Netweb Technologies Indias Ai Proxy The Spotlight Ft Vinay Menon

TITLE: Netweb Technologies: India’s AI Proxy | The Spotlight ft. Vinay Menon CHANNEL: Krish Kothari DATE: 2026-06-19 ---TRANSCRIPT--- In early 2000s, they started with a partnership with, [music] you know, somebody like Nvidia, Intel, AMD, these kind of chip manufacturers in the US. And that is what gave them a lead. And this is a stock which I believe can be a Nifty 50 stock one day, maybe in the next decade or so.

That is Vinay Menon of Monarch Networth Capital. Is Netweb truly a proxy play for Nvidia? In this episode of the Krishna Kothari Show, Vinay and I discuss Netweb Technologies, one of the few Indian companies considered a beneficiary of the AI boom. Revenue has gone up like 15x. What in particular has enabled this kind of growth? We discuss the business model, competitive strengths, and challenges facing Netweb, as well as the all-important question of valuation. A lot of people say that India doesn’t have any innovation or R&D. This is one of those companies who has done very well. And, you know, they need [music] to get that recognition. And I’m I’m happy that they’re getting it right. We hope you find the conversation insightful. Don’t forget to like and subscribe. Thanks for watching. Vinay, welcome to the podcast. It’s wonderful having you on. Hi, thank you so much for inviting me. Is Netweb truly a proxy play for Nvidia in India? So, I won’t point it towards just Nvidia. I would say Netweb is a proxy play on AI in in a in general. See, because Nvidia is just one of the design, you know, partnerships. Netweb looks at servers in a big way. So, first I’ll explain what Netweb does. So, Netweb manufactures servers in India, made in India, and they’ve utilized the Make in India scheme which the government bought in 2014. Now, Netweb has been in this business for more than 30 years now, and they started out with servers at the time where, you know, the computer boom was big, you know, late ’90s. Everybody wanted to set up a desktop or sell a desktop. India finally saw computers. It was a big revolution which happened. and that is the time where, you know, Sanjay Ji who’s the promoter and the managing director of the company, he thought that, you know, why not look at servers because he felt that the future is compute, you know, and he had that vision long back and the computer space was overcrowded, you know, because India didn’t have any OEMs who manufacture computers. So, in India, it was more about reselling computers, being a distributor of computers and a lot of people do it well there also, but Sanjay Ji said that, you know, this could be the future which is compute and in early 2000s, they started with the partnership with, you know, somebody like Nvidia, Intel, AMD, these kind of chip manufacturers in the US and that is what gave them a lead. Now, those partnerships are now starting to really show fruit. So, today if you see, there is, in terms of the design partnership, there is nobody who matches Netweb in terms of design partnership in India as an OEM on the server side and that is what their USP has been. So, I would say rather than this Nvidia being a play, I would say any of the chip manufacturers there who are making chips which go into servers because they make chips for other things also, but on the server part, if you want to play any of those, Netweb is the best play in the Indian market for that. You know, I was looking at the financials of the company and from ‘21 to ‘26 in those 5 years, revenue has gone up like 15x. Can you effectively just explain what are the different ways in which your company makes money, what are the different revenue streams and what in particular has enabled this kind of growth? See, Netweb has three revenue streams if you have to put it down to it. The first one is HPC which is high performance computer. There, basically, you’re looking at somewhere, you know, like you look at supercomputers. Today, if you want to build a supercomputer, you need a lot of servers. So, for example, the fastest supercomputer of India which is Airawat and it’s 75th globally has been delivered by Netweb and that took like a few hundred crores. So, that is India’s fastest supercomputer till date. And it has listed in the top 500 for the last 2-3 years. Now, that is the first division, that is the original division which started with Netweb. So, Netweb started with the idea of making servers for supercomputers and giving it to research institutions, to government institutions. And that was the earlier like 2000 to 2010 was a pure focus on that. And that’s why a lot of the, you know, demand was from government and all these research institutions. So, 80-90% of revenue came from there. Now, post-COVID what happened is they decided that it’s time to go into enterprise also because that is somewhere where they saw a lot of boom. And they felt that, you know, a lot of that there was a whole shift towards on-prem that time that you want to be on-prem, you want to set up our own servers so that we can actually control our data. And that’s where server boom happened post-COVID, you know, and that’s where you saw a lot of demand coming towards Netweb and that’s where they started scaling up their operations as well. So, enterprise was the next trigger. So, from government they went to enterprise. Today, enterprise-government is 50-50 which took a lot of effort to reach there. Thirdly, today if you look post-COVID what has happened is, you know, everybody was working from home. Hyperscalers really did well. So, you had a lot of data centers coming up in India globally. Today, we are just at 1.5 gigawatt, but this is going to go to five. So, you can just understand the kind of demand which is still left despite growing 15x. So, it played all of these cycles very well. And the company was well prepared for all these cycles. And that’s what I like to reiterate my point, the vision of Sanjay Ji was, you know, quite good and he anticipated these things coming. And then, you know, in ‘22 ChatGPT launched and the world changed forever, you know, it it never went back. Now, when you have these LLM models who are giving you, you know, brilliant responses on all these prompts which we put, obviously you need a lot of compute for that. You need fast servers, you need GPU-led servers. So, servers were uh typically a boring business because they were CPU servers, you know, because they used to just do the compute and give you whatever data you wanted. And that was the business for a long 15-20 years. When GPUs started coming into servers, you know, that that’s the whole Nvidia wave which came in, where we came to know that a GPU can not only just render graphics, it can actually also do matrix calculations, it can do a lot more. So, that’s where the whole game changed. When GPU servers came in, you needed that for everything, and especially for AI. So, today what happened is there was a big replacement cycle also. People who only had CPU servers shifted to GPU servers, and the prices of GPU servers today have like, you know, skyrocketed because of the technicality. So, earlier this was more of a quantity game, a volume game, where you know, okay, a company needs six servers, 10 servers, 100 servers, this is the requirement. Today it’s very different. Today it’s like, why do you need servers? What’s your requirement? Are you doing some research? Are you doing some pharma-related research? Are you doing something for markets? Are you doing high-frequency trading? Because everything has a very different requirement in terms of servers. So, this trend, and even high-frequency trading was another area that captured because last 5 years, you know, markets saw such a boom. You know, a lot of HFTs are set up in India today. So, that has also led to, you know, Netweb doing very well because those servers are also given by So, all of this they were able to catch the wave properly. So, is it possible that Jane Street is a client of Netweb? Not officially, but it’s they’re not disclosed anywhere. But, there are a couple of good HFTs who they’ve disclosed as their clients. So, they are very active in that space for sure. What I’m understanding from you is that Netweb at its core, what it does is it customizes its own products for as per the requirements of an individual client. Yeah. And is that the real value proposition of the company? So, the biggest value proposition and the USP I would say is that you know like the other large server manufacturer. So I’ll tell you one thing in volume terms Netweb is not among the biggest in India. It is still dominated by global OEMs like an HP, [clears throat] a Dell, a Lenovo, Asus. They all are part of the market. Netweb would be like 5 to 6% roughly market share in India or even it’s not an official figure. It’s something which you can kind of calculate in terms of the sales and the total market size. So they’re not a volume player and they’re not chasing volumes also. They are more into solutions. So they want to create the right solution. And if you see in 2021 their margins double from 5 to 10. And the reason was simple is because they started selling software as well. Now typically a server manufacturer will only sell the hardware and go. That’s the typical agreement. You have other players you know who are software players who will sell you the software separately. So you have separate OEMs who are selling you the software and then you bundle it and use it as a client. Now what Netweb decided in early late 2019 2020 is that you know we can create a solution where the problem of two three vendors comes down to this one vendor which is us. And they put money in R&D. They set up the whole software part of it and now they have at least five to six different softwares which they are selling along with the hardware. Now the beauty of that is it’s it’s similar to the Nvidia way. Today Nvidia wants to own the whole stack. You buy the GPU from Nvidia. You buy the CPU from Nvidia. You run it on CUDA which is their software stack. Something similar Netweb is also doing. So today when you buy Netweb server you’re getting the software along and that software gets you the max ROI from the hardware that they’re selling. That is something unique which Netweb is doing which no other OEM has done. You know there’s a huge boom because of this AI thing you know you see Korean markets doing so well. You see DRAM and SSD prices have like 3x 4x in the last one year, is because there is no supply of these components. Now, what happens in that is that for a hardware, pure hardware manufacturer, you can’t really pass on anything cuz you know it’s it’s a little difficult. But when you have hardware plus software, it gives you that leverage that okay, if I want something, I can kind of give you a better deal to a larger OEM and that’s where I can get your business. So, that is also something which is playing in their field. And having that sort of hybrid model would also help with recovering revenue, right? So, they don’t have a SaaS model typically because they are not trying to sell the software separately. So, that’s that’s not the idea. The idea is still to be a server manufacturer and kind of give you the best solution. Obviously, there are there are upgrades and softwares which keep coming in. Softwares are bundled for like 3 years, 5 years typically on how you want to use it. And the best part is the pricing with software and hardware is at par to any other OEM who’s selling hardware and software separately. It’s not like they are not premium or they’re trying to under They feel that the product is good enough. We will charge the same price. And then that’s how they’re getting those kind of deals. Mention some of the larger global players and who they are competing with on a day-to-day basis. Yeah. On what basis do they compete with them? Is it just that they are competing on price? See, the larger OEMs are volume players. So, an HP and Dell are looking for volume. So, they’ll not go after maybe 100, 200, 300 server kind of deal. They’ll look at 1,000, 15,000. So, that is the kind of game they want to play. Plus, they don’t have the software part of it. So, they are more interested in get you know, selling the hardware and getting out of it. As simple as that. So, they will continue to do well. Again, I’ll tell you that because there is enough volume for everybody. This industry is growing at 10, 12% CAGR. And to give you a context, today India has 1/10 of the server capacity as US. And you can you know how much data is consumed today in India. So if the data has to be brought back to India and maintained in India, the server requirement is going to be huge. So everybody will do well. I’m not trying to say Netweb will outdo everybody. There is enough demand for everybody to grow. That will continue. Secondly, because of the solution part of it, they have gone to a lot more niche clients who are looking for, you know, things like, okay, I’m doing, for example, I’m doing AI research. I have a sample of say 1 lakh you know, clients like, you know, patients. I’m you know, I’m doing research for a hospital, for example. Now, I need X amount of servers for these which meet my exact requirement for my research. This is something where a Netweb will do better because they will take the time, find out how it is, and give you that solution. And that is their play. And that is somewhere maybe a larger OEM might not want to cater because it does not move the needle for them. And that is why Netweb will win there. And then Make in India obviously has been a huge addition to Today, we are seeing, you know, like from a defense perspective, today the kind of global tensions are there, it is best, I feel, for India also to own its own software and hardware because these things can be controlled from anywhere in the world. So when that happens, you would always want that. And I think that it’s clear from the government’s initiative that we want our own servers, we want to make our own chips, everything. Like today, you know, like from the AI mission perspective, sovereignty, this it’s a start, but till we have our own CPUs, obviously it’s not completely sovereign because till now the technology is with them. Maybe we are 10-15 years away from that. But in servers, we are close to them. And that’s where we can still compete. So I think that angle will always work in Netweb’s favor. Isn’t that a really core challenge for a company like like Netweb, which is that they are effectively at the mercy of an AMD or Nvidia. Yeah. Where if tomorrow morning there’s a supply constraint globally and Nvidia just puts its hands up and say sorry, you know, we can’t have to supply someone else, we don’t have the capacity. Yeah. Their business comes to a grinding halt. Is that risk is that sort of sword always dangling over them? So in the 25-30 years it has never happened where they’ve not gotten supply. Now obviously if there is an issue at any of these chip manufacturers Yeah. all the server manufacturers will be impacted. So I’ll give an example that today if AMD is coming out with a new chip, okay? Now obviously, you know, the server is also designed as per that chip. So today Netweb is at the same level as HP Dell that if for example after the chip is you know, put out to the market in 3 months if HP Dell can come out with servers on that chip, Netweb can also come at 3 months or sooner. And that is the capability. Now what happens in this industry from a chip perspective? See because these chips are made in Taiwan, TSMC is the main chip manufacturer. Design comes from Nvidia. The architecture comes from ARM which sits out of UK. So it’s it’s a proper channel. Now anyone impacted in the channel the whole channel goes for a toss. Nobody is creating their own thing. Everybody is doing their part of the drill. So that means that if anybody is impacted all OEMs are impacted. So you’ll see server prices go up for everybody. So it won’t be a Netweb specific issue. It will be an industry issue. Can you just walk me through the profit and loss statement of the company and explain what are the key line items that you see that are contributing to this growth in terms of whether it’s specific products or services that they provide? Yeah. And also if you can explain how that’s translated to improvements in the balance sheet and cash flow of the company. Yeah. So I’ll take you from step by step. So in 2021 as I told you margins, you know, doubled. Now that’s when software started to come as an element and and that’s where you saw a big transition from just a pure play hardware manufacturer doing 5%, you know, EBITDA margins to a hardware plus software play. So, that that was the first transition which came in. Secondly, they have three segments. One is HPC, which is I told you more on the supercomputer side, uh more from a, you know, research, from a government perspective. That that’s the segment which is going there. They got into cloud because they realized that a lot of uh on-prem projects are happening, especially in, uh you know, BFSI and BFCs, you’re looking more from a perspective that people want to own the data, keep the data in on-prem. That was something which is another thing. Plus, Netweb has, you know, they they have a claim that, you know, the cost of doing business in terms of, you know, buying the servers Mhm. in the long term is cheaper than going to any of the hyperscalers where you’re doing the annual subscription. And that is something which they have proven to a lot of clients who have then decided to go with that kind of approach where they do the capex. In the long term, the ROIs are much better. So, that was something which worked for them on the cloud part. The third is the AI part, which post-2022, you saw that huge boom coming. And that’s where margins are also a little better because there you, you know, you’re making very specific solutions. You’re using GPUs which are a lot high-end. So, the servers there will be like, you know, 1 and 1/2, 2 crores, north of that also. You know, your typical server starts at 10, 15 lakhs. So, that’s the difference in the kind of server which is there. It will have probably eight GPUs, and, you know, it will have liquid cooling inside. Top end of things, you know, so so that those come with better margins also. So, that is where also they saw. Plus, they introduced more softwares over the last three, four years. So, the more software that comes, the better your margins become because that’s all one-time R&D cost. Mhm. And then that keeps flowing to the margin. So, the margin profile changed there. Secondly, on the growth, they’ve they’ve grown at 35, 40% CAGR, you know, and more also. This is what they always guide, but they always outperform that number. It’s because demand is there, you know, like today the demand like if I had to give an example to put the demand, we are at 1 and 1/2, you know, gigawatt of capacity data center. We plan to go to 5 gigawatt by 2030. You will need 5 to 6 million servers, which is 50 to 60 lakh servers. And today the server market in India is 33 to 35,000 crores. So, if you were to add 60 lakh servers to this, it’s it’s going to go 3 to 4x at least in terms of the size. And that is going to create more and that is happening already. And that’s where you see the kind of demand they’ve seen on the top line. So, the order books have grown plus the enterprise side has also helped the order book grow faster. That really did a big thing for them. AI mission, which is the last 2 year phenomenon, they won deals, you know, north of 2200 crores, which is the highest among the other competitors, which has also helped them, you know, grow from say sub 2-300 crores to 2000 crore plus in 2026. So, that that’s the top line growth journey. How this P&L has impacted on the balance sheet cash flow, I can tell you. So, earlier pre-2021, you know, like they were building all of this with an EMS player, with an EMS partner. So, it was not done on their facility. They bought the design, they bought all the components, and but the manufacturing was happening at, you know, an EMS player. It was outsourced. Now, this is something which they felt a little uncomfortable after a point because they felt that the design is theirs. It’s their IP, but they didn’t have the capacity to kind of go and put out a proper SMT plant, which they did post IPO. Now, when that plant came in, the whole manufacturing became in-house. And it’s it’s a world-class facility. I visited the facility. It’s a really good facility. And because this is approved as per Nvidia other standards also. So, you know, that is how it is made. And it’s state of the art. And that helped them to kind of get better margins because once you have your own manufacturing, you can you know, kind of do a lot more things than outsourcing. And what helped also is that they they got a lot of people from that EMS player also because they wanted that talent because talent is a little scarce in this part of him because how many people have you met who knows servers in and out? You know, everybody talks software, but this is a bit of a boring stuff, but that’s where the money is being made right now. So, you know, it’s difficult to get talent. So so that is something which they worked on. So earlier, you know, typically debt to equity was 1:1 because they were a small company. They wanted to grow fast and they were getting that kind of demand. So then when you started and cash flows were thin because you know, demand was high, inventories were because you have to keep inventory for all these components. Typically 50 to 60 days of inventory has to be kept because you are going to make you know, you are going to make the motherboard and you’re going to deliver the server which you have a timeline, but the components have to come earlier. And you know, with component prices being so volatile, you need to have you know, that in good inventory so that your pricing does not get impacted beyond a point. So that is one thing. Now, it’s a debt-free company because you know, IPO came in, these large deals have come in, cash flow is good, you know, 125, 130 crores in 26. So that’s another a very good cash flow number which they gave. So it just kind of shows that the company has now matured. Earlier, obviously they needed debt to grow because they were a small company. Today, they don’t need debt. They have internal accruals to grow. One more thing which is that you know, they’re an asset-light business. Now, you would think how can a hardware manufacturer be asset-light? But the beauty of the business is it’s more design-led. So I’ll give you an example that if I have to make a server and that server sells for 10 lakhs or 1 crore, for me nothing changes, right? For me it’s the same SMT plant, it’s the same process. It’s just better GPUs, better design, get better ROI from that product. So for me, I don’t have to put in another, you know, big capital to make that one crore server compared to a 10 lakh server. It is for the process is the same, which is different for a lot of other manufacturers. You know, they have to set up different things. So, they they don’t have to do that. All of that is already done. There is enough space to even increase capacity when the time comes and that that will come maybe in the next few years looking at the demand. So, then they became, you know, the debt went off, cash flow improved, asset turn has really kept improving and that will do going ahead also because the capacity of the plant is good enough. They are not even close to full utilization and they have a few years to get there also. So, overall the business really improved and there are few businesses you will find at 35-40% ROI ROC and doing those kind of margins with the kind of order book, but they are placed that way and again it’s thanks to the vision of the promoters that they saw this happening. Like well, one of the first things they did post IPO is they hired like double this they doubled their staff because that is the requirement they saw for the demand that was coming and it has been met very well. And they’ve got people from Intel, they’ve got people from couple of these EMS players who know the, you know, how this works and that has really helped them to scale the business. If tomorrow morning some very large entity, you know, an Adani or Tata or someone like that decides that they want to enter this business, okay, from a standing start, how long would it take them to reach where Netweb is today? From a capital perspective, it’s not difficult because these are large players. To put a few hundred to a thousand crores is not very difficult for them. The biggest challenge is to get those design partnerships with Nvidia, Intel, AMD because these are long partnerships. Secondly, you know, to know the server market, to know the clientele there, to sell a particular way, to understand the software which is required, come up with new software which is required. It won’t be easy for them to do it because there it’s a complex market. See, the thing about software like servers you need to know is it’s a 24-hour running machine. 365 days. Nobody will take a chance with a server. Today any enterprise will say, “Okay, I need the best of servers because it has to be there on time. It has to be on during, you know, anytime I want. It will see surges. It will require replacements like servers are replaced every 3 years because it wears out very quickly. So, that is something which you have to take care of it.” So, those things kind of give them that advantage. See, it’s not like software. Like today, you know, we see the AI boom software is really under pressure because today, you know, agents can write software and can be You can’t do that with it. You you have to actually have the knowledge of that 20 25 years to build those servers. So, the good part is what I see is that a lot of OEMs are coming to India in the Make in India attempt in the government that, you know, a lot of these people are setting up server manufacturing in India which is a positive because as I the demand is quite big and the market is big. So, it’s not a threat to Netweb, but it it’s just going to enhance them and, you know, kind of make them more aggressive to their approach. So, I think the growth-wise I don’t see any challenge there. And in terms of setting up another Netweb, it will take at least 5 to 7 years I feel to reach what they’ve reached over this time. And then also it will be a little difficult. You know, something that you sort of touched upon in many of your answers is that there’s so much like just the speed of innovation in the sector as a whole. Just, you know, the entire AI space. How confident can one be that Netweb or frankly any company for that matter that is doing what it does is able to keep up with that pace of innovation. So, what does it require in terms of R&D spends? How do you make sure that your hardware has a capacity like existing hardware that you’ve sold to someone, has a capacity to deal with exponentially improving software or anything else that gets added to it. How How do you think about that entire problem? So, the first thing in servers you need to know is that if your servers are not performing, your client is going to switch to another OEM overnight. Because that is how critical the performance is. So, you know, whenever you look at say a Netweb’s numbers and you see a 50/50 government enterprise split, I can tell you one thing because if the server doesn’t work, no government, no enterprise is going to take it. So, you have to prove that your your product is good enough, which they have done over the years. So, that is the first thing. Because it’s it’s a necessity, so that that’s all. Secondly, with GPUs coming, with AI coming in, the problem what is happening now is, say for example, today OpenAI or a Claude runs on a trillion-plus parameters. Okay. Okay. Now, that requires obviously a good amount of compute. Now, with agents coming in, the parameters might go even high. And that is something which is going to take more pressure on your server because more things will be kind of searched. Like, I just read this somewhere that today actually there are more bots than human beings on the internet. That is the amount of agents that have been created last 1 year who have taken over the internet. And for example, if you have to search something, you might go to two, three pages. Okay. An agent might go to 500 pages. Because the agent is tuned to find the best information and deliver that. Now, 500 pages will have an X amount of pressure on the server versus [clears throat] two to three pages. So, exponentially you will need the servers to actually show much better performance. Plus, Nvidia, AMD, and Intel are not stopping. You know, the kind of pace Jensen works with, he comes up with something every 6 months. So, when they are innovating, you have to keep on adding to that. And with the kind of stack which Nvidia has today with the CPU, GPU, and with you know CUDA with the tools, they are targeting everything in the in the AI, you know, data center in AI space. So, Netweb has to obviously keep up with that. So, R&D is a big part. So, 3 to 5% plus of revenue does go into R&D, and that will be for any OEM who’s trying to build something. So, you’ll always have that effort. And that’s why as I said that if you’re new, it is going to take that much time for you to kind of grow. So, that is that is going to be there because the software OEMs also only do software for servers, they’re also coming up with softwares every year. Upgrades, better things because see, this is the first time where the whole sector is changing overnight. So, it’s like tomorrow you’re like before L&M servers were something, after L&M servers have changed completely. So, tomorrow if, you know, agents come in a big way, which is still now not there. If tomorrow we all are using agents for our laptop saying that our agent will go through 1,000 things in the morning and give us reports, then you will need a lot more compute. The way of handling should be different. So, for example, one of the softwares which Netweb does does GPU slicing. Now, GPU slicing is very interesting because see, for example, you have an organization of 40 people. Now, 20 people are doing normal work and maybe 10 people are doing, you know, high GPU load work. Now, that GPU has to be sliced enough so that every server gets the right amount and the usage is perfect to what the demand is. And that is something which a software can do, which they have done. So, these are the kind of innovations you’ll require in software to kind of match up with the demand. So, that that that’s not easy to do. So, R&D is a big part of it, and that will continue to be. And a lot of feedback also goes back. So, sorry. So, you [clears throat] see a lot of feedback from clients that okay, we want this. We have an issue with that. And that’s where your R&D goes more active and builds these things further. So, that is something which is very crucial for a company like Netweb. But is like three three to five percent is let’s call it 100 crores of R&D. Is that is that I mean I I mean I don’t know I would imagine that something like this requires, you know, a significantly larger outlay in R&D just to keep up because of the pace at which things are evolving globally. see the thing is see R&D Now the thing is if I have an Nvidia chip, that chip is improved because of R&D Nvidia does and I benefit directly from it. So, I don’t have to do that part of R&D. So, these people do their own R&D. TSMC is doing their own R&D to make the chip better, smaller, more efficient, more power efficient. So, they are doing their R&D. I just have to do the R&D on my side, which is the software and just kind of making sure that the whatever, you know, kind of components which come in fit well with my motherboard and give the best performance. Yeah. That’s why the R&D is limited to just my part of it. And as I told you it’s a chain, so everybody does their part and then Netweb just does their part. Yeah. So, they don’t have to worry how to make the next best GPU. There are better minds to do that. So, they will do that. Can you talk a bit about the history of the management team? Sort of you know, how how this company began, what was their thought process, and what has led them to where they are today? So, it was started by Sanjay Lodha. So, he’s the promoter and managing director. So, Sanjay ji had a vision that compute will be big, which I also mentioned earlier. And he started ‘96 and he started servers and and they were that time just making servers mean from a government angle that, you know, for supercomputers. So, the idea initially was, you know, basically look at supercomputers. They were also part of the national supercomputer mission, which was in 2015. They participated in that with other OEMs. So, the idea was just to do that and learn with time. So, the design partnership happened earlier. Those design partnerships were so very crucial to kind of make the servers better. And as I told you they were using another EMS player. So, manufacturing was not in-house. So, all of this was happening in that process. The evolvement happened in say 2016 where, you know, two brothers from the family joined the business. So, you had you know, earlier you had Sanjay G Sanjay Lodha and Naveen Lodha and then you had two brothers who joined the business. So, four brothers now were running the business and their idea was to capture the Indian market. So, today also all four brothers look at four different parts of India, four geographies and that split then they went into enterprise, then into software. So, the journey was very clear is to be the best OEM for my making servers in India and looking at India as a market, not abroad. Right now exports are very small, but it was to cater to the demand which will come in India and it was prepared for this kind of a you know, blowout demand which you’re seeing for the last 2-3 years. They are not surprised because they waited for this. Yeah. What do you think are one or two things that the management has got particularly right over time when building this business? One is, you know, transitioning from a hardware to software. They got it bang on because you see today large, you know, hardware manufacturers who are doing servers, they are struggling because see that is a business where debt also is a big part of it because you want to grow on the on the volume part of it. So, you will continue to take debt to build your volume because your growth is coming through volumes. Value growth is not very big, you know, in this business because servers are more necessity. So, you can’t out-price yourself in the market and there are four three four large OEMs globally. So, your pricing is always in global in tandem with global pricing. So, you can’t really, you know, quote X and you know, they’re quoting X minus 20%. That can’t happen. So, pricing is not that important. So, overall the the transition to have software helped. They made use of the Make in India, you know, movement. Third is they really bounced on AI as an opportunity. When AI came in, they realized this is the moment we’ve been waiting for the last 25 years. Because now a server will not be looked as just a commodity which is lying in your office taking your data. Now the server will do a lot more. And that’s where things changed. Today you have you know like you look at hyper scalers coming up in India. You have two three good names who are in that space. And this is a revolution for India itself. You will have a lot more coming in the next 15-20 years. We’ll have our own Googles and Microsoft who are in hyper scalers, who are hyper scalers. Now they will require servers from somebody and Netweb wants to be that OEM for them. So they they’ve caught the demand cycle perfectly. One more thing which you mentioned that the return ratios were good earlier also. So they’ve been very fiscally proficient. They’ve never stretched themselves beyond the point. So and and typically you look at any business which is at a 2 to 5% margin. Distribution is typically a business which is like that. They have to be very good fiscally otherwise it’s very easy to go bankrupt in that business. So before this wave also they were like they they were very they knew what they were doing. They knew what they had and they never over stretched their arms that which kept them in the business long. They kept doing well and that is what led them and this is all one more beauty of the business. It’s all funded by the family. No VCs, nothing, no private equity. Purely a family funded business all through internal accruals and a bit of debt which was there. So that’s another big thing, you know, because typically if this company would have gone with a VC earlier, the trajectory would have been a lot different. Maybe they would have listed a lot earlier also and then you know it’s it’s not the same. So the control they’ve kept with themselves which has helped them to take the kind of independent decisions they’ve always wanted to. Yeah. You know, of you, a lot of people have in this I’m I’m not talking network now I’m just saying generally in the AI space is that at some point the cycle turns particularly for hardware players and you know kind of effectively at some point there’s a shortage and suddenly there’s a there’s a supply glut and things change very quickly. Yeah. To what extent is Netweb’s business cyclical? So I personally believe that AI obviously it’s it’s too new right now. We are still trying to find you know use cases which will help us we are obviously we are doing a lot of you know kind of experimental work to know what AI can do but I think the potential is still not unlocked. The problem what has happened is because hardware prices have gone up so much it’s the cost of computer is very expensive which I feel has to come down. Yeah. Now when that comes down obviously it will be a negative from a demand perspective so all these players will be impacted on the volume side of it and because this is like an unusual level which it has gone to it has to come down. It will. Good part I feel is that when it comes down the cost of computer will also come down. So you you know typically what happened in the dot com bubble. Mhm. Software was the next big thing everybody went gung-ho on software the the valuations went sky rocket and then one day when they realized okay these numbers are maybe a few years away Mhm. it kind of came down. Mhm. But software still won the technology won. Mhm. So I feel the same case will play out here the technology of AI will win. It is just that financially we have Mhm. just gone a lot ahead the valuations are unrealistic and the hardware prices have really dampened that kind of demand momentum. Which will when the prices come and the whole you know AI movement slows down a little the cost of computer comes down and the server manufacturers will again do well. So maybe one two years you will see a bit of a dip on in terms of demand when this cycle comes to us because see at the end they are selling servers. You know it’s it’s not something which anybody else aren’t make and it’s not something which will not have a cycle. Everything will have a cycle. This will also have a cycle. Last 3-4 years have been a really good cycle. Maybe you see 1-2 years where the cycle kind of comes down. And there, from a demand perspective, revenues obviously the 35-40% growth might be a little under pressure, but they still perform despite everything. But in our opinion, it could be, you know, that that could come down to maybe a smaller slab, but margins will be intact because the software component will help you to keep the margins. And then you will see the continuity of the business because from an Indian perspective, now again I’ll tell you from an Indian perspective, globally obviously data centers is the big theme, but global data centers is more about the AI theme. Indian data centers has two things. One is to get the data back to India, to localize it, and the second is AI. So, even if the AI theme stops a little or slows a little, the localization theme will continue. So, you have all the large conglomerates setting up data centers. They are not setting this up for AI. They’re setting it because the data has to be here. So, that theme will continue. So, that that part of the, you know, business which is 30-35% of, you know, cloud, that will keep doing well. The other part which is the national supercomputing mission which I said the government had announced, the NSM 2 will also come in the next few years. That will also be big because today, for example, you know, we have only two in top 500 in the supercomputers globally. We need better supercomputers because it will be required for research, for weather forecast, drug discovery, and the government will look at this at some point. And that will be another trigger for somebody like Netweb. So, for example, today, you know, the fastest supercomputer, you know, that is basically doing like 200x faster speed to a supercomputer which we have. But the cost of setting up that is 5,000 crore plus. We have a few hundred crore. So, obviously it’s a it’s a big thing. The power for that supercomputer is one small town of power in India. So, you can understand the difference in scale. So, it will take time for us. But that will be another trigger for a company that will sell to that kind of market. So, the good part is because the segments are, you know, there are two three segments. Mhm. Demand from one segment will keep pushing the revenue [clears throat] growth in a good way. So, when AI kind of comes because AI was like what? Less than 10% as a segment 2 years back. Mhm. Today, it’s 25 30 odd percent because obviously these deals have come in. Yeah. Maybe when this kind of slows down, the other segments pick up. So, something keeps moving for them. And would you say that Netweb is in some ways also a play on burgeoning space exploration and, you know, that entire area? Is it Does it How does How does it How does it profit from that in any way? So, if we have to set up data centers in space, obviously you need servers which can match that capacity and that’s where Netweb will come in. Mhm. Another area which I find very interesting is quantum computing which I feel is another few years away because we are at least in India, we are at a very, you know, nascent state. Globally, quantum computing is already, you know, going to different levels. Mhm. The beauty of quantum computing is when that comes in, you will see a requirement for servers just go, you know, in a big way, you know, they go up in a big way. Why? Because a quantum computer requires an HPC cluster of servers. It can’t work on just one or two servers. You need a cluster of servers who need to be GPU powered, high-end. That requirement will come in. And when that happens, Netweb will see another big wave of demand from there. So, these two things like, you know, space plus quantum computing is the future. We are nowhere close to it, I feel, right now. We are a few years away because, you know, quantum computers are still not able to give the right kind of responses which we need. Or maybe right not right, but in the way prompt from an AI is giving today, you know, that level of uniformity is not there. So, we are a few years away from that, but when that happens, that will be a new trigger for the stock. To what extent can you underwrite the future for a company like Netweb? Yeah. In terms of what does the company look like 5, 10, 15 years down the line? So, at the base, I think if you’re buying Netweb, you should know you’re buying a server manufacturer, which is a which is a global market which will go as and when the way the market moves, those things will get impacted like supply issues, all of that will play out for them positively and negatively as well. What I would say is that this is a company which will reinvent itself every few years, Mhm. which will kind of put money where the new trends are forming. So, you are buying somebody who is kind of has the insight of that before you think of. Mhm. So, that’s why this is a stock which I believe can be a Nifty 50 stock one day, maybe in the next decade or so, because that is the kind of vision they have, and that is the kind of numbers they are doing, because they know the next trend, and they are in line with that. And the good part is see, there are many industries where you have partnership with one, two vendors, and you’re working. They have it with everyone who’s there. So, they’re not missing out on anything. So, today, for example, you know, like for Okay, I’ll give you a real-life example. Nvidia got ahead in the chip manufacturing race ahead of AMD, and AMD was 2 years behind. Now, AMD is catching up. Mhm. Now, and Netweb has a partnership with both. Right. So, it it caught the Nvidia movement, and now AMD is doing well in inferencing. You know, Nvidia was good in training. Mhm. AMD is coming good with inference. Now, when inference takes over, they will benefit from AMD also. Mhm. So, that is the good part. So, it’s a well-edged bet. So, you’re not saying that, okay, only one thing will take you into the next orbit. So, that is the good part. The segments are good, the partnerships are good, all of that is it. Plus, see, there are two models in AI today. You know, you have the capex model, which is Netweb, where you buy the servers, do the capex today, and you run the business. Then you have the opex model where, you know, you just buy space. You know, there are, you know, cloud operators who are giving you on a rental basis. So, you buy on rent and you see how much computer you require. And that’s another trend which will do very well. Good part is that both the trends can coexist together and that’s where Netweb will continue to do well in that also. I don’t know how this question sounds, but I I kind of feel like I have to ask it. That despite the fact that the company, you know, as of recording is trading at something in the region of 130 times earnings, plus minus. Yeah. Despite that, I feel like I’m getting the sense from you that you think the best is yet to come for the business and people still don’t appreciate how fantastic a business this can be. Is that a fair statement? So, see, valuation-wise, I would say that obviously there’s a lot of the run-up has happened, most of it. I would always see from our perspective, like I’ll give you an example, we came up with an IC when the price was 1,800 and it was trading at 80 times trailing that time. Okay. And a lot of people who we spoke to said it’s too expensive at 1,800, who have now realized know the potential of this company is a lot bigger. So, I would say the entry price for any investor is very important at what price you’re getting it because at the end you’re buying a 35-40% growth company with always the optionality of another 15-20% growth coming from any of these deals which we looked at. So, at best you’re buying a 55-60% kind of gather company going ahead. So, now as per that, what multiple is comfortable for you, you assign and you get it. And at 35-40% ROIC, it’s it’s one of the best in class doing that growth with margins of 15%, you know, which is 13-15% range right now. So, very rarely do you get a company which ticks all the boxes that way. And especially for even a investor who’s looking at core fundamentals or a growth investor who’s looking at the potential. So, I would say that any kind of dip is where you should really look at it. But the long-term story is intact, for sure. What are some of the challenges the business has faced historically? Yeah. And what do you see as potential challenges going ahead? If AI demand really falls So for example, if you know, tomorrow we we are in a world where we realize that AI is too dangerous. We need to like shut this down or we need to like with what happened with Metros a few months back that you know, Claude came and said that you know, this is too dangerous. It can do things which we can’t predict. So we need to slow down. Now if something like that happens and we go back into a non-AI kind of era, which is hard to imagine, but if that happens, then obviously that will impact demand because you know, a lot of the AI demand will go away. Most of it will go away. Sure. That is one thing which will work. Second is if you see a global recession where you know, demand comes down, then this is something you will get a X amount of growth, but the extra [clears throat] which is coming from, you know, enterprise spending, that will kind of stop. So that that’s another scenario where network will be impacted because you know, you’re not going to replace servers or improve servers unless your business itself is doing well on the core part of it. So that is something which I see that you know, global recession obviously will be a negative and it will take time for it to come back from that also. The government, even historically, has accounted for roughly half the business as in in terms of revenue. Yeah. How does that piece of the puzzle evolve? Yeah. And I’m a little surprised by that where I thought that going ahead, the private sector would be, you know, the primary driver of revenue growth. But is that going to be the case or do you still think the government will be broadly half the business? So it will it will be I think 45-50% business because of few things. One is as I told you with the National Supercomputing Mission, you will need more supercomputers. That will be a big part of demand. That will happen. Secondly, a lot of these IITs and the other institutes will look for research so they look for better equipment and that will be another demand driver for somebody like Netweb. Plus defense will be a big part of it. So any defense related, you know, deals will also be a big positive and any related research also to defense, you know, from a warfare perspective or a you need good compute capacity to do those kind of things, you know, something like today a Palantir has really done well in US because they are doing very well on the defense side. So these kind of software companies also will require a good hardware base where Netweb will come in. So government will continue to be a big part of the business. ISRO, DRDO, these are all organizations which will require a lot more compute. And one more thing is that the order sizes for these will keep increasing. So for example, an ISRO or you know, ISRO might buy something worth X today. In 3 years when the refreshment cycle comes in, it might be 2X. Not because of the price but because the technology required is much better now. Like those servers don’t do well anymore. Those GPUs are not good enough. So now you need to upgrade. And obviously that’s what every organization will do even on the government side. So that’s where you’ll see those the government factor still contributing a good chunk to the overall revenue and the growth also. On the private side, are their clients predominantly BFSI companies or what what what are some of the largest sectors to which they provide solutions? BFSI and IT are two areas where they do very well. Obviously IT companies from a server perspective require a lot more, you know, compute. A lot of these tech companies who are in SaaS also require a lot of compute. So sorry. So for example, like TCS would be a client. is a client which they have publicly disclosed also. So Infosys is one of the oldest clients for them. So Infosys will buy every few years and those deals also keep increasing because obviously you need better compute, you need better hardware. So that is something which will keep taking place. I also feel one thing is that there is a whole new wave of AI companies coming in India. Today you have like a homegrown startup like a Sarvam who you know presented some of its models during the AI summit. There are many like that coming up and those will require good compute. Government is also helping them out. I think that is where Netweb will do very well because that is a new area. That’s a new demand which was never there. So these companies are funded, they are getting VC funding, PE funding. These companies will become a whole new growth trajectory for somebody like a Netweb. So there’s it’s a new segment for them. So something will keep coming. As I said, some of the growth drivers keep ticking for them. Right. Well, I mean that is absolutely fascinating stuff and I I think it’s one of those things where with every answer there’s five more questions that come up just because in many ways it’s just such a complicated field and it’s a field where very few people actually really know what’s happening. But I want to thank you for taking the time to explain Netweb to everyone and it’s it was absolutely wonderful talking to you and and and I hope it was an interesting conversation for Yes, yes. Really interesting. I hope your viewers enjoy the conversation and get to learn maybe few things about Netweb, but it’s an interesting company. It’s it’s one of those companies which you you have to also keep learning with the company to kind of be up-to-date, but really fascinating from and and from a tech perspective, you know, where a lot of people say that India doesn’t have any innovation or R&D. This is one of those companies who has done very well. And you know, they need to get that recognition and and I’m happy that they’re getting it right. Absolutely wonderful. Well, thank thank you so much for taking Thank you so much. Thank you.