ESDS Software Solution Limited (ESDS) Q2 2027 Earnings Call Transcript

Ladies and gentlemen, good day and welcome to the ESDS Software Solution Limited Q2 and FY27 Earnings Conference call. As a reminder, all participant lines will be in the listen-only mode, and there will be an opportunity for you to ask questions after the presentation concludes. Should you need assistance during the call, please signal an operator by pressing star than zero on your touch-to-one phone. Please note that this conference is being recorded. I now hand the conference over to Mr. Pratik Jhaftar from E&Y Investor Relations. Thank you, and over to you, sir.

Thank you, Rayo. Good evening to all the participants, and welcome to Q2 FY27, Earnings Call of ESTS Software Solution Limited. The results and presentation have already been mailed to you, and they are also available on the website of ESTS. In case anyone has not received a copy of press release and presentation, please do write to us, and we will be happy to send it to you all. To take us through the results today and to answer your questions, we have with us the top management of the company, represented by Piyush Somani, Chairman and Managing Director, Nadukuru Sita Ramayar, Chief Financial Officer, to whom we will be referring as Ram, Ram, Saurabh Daga, account and finance team, and Prasad Devkar, company secretary and compliance officer. Piyush will start the call with brief overview of the quarter on business, which will be then followed by Ram, who will take us through the financials. Then we will open the floor for Q&A session. As usual, I would like to remind you that anything that is said on this call, which gives any outlook for the future, or which can be construed as a forward-looking statement must be viewed in conjunction with the risks and uncertainties that we face. These risks and uncertainties are included but not limited to what we have mentioned in the prospectors filed with SEBI and subsequent annual reports, which you can find on our website. With that said, I now hand over the call to Piyush. Over to you, sir.

Hey, Pradeek. Thanks a lot. Good evening, everyone. A warm welcome and thank you for joining our Q2FY27 earnings call. So, we also have our H1 numbers to disclose today. Before I talk about the portals, I would like to take a moment to reflect on an important milestone for all of us. We became a listed company on 4th September this year and I would like to once again thank all our shareholders for trust and confidence we have placed in us. Our first earnings interaction as a listed company happened very soon thereafter on 25th of September. As we are still relatively new to the public markets, we thought it would be useful to spend a few minutes today explaining ESVS, our business model, and how we look at the business, particularly for shareholders and investors who have joined us after the IPO. If you look at ESVS journey over the past two decades, we have witnessed significant shifts in the technology landscape and we have proactively evolved alongside them. Alongside this, our approach to innovation and solving customer problems has remained consistent. Back then, we began with web hosting supporting businesses as they established their digital As the requirement became more complex, we expanded into data center and mission-critical IT infrastructure. We have been early adopters of emerging technologies. We recognized the potential of cloud computing early and developed our own cloud platform with patented vertical auto-skilling technology. As customers increasingly sought integrated services, greater control, and digital sovereignty, We developed an artificial intelligence based autonomous sovereign platform with Swaraj Cloud and AI-enabled sovereign products, bringing multiple capabilities together within one ecosystem. Today, as artificial intelligence transforms the technology landscape, we are building on this strength through AI factories, large-scale GPU clusters and end-to-end AI-enabled solutions. At every stage, our approach has been to look beyond individual products, understand the larger customer challenge and bring together the right technologies to deliver a complete end-to-end solution. This philosophy has remained the foundation of our evolution and continues to guide the future direction of ESDA. Today, our business can broadly be understood across three offerings, infrastructure as a service, managed services, and software as a service. Instead of looking at this as independent businesses, we look at them as an integrated technology platform through which we support customers across infrastructure, cloud, software, security and the ongoing management of their technology environment. Let me briefly explain how our three core business segments work together. Infrastructure as a service forms the foundation of our business. This is where we provide customers with the AI infrastructure or cloud infrastructure or data center infrastructure, they need to run their applications and critical workloads, including be it their data center, cloud, co-location, AI compute, AI factory. So, whatever compute capacities or whatever the hardware capabilities are required that we provide to our customers. Manage services take this a step further. Beyond providing infrastructure, we help customers manage, monitor, secure, and optimize their IT environment. This allows them to focus on their core business while we take care of their ongoing technology operations. And software as a service represents our proprietary software and platform capabilities, including our growing Swaraj product portfolio across cybersecurity, observability, automation, and enterprise technology. What is important to understand is that these are not three isolated businesses. They are interconnected capabilities that allow us to address a customer's complete technology requirements. A customer may begin with infrastructure, expand into managed services, and subsequently adopt our software platform. This integrated approach enables us to build deeper customer relationships, address evolving requirements, and deliver end-to-end technology solutions. Our customer base spans enterprises, government organizations, and BFSI clients. ESDS is also not only an infrastructure company, we have built our own technology and software capabilities over many years. Swaraj Cloud is our patented vertical auto-scaling cloud technology, and we continue to build our capabilities across AI, cloud, cybersecurity, and GPU infrastructure. What differentiates ESDS is that we don't look at technology requirements in isolation. We focus on understanding the larger business challenge and bringing together the right technologies, services, and expertise to deliver tailored end-to-end solutions. The integrated approach allows us to serve as an accountable technology partner, simplifying complex IT environments while enabling digital sovereignty and giving organizations greater Looking ahead, over the next three to five years, we believe this model will become increasingly relevant. AI is fundamentally reshaping businesses and enterprises need an integrated technology competition to build the flow and scale AI. To our AI factory ecosystem, we bring them capabilities together under one umbrella. TSDS represents more than a traditional cloud We are an end-to-end AI ecosystem and IT service provider, focused on strengthening our competitive position and delivering sustainable long-term growth. As we start our journey in the public market, I would also like to provide some context on how we believe our performance should be evaluated. Given the nature of our business, quarterly performance can be very based on project execution and revenue recognition. We therefore believe year-on-year trend Welcome to the Q2 FI 27, we are pleased with the progress made during the quarter. During the quarter, we continued strengthening our core business while accelerating our transition towards an integrated AI enabled sovereign technology ecosystem. Our priority is centered on expanding the Swaraj product portfolio, advancing AI factory business model and GPU infrastructure initiative and bringing together cloud managed services and AI capabilities to address increasingly complex enterprise requirements. Going forward, our focus remains on strengthening the core ESDS business while building our AI and GPU infrastructure opportunities in a major manner. We are excited about the opportunity ahead, but equally our focus remains on execution, capability, capital discipline, and creating sustainable long-term value for our customers and shareholders. With that, I will hand over to our CFO Nadupul Sitaramaya, you can call him Ram Elvet, to take you through our Q2 financial performance. Thank you.

Thank you, Kiyush. And good evening, everyone on this call. And thank you for joining this call. Really appreciate it. As Piyush mentioned, I'll take you through the key financial highlights for Q2 fiscal year 27. I'll start with the year-on year performance, which we believe gives a better view of our business strength, while also explaining the important sequential moment, whatever is relevant. So, revenue from operation in Q2 is 1383.4 million INR, whereas for Q1, it was 1336.6 million INR, resulting in a growth of 3.5%. Whereas H1, when I compare H1 fiscal year 27 with H1 fiscal year 26, the revenue in H1 fiscal year 27 was 2,719 million, whereas revenue in H1 fiscal year 26, This is 2183.7 million with a growth of 24.6% as we achieve. And the total revenue in H1 fiscal year 27 is 2788 which includes 68 million of other income against a total income in H1 fiscal year 26 of 2 to 19 million. So thereby we see a growth of 25.6%. EBITDA margin for H1 is 1137.5 million, whereas for H1 fiscal year 26 is 1012.7 million, so we again see a growth of 12.3% there. Similarly, PAT has also increased to 603.7 million in H1 fiscal year 27 against 477.9 million in H1 fiscal year 26, thereby we see a growth of

So total amount raised through

IPO issue is 7200 million. Utilized out of the same is 469.13 million. The unutilized amount stands at 6730.87 million as of date, which is most of them we have close to 6500 million. We have put that in fixed deposit and remaining amount has been there.

In the monitoring account, we had a cash equivalent

amounting to INR 16368.31 million and other bank balances equal to 3419.80 million as on September 30, 2026, which includes FDR due to IPO proceeds close to 6500 million.

Thank you very much. We will now begin the question and answer session. Anyone who wishes to ask questions may press star and 1 on their touchtone telephone. If you wish to remove yourself from the question queue, you may press star and 2. Participants are requested to use handsets while asking questions. Ladies and gentlemen, we will wait for a moment while the question queue assembles.

To ask questions, please press star and 1. The first question is from Deepak Potar from Sapphire Capital.

Please go ahead.

Yeah, hi. Am I audible, sir?

Yes, you are audible.

Yeah, thank you very much, sir, for this opportunity. So, just wanted to understand on the Sharon part, I mean, we were expecting to get commissioned the Sharon GPUs this month, right, November. So, just wanted to know the status of it and how much revenue it can contribute in FY27 and FY28 and corresponding EBITDA margin we can expect in the future. Yeah, that would be from my side, yeah.

Yeah. So, we are in the final stages of testing now. So, NVIDIA testing is happening simultaneously. As we speak, NVIDIA is doing the cluster testing. So, this is expected to complete within the next 2 to 3 weeks time, maximum 3 weeks time. So, on the higher side, they have mentioned that by 27th of November, they will hand over to us for testing. And after that, we have to hand over to our customers for 20 days of testing. So, on the higher side, the go live would happen. RFS ready for service would happen on 17th of December. This is on the higher side. Lower side, it can happen somewhere in the first week of December also ready for service. We are hopeful we should be able to in this quarter, we should be able to book one month of our revenue and one month of our profits. And next quarter would be full all three months for the next quarter. So, somewhere around like 50 to 60 crore rupees in that range is going to be our monthly fat once the cluster goes like.

So, 50 to 60 crore per month,

I mean after the entire 8200?

Yeah, all 8200 once they fire up, then somewhere between 50 to 60 crore will be our fat per month So, I will give you an approximate number only, cannot give you an exact number.

288 crore is monthly revenue.

Yeah, 288 crore would be the monthly revenue.

So can you just repeat a misdeaf number?

Approximate number would be 288 crore of revenue.

Okay, so annually around 3000 crore revenue we can do it at, I mean, next year?

Yeah.

So, so the PAD margin that is coming is close to what, 15, 15%, around 3000? So 500, 600 crore kind of a PAD we can generate from this entire utilization of this 8200?

Yeah, so domestic pat margin would be like in the range of 26-27% and this one AI factory pat margin would be in the range of 16-18%.

And once our AI factory also goes live in India then that will have a higher margin.

Sure, sir, that would be it from my side. Thank you so much, sir.

Thank you.

Thank you. The next question is from Yash Sadani from Integrity Please go ahead.

Hi, sir. So, you told us about the revenue and the profit numbers. Can I get to know the per hour revenue rentals that we are earning from this new customer?

Hello? Yeah, can you hear me?

Yeah, now I can hear you. Yeah.

Yeah, asking, can you just let us know the hourly rentals rate that we are charging from this customer for this share in AID?

Yeah, so as I mentioned, we can't be giving far too many details about it, but yeah, our monthly revenue would be 288 crore and monthly pack would be in the range of 50 to 60 crores.

Okay, got it. Just another question on the Sharon AI deal only. As you know, Sharon AI is also selling the services to other customers. And in the light of recent increase in GPU prices of B300, how do you see Sharon themselves competing with you guys in terms of selling the services?

This is a committed AI factory and private dedicated AI factory for us. It is not shared with anyone else. So in this AI factory, there is no competition. These 8200 GPUs are fully contracted to ASU.

Okay. And what's your view on recent increase in B300 GPU prices?

We are also committed to some clients. So that's the reason we also cannot increase our prices. Looking at the current prices, we could have had a potential to earn almost 150 crore piece pad every month, even more than that was a possibility. But unfortunately, because we are already committed to our clients, Sharon AI is committed to us, so we cannot charge more.

So, for 8200 is committed to another client from you?

Yeah, certain number of clients, So, that's why we now don't have an opportunity to increase the price.

Okay, but we will be able to increase the prices in our own data center that we are building in India.

Correct, in India what we are building, so there we will be able to charge higher prices.

Okay, got it. Thank you. That was so much. Thank you.

Thank you. The next question is from Amit Chandra from HDFC Securities. Please go ahead.

Thanks for the opportunity. My first question is again on the Shedon AI. Obviously, you have mentioned that there has been a deal. So, any specific reason that you want to highlight why the deal has happened? Especially if I see the Shedon AI performance in terms of their performance, one month they have been under stress and the market cap of Shedon AI has gone down. So, is there any kind of a funding issue from?

Yeah, Amit, thanks. Good to connect with you. But Amit, we should never compare a company by one-day sale, but Sharaniya is a very, very strong and ethical organization, very much firm on their commitments, their delivery timelines. What they have committed to us, they are in the final stages of delivering also that. They have more than $2 billion sitting with them today. So, though their market cap is like $1.5 billion, but they have in their bank account the cash equivalent of $2 billion sitting with them. So much hardware which has already come, multiple AI factories are finally going live. So, cash-wise and financial positioning-wise, they are lucky that we identified them as our partner very early. Because now the same dollars an hour, what we got at three and a half dollars an hour from them, now the market is demanding seven dollars for the same D3 and it. So, they have come to spread with them, we have done with them for seven years, since they have come to coming again back to India, we are again next month going to Australia. So, we keep on meeting on a regular basis. A year back, we initiated a concept of intelligence officer. And what happened was that young people were brought. We did not have a clue that the CAIO would eat up a period of six months. There was nothing left for the CIO and CTO. Since the arrival of CTO in the company, we further realized that massive changes and the CAIO that we brought and under the CAIO and CTO or the AI, they made certain roles happening in ESDS that a role has become redundant. We have seen many such roles in the past becoming redundant. Probably, we will not need to hire. CIO role is 100% redundant now. CTO, we may decide whether we want to hire, but at the moment, we don't see any reason to hire a new CTO.

Answer for the India business, please.

So, it would be difficult to give guidance, but I can tell you that we have a pipeline of 3000, and we are also doing from our IPO proceeds plus customer advance, we will be doing capex of 1500 proceeds, where we expect that that would be definitely a smaller AI factory, but that will go live from India, and we are looking for delivering more AI factories.

So, this kind of

AI factories are consumed by AI labs only at the moment, so majority of the consumption

Before we take the next question, a request to participants to please limit your questions to two per participant. For follow-up questions, we request you to rejoin the queue. The next question is from Pratik Srivastava from Nivesh Wisdom. Please go ahead.

Yes. Thank you, sir, for giving me this opportunity. Sir, I just wanted to understand the share and the economics a little better. I see that in the deck you mentioned, it's a, you know, USD 1.25 million, 12.15 million contract. And you mentioned around, also around 288 crore of, you know, revenue per month once it starts, you know, once it starts getting fired up. So, I just wanted to sort of tie these two, sir. How is this? Is this committed contract and is this over a five-year or seven-year? Earlier, I was hearing also that, you know, it was seven-year lockdown. Now, I'm hearing five-year plus two-year. So, can you please clarify some of these multiple things where I have, where I have confusion?

So our commitment to Sharon AI is for five years, but we have locked the rate for seven years. So if we will renew with Sharon AI after five years, we will renew at that lower price. So 30%, 40% discounted pricing to the first five years. But with our client, we have kept the price flexible. At the end of 5 years, we will increase the price or maybe we will retain the same price that we will be willing to them for the first 5 years.

Got it. And sir, this contract value, sir, this USD 1250 million contract, this is the one which we are talking about when we are seeing this 3000 crore revenue per year once it goes live.

That is our liability.

3000 crore per year is our liability to Sharaneya. And 288 crore per month is our billing to our customer. Okay. Got it. Got it. Got it, sir.

And any color on our core bill, sir, I see that our cloud, you know, services sell 9% Q1Q. The infrastructure is as per 4%. BSSR revenue is down, and customers have also dropped from 2012, 2016 to 1709. So, is there any color on that? How are you seeing this business, you know, from here on?

Yeah. So, what we've done is that we've created an internal subsidiary, which is a virtual subsidiary. It is not an actual subsidiary. And all the small customers We are transferring those. So, we are trying to keep only 400-500 customers in ESDS, not more than that. And all the smaller customers go to an internal subsidiary. They sit in a separate office. They have their own operations. People come at a lesser price. The support services costs are lower there. And they are given a certain profitability target that you achieve so-and-so profitability or ask your customers to go. So, out of this 2,500 customers we had last year, we have actually asked some 50-60 customers to exit also, and we are looking to make a strong base between this top 400-500 customers Maybe in future, we may want to limit ourselves to 250-300 customers only, but at the moment, we are wanting to focus our energies only on the top 500 customers.

Got it, sir. Thank you very much.

Thank you. The next question is from Deepak Pandey from Saghun Capital. Please go ahead.

Hi, sir. Thanks for the opportunity. Sir, on the international demand side for the GPUs, you had mentioned around 50,000 GPU pipelines from your existing clients last time. So, how is that shaping up for you, and will the setup be similar as in it is in ShareNAI, or anything else can be done?

Yeah, so we are targeting to have similar arrangements. We are looking for American partners and South Korean partners, Taiwanese partners. So we are working very closely with partners from multiple regions. We recently signed an MOU with a partner in Sweden as well. So, we are aggressively working on how we can conclude on this pipeline, have firm MSA signed with the customers and advance coming from the customers and how fast can we deliver. So, very important to note that ESVS carries a reputation in the market that they sign the large deal and they deliver the large deal so that perception that ESVS has in the We have to retain that. So, we all the time when we communicate with various AI labs, we very clearly give them a deadline that this is the timeline by which we will deliver, be it V300, GB300, Vera Rubin, or Rubin Ultra, whatever the customer is wanting, we are giving them a timeline and we are very calmly telling them that we will deliver as per this timeline. So, these things take time. Large AI factories, reports of 8,000, 16,000 vehicles kind of AI factories, they take time. Signing an MSA with the customer can get tracked for almost two months. So, their legal team, two, three different legal account sales, we have our own legal account sales. So, that is where it takes time, but we are on track.

Understood, sir. Sir, on the balance sheet, there are some elements, other non-current liabilities towards 670 crores, and there is some other current liabilities around 1900 crores. What does this pertain to?

Yes, so we have paid an advance to share an AI of 140 million dollars, out of which the 50% will be amortized over a period of 3 months from go-live, and the remaining 50% will be amortized in the month of 34, 35, and 36. An exact similar arrangement we have done with our customer. So, customer have given us the advance which is sitting in the liability. You can see the exact replicating amount in the non-current asset and current assets.

Okay. The last question is on the domestic side. When you say you have a pipeline of 3000 crores, can you elaborate a bit on that, where does it is coming from and how the supply side of it will be taken care of? Is it around managed services or again infra as a service largely?

Mostly infra as a service only at the moment. All the demand that is coming to us from different PSUs or different departments, central government defense departments, this is mostly for AI infra. So these are obviously small AI factories that they are wanting. So, we are at the moment positioned very strongly for AI infra workloads of government and also enterprises and banks in India. Managed services and SaaS funnel is not as promising as the AI infra funnel is at the moment.

Thank you, sir. Good luck for the future. Thank you.

Thank you. Thank you. The next question is from Ashish Das from Systematics. Please go ahead.

Thank you for the opportunity. Sir, I just want to understand your expansion plan in the domestic site. So, you have existing data centers, and also some data centers will be coming, upcoming data centers are there. So, Sir, could you please guide us, like, what is your planned expansion in existing as well as occupying data centers?

Yes, Ashish, very good question, but I will not disclose our business strategy, but I will tell you that we are in the process of acquiring data centers with a lot of power, 50 megawatt, 100 megawatt kind of power capabilities. At the moment, it is the most vital thing across the world to get power allocation. So, get a land which is powered. So, in our business language, we call it powered land. So, if you are getting a powered land, immediately get the contracts done, MSA done and get that powered land blocked with you. So, that is what is our focus at the moment. But the detailed strategy that we have for domestic and detailed strategy that we have for international, that obviously cannot be shared.

Thank you so much, sir. That's all for my job.

Thanks, Ashish. Thank you. The next question is from Vedant Sarda from Nirmal Bank Securities. Please go ahead.

My questions have been answered.

Thank you. We'll move to the next question. The next question is from Anmol from Dam Capital. Please go ahead.

Yeah, hi. Thanks for the opportunity. A couple of things I wanted to understand. Firstly, in this quarter, was there any international revenue in terms of, you know, design sort of revenue for the international contract that has been come to us?

No, no, Anmol. Nothing of international customers for design or anything.

Okay. And secondly, I wanted to understand that you've spoken about the GPUs, you know, that you will be building the air factories. Now, from perspective of that, will this be international customers or will this be for domestic customers largely?

We are speaking to both simultaneously, but because we don't have a delivery date at the moment that can we deliver this by April, May, June, what date we are going to deliver. So, the negotiations are going on. Whoever gives us better pricing and better advance, we will give it to them.

Right, right. So, this 1,500 crore worth of GPUs, this is not for one particular customer. This would be for a particular set of customers.

Maybe two customers, three customers can be one single because in the funnel, we have international customers who are looking for this kind of capacitor even two times, three times, four times more than this capacity. So, we have international customers who can take one single AI factory of 1000 Vera Rubins or 4000, 8000 Vera Rubins also.

Understood, understood. And just one thing more into this is that if you can sort of explain us that what would motivate an international customer to come to India and set up its GPU factory rather than, you know, in its sort of country. So, is there any benefit for this?

Power shortage. So, wherever NVIDIA GPUs can be sold, in none of the countries, power is available now. So, be it America, be it Europe. So, Europe had this 30-40 gigawatts 3-4 years back. It's all gone now. So, Europe is left with no power. Australia is left with no power. So, every country which had power, Your AI factories are there now, and they don't have enough power left. So, India is the only country which has got 40-50 gigawatt surplus to be given for AI factories over the next 4-5 years.

Understood, understood. Yeah, that's it from my end. Thank you so much, sir. Thanks, Anamu.

Thank you. The next question is from Shiv from Aryan Investing. Please go ahead.

Hi, PhD. I have a question.

I know that meaningful numbers will come Q4 onwards that you have already explained in a number of calls. I have a question on the future engagements. So since you're saying that power is a major bottleneck in terms of our expansion and also in terms of The future of AI Compute. So what are the new geographies we are targeting? For example, I see that you have been traveling a lot also in other South American countries, and also I saw a report that in the existing Sharini facility in Sydney, where this installation is happening, there also there is a possibility of another So, I am wondering why we have a question about the 80 megawatt expansion cleared by Australian authorities. So, can you throw some light on these facts also?

Yes. So, obviously, we have put up our strategy in place that how we want to bring 1 gigawatt of power under our captivity under our control. So, we have put up a strategy in place and we are targeting all those countries where hydro power is available, wind power is available or geothermal power is available. We are targeting India also because 40-50 gigawatts surplus power will be available for AI factories and government is also trying to promote AI factories in India. But, will we give this to Indian customers or will we give this to international customers is completely dependent on how much profit we can make from each of the AI companies.

Got it.

So, in terms of new inquiries from across the world and the GPU funnel which you already have had, so from which kind of geographies these demands are coming to us from which parts of the world? That would be my last question.

So, demand is coming from AI labs. Now, there are these 17, 18 top AI labs in the world, and very easy for anyone to find out which are the countries where most of the AI labs are there today. So, we are talking to all of them.

Okay. No worries. Thank you so much, and all the best. Thank you. Thank you, Shiv.

Thank you. The next question is from Akash Jain from Visit Global Services. Please go ahead.

My question is regarding the Indian business. We are talking about client GPU expansion in India business as well. So what is the expected timeline and what is the expected revenue potential from that business?

Timeline target is June, July next year, and we are targeting 1,000 crore rupees of revenue from that business per year. Okay.

And with regards to technology leadership transition, our CTO and CIO Sanjeev Nair resigned. So what led to this transition, and who will lead this technology product development functions?

So, the CPO was onboarded two months back and all the product portfolio is with the chief product officer. While all the technology operations, we had created a position of chief AI officer. This was done almost a year back and the entire technology operations were like taken over by So, we started by the CAIO function and people started like all the new generation people who joined the organization, be it the IITN or those who were hired to our autonomous human hiring platform. Those people also started reporting into the CAIO, which made the CIO Chief Information Officer and CTO Chief Technology Officer, those two functions became redundant because of the changes that we did. So, the chief product officer, which was mostly AI product officer, and chief artificial intelligence officer, they have made these two positions redundant.

Okay. And, sir, my question is regarding customer advances. As we have already tied up 8200 GPUs with global customers, so any customer advances you received in Q2?

No, no major advances received in Q2. We have not signed any major contract in Q2, so that's the reason we don't have any advance in Q2.

Okay, so for this 8200 GPUs, the contracts have not been signed yet?

No, no, that was signed. It was signed earlier itself. 8200 GPUs signed long-term back.

Okay. And with global rates going up for GPU, so when can we expect that rates would be revised for our customers? Is there some lock-in for the price?

The Sydney GPU AI factory, we cannot change the rates now because we offered that to customers who were available that time to give us the price which we wanted. Now that the rates have increased, so we cannot, so the rate is not floating. Today the rate has increased, after the year the rate will decrease, so then again the client will negotiate with us, so that's not the right way. These are some largest of the companies in the world which are ready to commit this kind of money for AI lab, AI factories. So, they stick to the contract. Same as with Charon AI, they have committed to us, they will stick to the pricing. We have committed to our customers, we will stick to our pricing for five years.

Okay, and my next question is regarding the India data center business. So, what is the utilization rate as of now and any capacity addition

we are expected in this business?

Yeah, so currently, we are utilizing close to 85% of our existing data centers, and we are also increasing capacity in other three data centers, which would be up and running in next two to three months.

Okay, and my next question is regarding technology of silence, like with NVIDIA continuously launching newer GPU architectures. So, how does the company protect the residual value of its existing GPU fleet? And what is the explicit useful life of GPUs?

It is very important to understand that NVIDIA GPU is not like a car model that it becomes outdated. With NVIDIA GPU and CUDA platform from NVIDIA, they both have a fantastic combination and what CUDA does is every GPU that gets old, CUDA improves its performance because CUDA improves its time. So, typically in an iPhone, the software becomes slow after some time, but in case of NVIDIA, OODA makes the hardware faster and faster with time. That is the reason H100 and H200 in the retail market also is selling at 3 to 4 times of its buying price 2 years back. So, whatever price at which you bought an H100 or H200 2 years, 3 years back, anyone in the market is ready to pay 4 times today of that price. So, this is the reason, Pura is the reason why NVIDIA GPUs continue to have, continue to demand so much premium in the market. Same thing is going to happen with the B300 or GB300. They will also continue to demand price for next 6 years, 7 years, 8 years also.

Okay, sir, my next question is regarding Swaraj cloud business. So, what is the scalability do you expect from this business, considering the thrust on data protection in India?

We are very, very hopeful that DPDP, as soon as DPDP kicks in next year, there will be a kind of a chaotic situation where people may want to exit the foreign cloud platform And come to ESBS. So that would create a kind of a demand for us, which we've been anticipating. We are preparing for that kind of an opportunity to come and we'll definitely love to grab that opportunity.

So as of now, are we in talks with some customers for this business? Any visibility?

Yes, so that is what we mentioned that the

domestic cloud and co-location opportunities that we have, majorly cloud is around 3,000 crore kind of a planet.

Okay, and so going forward with the share and AI deal, are we in talks with some other players who have contacted GPUs from NVIDIA?

Yes, we are talking to other new cloud companies, AI, infra companies also. And looking to contact with them.

Okay. I will come back in queue. Thank you, sir. Thank you, sir.

Thank you. The next question is from Bharat Seth from Diva Investments. Please go ahead.

Yeah, thanks very much for the opportunity. I just have one simple short question, sir. That our deal with Sharon, with our customer for the Sydney Data Centre, is on a take-or-pay basis, right? So, which means every available GPU per hour will be billed to the customer, irrespective of the fact that how much ever is the utilization, right? Correct. Okay. Thank you so much, sir.

That's all from my friend.

Thank you. The next question is from Raj Mohan V, who is an individual investor. Please go ahead.

Yeah, thank you for the opportunity. It appears actually you indicated to NVIDIA evaluation currently happening for the Sharon AI project. Has it revealed any challenges vis-à-vis your initial assessment in terms of timelines and other technicalities? Or has your hardness been really satisfactory? And then you have indicated two 50,000 GPUs in the pipeline. In this light, based on the 7a project, when can one expect you to garner the next deal of this kind of a cluster? Also, based on your internal assessment of your capacity and capabilities, what kind of frequency do you see in garnering first deals? Logically, you should improve on this incrementally. So, broadly, I wanted to know the challenges you faced for this, the next deal anticipated, and the future frequencies.

Yeah, thanks Mr. Rajamohan, fantastic question. It is very important for people to understand that such a large AI factory which conduits 16 megawatts of power 8202 GPUs in one single data center. So, it's not an easy thing. Many things go wrong in this process. And NVIDIA is a very, very, like, strong-headed, like, their auditors are very, very strong-headed. They will not tolerate even smallest of the configuration mistakes. So, they run through, run your entire cluster through very, very stressful tests, and they will make sure that the cluster goes up to 100% utilization, 100% power, 100%. So, the first time we have been able to take over 100% CPU, GPU, RAM utilization for 1 hour continuously, and then from that they will decide whether this cluster can be certified or not. So, fortunately, all of that has now been cleared. The toughest of the tests that our cluster had to go through, those have been cleared now. And we are just hoping now that they will be submitting a report. And as soon as they submit a report, then we should be in a position to take over and then hand it over to the customer for testing 20 days.

Okay. So based on this experience of yours, which has been certified, which will get certified by NVIDIA, your next deal in terms of grabbing or garnering the next deal should happen at a slightly, If one would expect, objectively, for it to happen at a more quicker frequency, is my understanding correct? Okay, and when do you anticipate, like, the next deal of such a proportion? Like, obviously, one cannot expect it to happen regularly, but broadly, I'm seeing in a matter of one quarter, two quarters, one deal. Is that the kind of frequency that one should be looking at?

Yeah, so we are preparing the ground for the next few deals. We obviously, because if we disclose everything today and then from here it takes 8-9 months to deliver the next AI factory, again that has butterflies in the stomachs of the investors. So that's the reason we are trying to make sure that how we can bring the timeline to less than six months time and then we make the announcement. So we are very closely working with NVIDIA, Supermicro, Dell, and the NeoCloud companies, data centers where power is there. We are ensuring that everything that is required in the factory is already there. So, that's the only thing that has to happen is the GPU system and then NVIDIA to do the testing and hand it over to us. So, all that is happening simultaneously. We will definitely come back and share with the regulator first that once we have signed a contract and the delivery timelines are there in the core site, we will accordingly come and notify the regulator.

Okay. Finally, what are your holistic views, which you are normally generous with, on the U.S. midterms? If, as anticipated, Democrats in any of the House or the Senate, do you think it could undergo some change in terms of deployment of data centers? Generally, your perspective on this?

In both scenarios, they don't have power in US. So power companies in US and Europe have nothing with them at the moment.

So whoever wins, will he be able to solve power problems?

Can they immediately give 300 gigawatts of power in America? No. Can they give like 150 gigawatts of power in Europe? So the answer is no. So where do they have to come? They have to go to those countries where power is available.

Thank you very much, and best wishes.

Thank you, Mr. Arun. Thank you. The next question is from Gaurav from GFS Capital. Please go ahead.

Thank you so much, sir. You just gave a guidance that by next year, June, July, we are expecting 1,000 crore revenue. Is it per month or per year?

1,000 crore per year will get added to our India business.

Any expectation on number of GPUs to be increased in India?

In India, it will go up by 1200, 1150 to 1200 GPUs is what we will be adding on an immediate basis in India. Immediate means like by the time they go live again, it is June, July next year. So, there will be more that we will add in India, but then that will be maybe August,

So, by June, July, 1150 is fixed, and by August, September, it will be plus plus, add on to the existing one. Okay, and I know you have answered it in quite a different question as well today. I'm not asking a new deal or nothing, but just asking that in the next five years, can I expect three to four share on air kind of deals? I don't need timelines, nothing, just a broad expectation.

Aspiration is to do many more such than what you are targeting. You have a positive aspiration for the company in which you have invested. We also have a very strong aspiration to deliver more such AI factories. Let's see where we reach. We all work together, create a kind of a positive mindset and things will automatically happen.

Lots of meta to you, sir. Thank you so much.

Yeah, thank you.

Thank you very much. We'll take that as the last question. I would now like to hand the conference back to the management team for closing comments.

Thank you, everyone, for attending this conference call for our results of Q2 FY27 and H1 FY27. Thank you for showing trust on us. We would continue to work for the purpose of investors and our shareholders and the company. Thank you.

Thank you very much. With that, we conclude today's conference. Thank you for joining us, ladies and gentlemen. We will now disconnect your hands.

Thank you. Thank you, everyone.