E2E Networks Limited (E2E) Earnings Call Transcript

Ladies and gentlemen, good day, and welcome to the E2E Networks Limited Q2 H1 FY '26 Earnings Conference Call hosted by Go India Advisors. [Operator Instructions] Please note that this conference is being recorded. I now hand the conference over to Ms. Raashi Khatri from Go India Advisors. Thank you, and over to you, ma'am.

Thank you, and good afternoon, everyone. We welcome you to E2E Limited, H2 and -- H1 and Q2 results and earnings call. We have with us on call today, Mr. Tarun Dua, Managing Director; Mr. Nitin Jain, the Chief Financial Officer; and Mr. Ronit Gaba, the company secretary. I must remind you that the discussion on today's call may include certain forward-looking statements and must be viewed in conjunction with the risks that the company may face. I now request Mr. Tarun to take us to the company's business and financial highlights, subsequent to which we will open the floor for Q&A. Thank you, and over to you, sir.

Yes. Thank you. Hi, everyone. Welcome to the Q2 financial year '26 call for E2E Network. Let me very briefly talk about our business. So we are a compute infrastructure player, including cloud GPU, based out of India. We have built capacity of more than 3,900 cloud GPUs and we have been in the business for providing compute infrastructure since 2009. So we have built a very strong engineering team over the last many years and we have built our -- from the ground up self-service platform for running our entire cloud infrastructure, which is accessible on the web on myaccount.e2enetworks.com. And TIR is our AI/ML platform, which is a part of MyAccount infrastructure. So our platform allows for like a lot of features, which are typical of any large cloud provider on any self-service basis. So let us talk about some of the updates -- recap some of the updates, some of which have already been shared with you. So your company received 2 reasonably large orders from IndiaAI Mission for INR 88 crores and INR 177 crores. And these are primarily for customers who want to run their own LLM trading for building large domestic LLM models, which are primarily focused on India and of course, like depending on the choice of the customer, they might be usable globally as well. So these 2 large orders, like essentially will -- we are in advanced discussions with both the IndiaAI team as well as the customers on helping to go live with these and we expect these to go live very soon. Now we had made a guidance for FY '26 March, for the monthly run rate, where we mentioned a number of anywhere between, say, anywhere between INR 35 crores and INR 40 crores per month. So with these 2 orders, like -- it looks like potentially, we should be able to meet our monthly run rate numbers, hopefully, much earlier than March. And obviously, there is a great deal of confidence because of the IndiaAI orders that it should be possible to do this sooner. Now, we are also continuing to see a very robust global demand for cloud GPUs. The demand continues at a price point at which Hoppers are available globally. And at a reasonably higher price spread for Blackwell GPUs, we are also seeing a lot of demand globally. So basically, like what we had positioned was that as we have visibility into our capacity utilization for Hoppers, we would be looking at procuring more Blackwell. So given the market scenario of like a lot of demand picking up for Blackwell, we are already in a very advanced stages of placing the orders for nearly 2,048 Blackwell, majorly B200. This is primarily funded through our internal accrual previous fundraisers and debt from institutions. Now we spoke about like how the Chennai location had become a bit delayed. So our Chennai location went live from 1st of August. Now all of our capacity today is completely online and available for use by our customers. I would also like to once again share the news about the ongoing acquisition of Jarvis Labs asset. So primarily, the reason for kind of like acquiring the business of Jarvis Labs was essentially to be able to service like a lot of additional self-service demand for a cohort of customers we were currently not serving, especially the customers who operate globally. And the -- even more importantly for us was the people who were a part of Jarvis Labs, now they are a part of E2E Networks. And essentially, we believe in their ability to contribute overall to building up like a very fine DNA for E2E Networks. So that's one of the reasons for the Jarvis Labs acquisition. And hopefully, in the medium and long term, we will see a lot of benefit from having access to the great talent that we have already built within E2E and acquired through Jarvis, and where we will continue to work on acquiring more tech talent in India. So over a period of time, we will see like great benefits of that. Now that being said, like our focus continues to be like from a management perspective, our focus continues to be to rightsize the cloud GPU infrastructure acquisition in a way that is prudent and as well as also aggressive, so prudent and aggressive. I'm trying to reconcile that. So basically prudent from the point of view that like once we have visibility on how the capacity would be utilized, we want to buy more. And buy as much as that, like we don't lose business because of lack of capacity. So that's the balance we have always tried to maintain. And in future also, our strategy remains to balance these 2. Now that being said, like overall, the global market that we look at, like we are still very, very small. And this essentially means that like there is plenty of addressable market for us to pursue. And whatever capacity we build, we are very confident that we will be able to utilize that capacity for revenue. So that continues to remain our strategy. On the platform side, on technology side, we are committed to continue to invest in our sovereign technology. So when I use the word sovereign, I essentially mean that we are building technology, which is primarily reliant on open source, with minimal reliance on international software, which could be subject to, say, encumbrances in the future. So essentially, we continue to invest in our technology, home grown technology for our customers, for our cloud and we'll continue to do that. And we foresee a long-term benefit for continuing with that strategy of having full stack sovereignty. So, essentially, sovereignty is -- we have defined it like for other customers but like one is that like we are an Indian cloud GPU provider, where the infrastructure is owned by us, the data centers are physically located in India. And the software being operated on our cloud is primarily built by us or is open source with an ability to redistribute that open source howsoever we see fit. So we have 3 levels of, sovereignty over there. And we believe in India's strategy of building sovereignty over the AI model, which is what we are seeing the investment from IndiaAI Mission from, they are trying to build a sovereignty of Indian AI model. So that becomes like 4 levels of sovereignty. And of course, like as a country, we will continue to invest into the entire sovereignty platform, just like in the past, India has invested into the digital public infrastructure. I think like the partnership between public and private enterprises would lead to autonomous and self-sufficiency. And we believe that like we are a big part of that equation in India. So with that, now, I would like to hand over the conversation to Nitin, our CFO, who is present. And then we would open up the discussion for a Q&A session.

Thank you, Tarun. Good afternoon all. So I would just run you through the financial highlights. So in the quarter ended September 30, 2025, we reported a revenue of INR 43.8 crores, which is 21% over the last quarter, which is April to June 2025. EBITDA margins have significantly improved to 41% from 29% in the previous quarter. Other income comprising interest -- mainly comprising of interest on deposits has declined on account of utilization of deposits for the payment of CapEx. As a consequential to the depreciation and our -- we reported a net loss of INR 13.5 crores. Now we can open the floor for the question-and-answer session.

[Operator Instructions] The first question is from the line of Bhavya Gandhi from Dalal & Broacha Stock Broking.

Sir, my first question is regarding the CapEx, since you have fully utilized with the 2 contract that we have received recently. So what is the CapEx plan for this year and next 2 years, if you can provide, in terms of amount as well as in terms of GPUs?

Let's talk in terms of GPUs. I think like -- so what we believe is that like globally what we have seen, and we hope that, that would be replicated in India is that the Blackwells are going to be much bigger than Hoppers in terms of like global utilization of that infrastructure. And the price point for Blackwell is different. So there is no cannibalization for Blackwell -- cannibalization of Hopper because of Blackwell. So we don't expect that to happen. So overall, we believe that like whatever number of Hoppers we have acquired, like it could be -- the Blackwells could be anywhere between 2x to 8x, that number. So when we run through the entire cycle, we will know. So like I mentioned that like we are both prudent as well as aggressive. So we'll continue to operate both prudentially as well as aggressively to acquire more capacity opportunistically. So as we are immediately planning to acquire about 2,048 Blackwell, so as soon as they are deployed or while we are in the process of deploying them if we see that there is potentially more capacity that can be utilized based on the input that we keep getting from overall, the market and the ecosystem, we might decide to kind of double up that number all the way to 4,096 in the -- before March. And we believe that like another 12 months or so, we may keep acquiring Blackwell until we are ready to -- until the next generation becomes ready to ship and become available for receipt in the Indian market. So we believe that there is a long runway to be able to answer that question today about what's the maximum capacity but I think on an immediate basis, I think anywhere between 2,048 to 4,096 GPUs should be there.

This, you was mentioning before March -- this March you're saying, March '26?

Yes, I'm not talking about deployment, but like in terms of like -- so one is, of course, like 2,048 is already underway, under process. So potentially another 2,048, even if it goes under process, so whether that deployment will happen in Feb or March or later, that would be hard to say today unless we place those orders with the vendors.

But can we assume like 2,048 if average pricing for 1 Blackwell GPU is closer to INR 50 lakhs, can you assume like INR 1,000 crores CapEx by FY '26, before the closure of FY '26?

See, it's hard to kind of like put numbers like that, like I said, that's like -- look, the pricing is dependent on a lot of things. So let us not put kind of a hard number to these things. But we are committed to building sufficient Blackwell capacity with a view not to lose any opportunities for Blackwell infrastructure coming our way.

And sir, second question is regarding the purchase of service and consumables. I believe that is largely with respect to Chennai facility and that is largely done with. So can we assume that INR 14 crores would be the run rate going forward in future quarters as well?

See, some difference would be there before the Blackwell capacity gets added in terms of when your utilization goes higher, the cost obviously goes higher. And we continue to build capacity for non-flagship GPUs or GPUs for storage and many other things. So that being said, that like broadly, that is in range, that it will not change drastically, unless we add a few more clusters of a couple of thousand GPUs. Till that time, we don't expect it to change drastically, but like that number obviously would change. but not to a significant extent.

Got it. And sir, just regarding the data center capacity, how much megawatts do we currently have? And how much do we require with the Blackwell capacity coming on stream? If you can throw some light on that?

We have some of the older capacity that we classify close to 1 megawatt or so. Then the newer capacity that we built in the recent past, like last 2 years or so, including the Chennai one, the capacity available to us is about 9 megawatts plus 1 megawatt older capacity. So that's an overall of about 10 megawatts or so of capacity. So broadly, if you look at 10 megawatts, we're looking at like something -- somewhere between 8,000 to 10,000 cloud GPU can be made available over there. That being said, our Chennai facility, like we have the ability to kind of get more capacity over there at a reasonably short notice.

Got it. So just a follow-up on the depreciation part, is the depreciation charge for the new gross block already capitalized? I mean, it's there in the P&L right now? We are at INR 42 crore run rate. So that would be the largely depreciation number for following quarters as well?

Let Nitin answer that question.

Yes. So I think the number will be broadly similar, but only 1 point that Chennai cluster became active on 1st August. So 1 month depreciation would be an additional number.

Got it. So we can assume like INR 50 crore depreciation run rate for the entire year -- sorry, for the next coming quarters?

Yes, it should be on the same lines.

[Operator Instructions] The next question is from the line of Anand B from Ksema Wealth Private Limited.

I just want to get this thing, one thing, the data centers, which you are operating, that is of third party, is it all of them?

Right. Yes. We don't own any data centers. So these are all third-party data centers, yes.

Okay. So recently, TCS announced that they'll be installing 1 gigawatt data center in Andhra. So -- and they said that the main customers will be for accelerators, so cloud GPU accelerators. So will E2E Networks be a client or will it be participating in that data center?

I think this is a question which is a bit premature. I think like we would like to -- we like to see a data center where we are buying capacity. So once they build those data centers and they have the capacity available for sale, we would be happy to consider them along with all the other players. So we essentially go by a merit of any player before buying any services. So if TCS is in the market to sell those services, we are happy to speak to them, of course.

No, I'm saying, are like talks underway or something or like...

No, No. Not currently.

Okay. My second question is, you said the MRR for March 2026 would be the same as you mentioned before '24 but the MRR...

Yes, yes. We've reiterated our guidance based on the confidence that we have received from IndiaAI Mission, where if we put the 2 orders together, then that itself is about INR 20 crores-plus MRR coming, I mean, and then we have another 1.5 quarters to build up more run rate. So we are reasonably confident of meeting the guidance that we have given earlier.

Okay. So would you say that in next quarter, sir, would you reach that MRR?

You see, the guidance we've given is for March '26. So we would like to stick to that. Hopefully, it should be earlier than that, obviously, because we feel that these are -- these deployments are imminent. So as and when they go online, obviously, they will start showing up.

Okay. And also last question. I just wanted to get some information right. So currently, you have 3,900-plus GPUs available, right, that sort of capacity. So previously, you mentioned 8,000 to 10,000 cloud GPUs, that is apart from the AI GPUs that is available?

I think somebody asked a question about like what's the data center capacity. So I was like essentially talking about the data center capacity can contain that many GPUS. Doesn't mean that we have that many GPUs today.

The next question is from the line of Keshav from Niveshaay.

Sir, regarding the recent orders that we won to IndiaAI Mission. So as you can see that realization for this order appears to be on the lower side, which potentially extend the payback period. So could you provide some color on the margin profile for this project and we should -- like what margin should we expect going forward?

I think overall, like once we actually go into implementation and then we still look at like all the requirements that might come through whether from IndiaAI or through the customers who have received these allocations, only then we will be able to comment on like margin profile looking back. So currently, we are not able to kind of say whether it would be too low or too high or whatever. So I think we'll have to do a look back to figure out that, how the overall margin profile turns out to be.

So sir, like are we in line with the margin guidance that we earlier guided?

You see EBITDA guidance like has been always that...

I think like we have the ability to hit somewhere close to 70% plus/minus something. So we stand by that from an overall company perspective.

Got it, sir. And sir, any visibility that we have got on the software side on the sovereign AI cloud?

No, not correctly. There is no immediate visibility of revenue on the software side.

The next question is from the line of [ Shubham Agarwal ] from YES Securities.

So I just have one question. What is the useful life that you take for your GPUs for the depreciation consideration?

See, for the depreciation purpose, I believe it is as per the company's act which is 6 years, for all services as far as the GPU and mostly useful life maybe that for GPUs or CPUs is at least like happily and easily about 7, 8 years.

The next question is from the line of Varun Gandhi from Fident Asset Management.

So Tarun, my question to you is on the AI development side in India and as far as my understanding is, if someone is using an LLM API, the computing happens on the LLM's end. So the essential computing can happen on foreign waters, foreign land. And what I wanted to understand is, in India, do you see that significant of compute requirement because if AI development is happening through APIs and I'm not sure how much is happening through APIs, and if it is happening, then it's not much of a big market for us. Just some clarity there.

There are like so many nuances over there in terms of what people are trying to do with AI. So if you look at the use cases, where people need to have a control over the outcome where people need to have security, where people need to have the sovereignty of how they are processing the data. Let us take the case of people with a significant number of users. So eventually, with a significant number of users, if you end up doing financial transactions in India, then I believe that a lot of that infrastructure will would to be set up in India itself. Second is that these are not the people who would use a one-size-fit-all generic inference APIs to do their business or to run their business. So they would obviously look for an edge, which can only be achieved by having a dedicated back-end infrastructure or the inference APIs, where you are able to make changes to the underlying platform with the help of retraining, fine-tuning and being able to teach the AI model. So we see a lot more development in the open source AI, which enables all of these things that we are talking about. So while we have seen that proprietary AI definitely comes up first and says that, yes, we are doing something that nobody else can do. But what we have seen is that the rate of change in open source is very, very rapid today. And mostly, we see that with a 1 quarter, 2 quarter or 3 quarter gap, we see open source, which is able to do 90% of what proprietary AI models are able to do. Now the way to get -- drive maximum benefit out of those open source model, is by running your own org boundary around those model and being able to take definite and direct benefit of those models. And health is not a major use case alone for AI. AI is present everywhere. So all said and done, where the developers are using AI is not the only question we should be asking. So there is a need and demand for sovereignty over AI, which is very important. And that I believe...

Sorry to interrupt. Sir, the implication is that we're largely dependent on open source architecture and developments around that architecture.

Okay. So sure, that is one way of putting it.

Understood. And just a follow-up on this. So OpenAI has opened an office in Delhi and still very premature...

We will not be commenting on OpenAI over here in this call. So yes, OpenAI is a large company, but we are not talking about that over here.

I see. So is there a chance we could be fulfilling their compute resource? Or you won't be commenting on that currently?

Too early to say what I would say.

Got you. And the last question would be, although the economic life of a GPU, you said was 7 to 8 years. Do you think going forward since the technology is advancing so rapidly, do you think 7 to 8 years of economic life would be an appropriate assumption?

Yes, it has continued to be an appropriate assumption in the past. Now future, of course, nobody can predict. But what we have seen is that the entire ecosystem of what software has been built on top of what AI and the stack, so all of these things put together continue to operate best at a particular price performance over a particular model of a GPU. So that doesn't change very rapidly without investing rapidly into development and data science and AI/ML effort. So net-net, what we have seen in the past is that, that continues to kind of be the future as well, where any GPU model that is existing, that will have a sweet spot for certain use cases, which kind of like came and go during the peak prime time of that particular GPU. So, for instance, currently we see all 3 big models, big GPUs, the Ampere, the A100 series -- A100 and related series, and the Hopper series and the Blackwell. So there is a demand for all 3 of them. So, essentially, we are looking at a -- almost a 6 year life cycle across the 3 of them and Blackwell, we are simply entering into that life cycle. So, that being said, we believe that our assumption that GPUs will continue to be useful is grounded in what we are seeing.

The next question is from the line of [ Aminesh Jain ], an individual Investor.

Sir, I just wanted to understand that other current assets in the balance sheet in H1 have gone up significantly. What I understand is that it is a GST input receivable item. So, I assumed that it was for our GPUs that we had in CWIP. So, I thought it was fully registered, if you can address that concern. Why has it almost doubled in amount?

Nitin, can you answer that question?

Yes. So, the preliminary amount includes the GST balances and the fixed deposits, other current assets, it's a advance to vendors that we have given across for procurement of the new GPUs. That is the other reason along with the previous balances of GST receivable.

So, the current increase that has happened from the previous amount which has doubled, it's not -- GST input hasn't gone up, some other items have gone up?

Yes, yes. GST input was already factored in the March financials because the capitalization was under CWIP and GST input was already taken. During the current quarter, it has gone up primarily because of advances given for the procurement of new GPUs.

Got it. And the next question was regarding our debts that we have taken about INR 100 crores. So, can you guide what our full year debt would be and what our debt utilizations would be for next year?

So, debt what we have taken the facility is INR 450 crores, which would be utilized as per the terms of agreement with the vendor during the payment cycle. So, by end of March, that completely should be utilized across, by end of March or early April.

Got it. And just last question regarding our initial 2 contracts that we have won, it has been more on the basis of GPUs being given out on an hourly basis. Going forward, if we have any enterprise wins, will this continue to be a model or will we move towards a more traditional cloud- based model, which is our pay-as-you-go model, which will actually avail our platform benefits, if you can throw some light on how our model is going to shape up going forward?

So, we continue to work on all aspects of how people want to utilize the GPUs. So, there would always be customers who want to kind of use GPUs for short periods of time where they would want to do hourly, weekly, monthly, three-monthly, and there would always be customers who would look for longer-term engagement in terms of how they want to utilize their GPU, so, that is there.

The next question is from the line of [ Kartikeyan Vaidyanathan ] from Raj Investments.

I think I have 2 to 3 questions. First is, what exactly is the moat that you enjoy compared to other data center players and how you are sort of -- this whole L&T buying some of the things in your company, how are you sort of leveraging L&T Tech into this whole place, can you throw some light on this? Just want to understand, because there are other data center players what is that unique solution that E2E offers?

See, we are a cloud player, not a data center player. So essentially, we have this whole stack of services. So we have been supporting GPUs all the way since 2019-20 kind of timeframe, and we have been working on our software platform for the last so many years, and then we have the ability to provide the solutions of the kind that customers need. So, that's what we would like to believe is our moat that our customer orientation towards providing the support, providing the solutions and providing the software platform on which to run either training or inference or model endpoint deployment and the ability to integrate with our general purpose cloud platform that we have built up over the last so many years. So, all of these are a part of our moat. And of course, running a cloud involves doing a significant number of activities that all need to sync up and work together well. And essentially, we have been a team that has been doing that for a long time, and the team itself is also a moat that we have.