell, which is, let's say, 15-20 watt-hour cell, you will end up drawing power from it within 2-3 minutes, which means you're talking about 20-30C of rating, which these cells, to my mind, unless they are redesigned completely, you can't draw that enough power. From that standpoint, even though the capacity is there, I would imagine there will be some significant design changes happening to capture not the current market, but the market in 2-3 years because that is where the technology evolution is happening, and that is where the higher voltage architecture is coming. In the announcements, there is a talk about developing or converting to new cells, and I would imagine that's what Panasonic possibly is doing today.

00:21:23] How has Panasonic Energy's data centre battery product business evolved? Could you outline the company's strategy for expanding the battery product business? What strengths and weaknesses did you observe during your tenure?

RS [00:22:12]: I want to be very clear. I worked in Panasonic in the capacity of EV business. During that time, the data centre development, if anything, must have been happening in Japan, but not in US where I had been. It will be more of, let's say, speculating about what Japan would have been doing, which was not actually the purpose in Kansas. We were all focused on the EV business and which is doing fine. The success, I would say Panasonic, they've been the largest battery manufacturer in Japan, very successful. I would imagine in Japan, there's a lot of different battery formats they supply. The capacities are not necessarily largely to the gigawatt scales today, limited ones, because this data centre hype is not very old. People were using what is available in the market. A player like Panasonic, when they have batteries here available in US, lithium-ion batteries, anybody would design their system to fit to those batteries' performance levels.

Since the LGs or Samsungs or SKs of the world, they came rather in the beginning of this decade, the '20s while Panasonic has been there from 2013-14 timeframe, so obviously, there was a long catch the others would have played. The technology maturity, which happened with the EV support, those cells automatically got a default path for Panasonic into data centres. That gave an obvious lead. Then the backup from Japan because the energy business has been very centred around Tesla or US operations. The data centre, there is not any specific requirement until recently for localisation. Obviously, these companies were also not really growing at the pace what they are thinking now. Those days, batteries must have been coming from Japan, so would have been the Koreans and anybody else. Interesting thing with Panasonic is because it's a big conglomerate, many, many companies, and they have been an electronics company, all the power electronics, all battery manufacturing, obviously, even the software, this is all part of Panasonic Group's strengths.

It is a matter of integrating these strengths, which actually was visible in the recent announcements also where Panasonic is putting assembly lines in Mexico, where they will assemble all these BBU racks or units with the cells coming from Japan or from Kansas. Then all the power electronics integration and the battery management system, BMS, will be integrated down there. Panasonic, as far as I know, I think there is a company called Blue Yonder, which was acquired a couple of years ago. They are very good at software. I think that capacity also would be available. Not necessarily with the purpose with which now they are growing. The battery capacity in US played a big role. The market reliability helped to get them to a stature of where all the data centres who wanted to grow faster, they could use whatever is available. That's how, in my mind, the business has grown.

00:26:49] Panasonic Energy's 2026 integrated report mentioned that net sales for data centre battery products would be threefold in FY29. How do you evaluate this net sales trajectory? Do you think it is a realistic estimate?

RS [00:28:06]: I would say yes. I'll make some assessment here. The thing is the report mentions about supply from Japan and then from Kansas. The Kansas plant, which started production only last year, obviously, you can't reach 30GW in six months. It takes time. It's a process of about 1.5-2 years to get to the full capacity. We keep on adding lines and then capacity doubles, triples and then eventually, you hit where you are supposed to be. Today, they might be half the capacity, so the plant might be running at 15GW. You still have 15GW available and you can flip it to whichever direction you want to go. The plant was conceived for EVs, but now converting it to data centre-bound batteries, which obviously will be different in design because of the requirement which we talked earlier, that the depletion, the discharge rate of the batteries have to be very high, which means the chemistries have to be tweaked.

The announcement did mention something about 2027 or 2029 timeframes, which tells you that they are taking all the time available to convert to those type of batteries with the 15GW capacity available in Kansas at least, if you see from the timeline today onwards. It's clear that the EV business is not growing in US… It's slow, Tesla is not growing, which means the capacity is available. Diverting that, I think, is a pretty good strategy. It's just a matter of time. They will secure that kind of capacity.

00:30:11] How much of the manufacturing capacity that you mentioned is transferable from EV to data centres?

RS [00:30:33]: This announcement was in mid-year this year. By that time, I would say Kansas possibly is operating only at half its capacity. When you put a new line, yes, you can redesign to whatever the format is going to be. Because whatever is in the media, it's all cylindrical batteries. There are only specific formats, 2170, 1865 or 4680. In my understanding, whatever Panasonic will put into the data centres from US side would most likely be 1865. There is a specific reason. Because you want to have a high discharge rate, which means you need to have a thinner cathode to draw that power quickly, a thinner cathode would mean you don't need that big diameter of the cell, you can live with a slimmer cell, and 1865 in my understanding is the slimmest battery prevalent in the market. Maybe they will convert it into that format. It should be possible. You can convert all the 15GW without any issue.

00:32:08] Does the factory transformation we mentioned mean Panasonic Energy has strengths vs others, or does it not imply a future competitive edge for the company against others?

RS [00:32:30]: No, I think definitely, they will have a competitive advantage. The battery business, there are government incentives today. Everybody knows that there is a 45X, the IRA benefits credits. Means as long as you're able to produce batteries in US, you get tax credits. Most of the players struggle with the fixed cost, the CAPEX which it takes to get to these plants. For that part, Panasonic has already taken the bullet many, many years ago. All that is now existing. For any new player who wants to get into this business, the initial costs are going to kill the business case. For Panasonic, that is a big advantage. The whole structure, everything is there. It's just the machinery you change or tweak to the new format and then you're good to go. I would imagine for Panasonic, they are in a very good position as compared to others.

00:33:40] In the 2026 integrated report, Panasonic Energy's CTO mentioned that he will target North America local procurement of above 50% for power supply components by FY29. Would that materially improve delivery and supply flexibility? What could be the remaining constraint even with sufficient sales capacity?

RS [00:34:30]: I think here, maybe the perspective for this question has to be the end product, not just the cells. In the cell, the main thing is the graphite on the anode and the NCMA chemistry what Panasonic has announced. They will stay with NCM. They will not go to LFPs. That powder, the supply chain, I would say it will stay the way it is today. I don't think anybody is rushing to localise that. The end product for data centre is all the power electronics, the BMS and everything altogether put into a module or a power shelf for BBU rack fitting system. That is where I would say, not only Panasonic, but people like Samsungs and LGs also, who have traditionally been electronics companies having very deep supply chains and have been existing in US for quite some time, I think they will flex their powers here because all these businesses are already set up. Now, it's a matter of changing the part numbers or whatever the little changes are required for something to be made for data centres, that they will do.

I actually was very intrigued with the announcement of putting these production lines in Mexico. Today, whatever the FTA rules are there with US and Mexico, although the administration is trying to control more, but at least given at a point of time today, that strategy could be very interesting that you send cells there and then get everything structured all from there back into US because Panasonic, in my understanding, would have a lot of electronics operations in Mexico. The strategy on paper looks very promising. The logistics, obviously, will help. The localisation content, if the CTO has mentioned 50% in US, I would say that should be very much possible given the size and scale of the business Panasonic has in US, all the business, all the conglomerate, all the companies they have here.

00:37:10] Panasonic reports approximately 80% market share in distributed power system for data centres. How should we interpret the figures when assessing its position in AI BBUs?

RS [00:37:47]: I'd be speculating a little bit because that's a very tall claim. That 80% figure actually is intriguing in certain ways. At the time of announcement, maybe that's what the reality is. Today, they are 80%. If you see three years down the road, will that be sustainable could be a question. There are two things happening. One is, obviously, the competition from other companies that they are also preparing in parallel, but the second part is the technology evolution. The sheer time it takes from conceiving a cell design to then fully validating it and then setting up the process of it, it's quite long. It could take easily 2-3 years before you can come up with a new battery cell launch and get to some meaningful volumes.

With the cells, what are actually originally designed for EV systems, converting them into power cell type of architecture where you manage all these power surges, will it keep pace with how the industry is looking at it? That, I would say, could be a little bit questionable. Yes, on paper, it looks all fine, but in my mind, it's a bit of stagnant state of evolution or a very linear state of evolution, whereas the way I see the customers evolving… A model cycle for Nvidia or Meta or somebody else is almost one year or 1.5 years, they want a different architecture for the data centre. The question would be, these cell designs and chemistries being conceived today, will it be able to take three years down the road that competition? That is yet to be seen.

KM [00:40:11]: What does this figure represent – 80% share in power system for data centres? What's the exact scope of the number? Does it measure its cell shipments, battery module, BBUs, racks or completed power shelves, including power electronics?

RS [00:40:43]: I think it's built-in shelves, and that's possibly… Because there is an asterisk mark in here when I see the report, where they have this bubble (ph 40.55), overwhelming market share and that's distributed power systems for DCs, which suggests me it could be maybe on the UPS side. It doesn't specifically say that it is in the rack architecture. Maybe if you take one segment of the whole architecture, there, they may be because these cells are not necessarily designed for impulsive power surges, but more of a stable supply with energy capacity what they have. I would see it as more of, I would imagine, on the UPS side. That's where probably they are. Even on that, that is today's state. It's not talking about future state. Future state is just an ambition, and that's where I mentioned that that's questionable.

00:42:10] You have questions around how sustainable Panasonic Energy's market share is for the next 2-3 years. What limitations do you think the company has vs others? How would that impact the 80% market share?

RS [00:42:46]: I will probably base my assumption on what happened on the EV business side. At some point, even before I ever worked for Panasonic, Tesla relationship was there. When Tesla brought 4680, Panasonic also had 4680. They have those cells available. That business never grew, while Tesla ended up building their own ecosystem and own batteries. What is the real reason, I don't know, but as a sole supplier at some point, Panasonic having the biggest market share, and not being able to address the customers' requirement tells you something about the adaptability speed of the organisation, how quickly you are able to pivot and then answer the need. Secondly, even in the BESS systems, where LGs and Samsungs got entry into Tesla battery business with LFP chemistries, Panasonic's decision of not going into LFP chemistry itself is… Okay, maybe they want to stay in the premium segment, but then that also tells you something about the adaptability and trying to secure your customer.

When you're applying the same model here, 800V architecture is a pretty old story. In automotive, it was done many, many years ago by Tesla, five years, if I remember well, when this Cybertruck and others were even conceived. The data centre 800V has been only very recently, a couple of years, 2-3 years maximum. For Panasonic to now build the cells to meet that architecture once again comes to how quickly you can do it. Anybody new who is getting into this business now, obviously, will work with what is in the future, that vision, whereas for a player like Panasonic or even some other established big brands, unlearning is also a bit of a challenge because the bigger the company, the bigger the system, the bigger the business size is, harder it is to adapt very quickly.

One possible challenge could be that today, okay, you can put your bets with what you know about the market, but given the history of evolution, one more example I would say, Panasonic used to be the largest producer of batteries in the world after Sony when Sanyo, everything was broken down and then Panasonic had a big market share, but within the last, what, 10 years, all those IPs and I imagine the technology have been taken over by Chinese in such a speed that Panasonic is now a niche player in certain segments, certain markets. That possibly, unless the CTO, Mr Watanabe, I guess that's the gentleman, has changed the organisation, the way things happen, you could live with an element of doubt.

00:46:40] Based on your estimation for Panasonic Energy's weaknesses, in what areas could the company be substituted by other competitors? What could the 80% market share number look like in the next 3-5 years?

RS [00:47:11]: I would say sustaining 80% in my mind would be very challenging given the Taiwanese players, there are so many data centre-focused players trying to win business. Panasonic will eventually find a sweet spot, maybe 30-40%-ish level eventually in 3-5 years. Short term, yes, I would say short term, the business is going to grow because they are the most ready company in terms of something which can respond to the customer needs. They will grow… 2028-29, I think their ambitions will be fine, but long term, the market share will reduce. I would say the challenges are… One thing I also feel that the battery format, the cylindrical battery format alone, will it sustain the requirements in future? I would see a bit of a question there, unless Panasonic will bring some different battery formats for which obviously the plants in US are not ready, so maybe from Japan. That's one. Second is the chemistries. Panasonic is very clear about NMC, that's what they're going to use.

While the others may bring… Recently, in the market, a player like GM now. GM is into this battery storage business, just like Ford. GM is big on LMRs. They have announced manganese-rich lithium batteries. That chemistry is somewhere between NMC and LFP, somewhere in between there. If you can work on the packaging, you can come with an innovative format, like pouch cells, which can then beat the cylindrical battery volumetric efficiency, then possibly you can compete with an LMR to an NMC and then be at the same level where Panasonic is. Those kind of evolutions could be a potential threat. Once again, I would say it's hard to forecast long term, but short to mid-term, Panasonic will definitely win, but after that, I would say it will be a fair competition. People will catch up.

KM [00:50:01]: To summarise, its market share will be down to around 30-40% in 3-5 years?

RS [00:50:13]: I might want to correct a bit. This big dip, three years may not be possible. Five, yes, you can imagine that. Take it mid- to long term. That's where it is going to stabilise.

00:50:34] For what customers or applications do you think Panasonic Energy's competitive edge will sustain in 3-5 years?

RS [00:50:58]: One is the footprint in US, that's significant. Chinese still have these policy issues. It is hard for them to compete here directly. The policies, I would say, being local, obviously, and then as you mentioned, the localisation of all the power electronics and everything right here in the United States is one big advantage they have, where the Taiwanese and Chinese will take time to catch up to. Secondly, not all the data centres are planned to be installed here only in US. There is development in Europe. There is development in rest of the world also. The same Nvidia's data centre will be sitting probably somewhere in Europe or Google or Meta. The competition can get an advantage there. They can win significant business in those markets where Panasonic may not be that strong yet. That's why the short term, where the footprint is, yes, you have a win, but when you look at the global scale, this data, for instance, is it all centred in US, 80%? I'm sorry, I haven't really looked deeper into it, but if you look at it from a global standpoint, I would say that's US-centred advantage, definitely Panasonic has that, but from a global standpoint, may not be that big. I think it's something to wait and watch. For now, they have capacity, so I'm pretty sure they should be able to win the immediate market shares.

00:53:04] How should we estimate the revenue composition of Panasonic Energy's data centre battery business across bare cells, modules, power shelves, installation and after-service?

RS [00:53:43]: That's quite a large, wide-spectrum question. Starting from battery cell, right at that point, I personally am not aware how the transactions with the data centre customers are happening. Is it happening in dollars per kilowatt or dollar per rack or dollar per something? The reason for that is for normal EV business or normal battery storage system business, the capacity is seen in a kilowatt-hour because principally, you are working with 0.5C or 1C rating of the cells and then you are wanting one hour of availability because the energy capacity of the cell is important. When it comes to data centre, the C-rating, the power of the cell is way more important than the capacity. The same cell, which in an EV, it lasts one hour for X watts, in data centre, you are probably depleting it in 3-5 minutes. A three-minute depletion would mean that it can work worth 20 cells, which means for a customer of data centre, if the unit of a sale is a cell, how much does a cell cost? Let's say, X dollars. They will negotiate based on, "Okay, if 20 cells get me X kilowatt-hour, I would rather buy one and pay the price of that exact one cell to limit the cost" because they will never use it more than 3-5 minutes by design, where the battery is being used.

From that standpoint, it will make more sense to do the transaction at their system level, at a rack level, dollar per. For a company like Panasonic, I would say they will obviously have a strategic angle, which is they are an integrator, because they can make these things because of all their other resources they have. They have other companies to help them. To defend their price points and I would say, the competitiveness, I would say they will transact at that level. For a data centre customer, it's very important to have one guy who takes responsibility of all the integration on their side and pay a premium to that person.

Strategically, Panasonic, I think, is in a strong position there. They can bring the full system and take responsibility of whatever is in the back end and still defend the price. Now, I'm not aware about what is the price for the power electronics or the BMS and all that. That is something, I would say… Maybe some data points from the general industry trend where people sell power electronics stuff. That mathematics has to be done. Cells to me, I don't think it's that big a money-making business. The value is in the integration, the value is in the system level, and that's where I think Panasonic is equipped to do it. That's why all that Mexico announcements of making all these systems. That's part of the bigger strategy, which is, I would say, a pretty smart move.

00:57:34] What is your outlook for Panasonic Energy's battery business, considering the move towards 800 VDC?

RS [00:58:17]: I personally like the strategy how they have positioned themselves to capture this market. The ambitions may be too aggressive, but that has to be seen. Time will tell. The adaptability, I would say, is the key here, as long as they are able to mobilise all these resources at the same time as, actually, the strategy dictates from the President, Mr Kusumi. I would say, at this point, it's more of the foundations of the business, which is exercised with a very strict regimen and goal. I think they will grow, they will grow significantly with the business. The localisation, as mentioned by the CTO, if they have announced 50%, yes, they should definitely make sure that happens.

The risk part is the policy changes. If there are some significant policy changes, let's say, tomorrow, the administration put tariffs on Mexico, that could play a bit of a role, could be problematic to directly hit the competitiveness. Otherwise, I would say the readiness standpoint, in my mind, Panasonic is pretty ready, as compared to the competition, to approach these markets with an integrated capacity. Against the Chinese or Taiwanese competition, being local and being the most established in terms of supplying batteries in the US market, I would say they will cash on their market position today. Short to mid-term, I would say the strategy is workable. It will deliver good results. For long term, I would say the time is going to tell.

KM [01:00:38]: We're going to end this Interview. Thank you so much for joining this Interview today. See you next time.

RS [01:00:49]: Thank you very much. Appreciate it. Bye.

Transcription ends at 01:00:53 of the recorded material