E2E Networks Ltd (NSE: E2E) Q4 2026 Earnings Call dated Apr. 20, 2026
Corporate Participants:
Raashi Khatri — Investor Relations
Tarun Dua — Managing Director
Nitin Jain — Chief Financial Officer
Analysts:
Bhavya Gandhi — Analyst
Keshav Sureka — Analyst
Bharat Gulati — Analyst
Nishant Joshi — Analyst
Deepak Poddar — Analyst
Varun Gandhi — Analyst
Abhishek Shindadkar — Analyst
Neil Munot — Analyst
Hardik Gandhi — Analyst
Srinivas K — Analyst
Rohan Nagpal — Analyst
Presentation:
Operator
Ladies and gentlemen, good day and welcome to E2E Networks Limited Q4 and FY ’26 Earning Conference Call. As a reminder, all participant line will be in the listen-only mode and there will be an opportunity for you to ask question after the presentation concludes. [Operator Instructions]. Please note that this conference is being recorded.
I now hand the conference over to Ms. Raashi Khatri from Go India Advisor. Thank you and over to you, Raashi.
Raashi Khatri — Investor Relations
Thank you and good afternoon, everyone, and welcome to the Q4 and FY ’26 earnings call of E2E Networks Limited. We have on the call Mr. Tarun Dua, Managing Director of E2E Networks; and Mr. Nitin Jain, CFO of E2E Network. We must remind you that the discussion in today’s call may include certain forward looking statements and must be therefore viewed in conjunction with the risks that the company may face.
I will now request Mr. Tarun to take us through the financial and business updates. Subsequent to which, we can open the floor for Q&A. Thank you and over to you, sir.
Tarun Dua — Managing Director
Yes. Hi everyone. Hi to all our team members, all our investors, and we hope that you have had a good last financial year. So I’ll do a brief coverage of all the progress we have made during the last quarter and last year and then I will hand it over to Nitin to talk about the financial highlights for this quarter and the overall year and then we’ll open the floor for questions. So in this year we have a lot of milestones as a company that as a team that we have achieved. So we have been able to use our TIR platform for large scale GPU clusters and we have helped LLM teams operate trainings on large clusters over the TIR platform. So with this, we have demonstrated a full stack capability of operating both bare metal and container-based large scale deployments for Nvidia powered GPU infrastructure at a bigger scale. And we have not only demonstrated that we have the ability to operate the infrastructure, but also successfully monetize the infrastructure at a much larger scale than we had proven last year.
And this is a testament to the strengthening of our technology and talent, which has been used to build very deep in-house capability across the entire five layers of the stack and we have done major software improvements to improve the reliability, performance, and scalability of our GPU infrastructure in the last many quarters. And going forward, we feel that we have been saying this a lot that look, the future is all AI. This is a decade old story. Now given how the market all across the world are reacting to the AI infrastructure usage being increased very, very drastically so I think we have been validated in kind of putting up speculative GPU infrastructure. So where the world is divided between haves and have nots, we are the haves in terms of having access to speculative GPU capacity where capacity is running out everywhere. So that allows us to kind of continue our growth.
So with regard to where we see medium-term and long-term kind of growing in terms of the world shifting to GPUs from the CPUs, from rule driven software to AI driven software. So we continue to explore how we can build more infrastructure under our management. So where we have explored equity, where we have explored debt, now we are also exploring a variety of private satellite models to kind of bring in rapid expansion to our GPU capabilities. So we have consistently met the industry benchmarks for performance for inference or training or GPU deployments under our management and we expect our cluster one (B200) 1024 to go live somewhere in the mid May. And in a couple of months, we are expecting to kind of be able to deploy another cluster of 1024, which has already been planned. And then we are very strongly positioned to capture significant uptake in the Blackwell generation of GPUs apart from continuing to run very, very strong on the Hopper generation.
Apart from this, we continue to build a plan around B300, GB300 and Vera Ruben deployment and we will continue to keep increasing the GPUs under one management and under our direct deployment both through partnerships, through direct acquisition, and by exploring various financing models that can help us secure more GPUs. So we’ll continue to focus on that. We continue to focus on our end to end project management capabilities to deliver AI infrastructure to AI services and AI product companies, AI digital native enterprises in the BSSI and other sectors. So we’ll continue to work towards that. So I think that pretty much covers it. So this has been a good year and a good quarter so pretty much our strategy has been validated over last many quarters where we significantly continue to add to our technology heft in terms of having deep tech talent in-house to be able to do stuff. So we gave a certain guidance in terms of where we would be by the end of this quarter a couple of quarters back.
I think we have more or less met and exceeded that target and we hope to continue to do better in the future. So given that the market is growing very, very positively and it’s a broad trend. There can be again like we have mentioned in the past that there are [Indecipherable], but ultimately this is a market where the GPU utilization is going up, the demand for tokens is through the roof. So India being a country of more than like a billion people on phone and of course even if they were not on the phone, there is an inherent ability to treat every possible piece of information that is going in and out of people, in and out of machines, in and out of enterprise systems to be described in the form of tokens. So tokens can be representing voice, text, any other form of data whether invisible spectrum, audible voice, inaudible sound waves, ultrasound, or whatever. So everything is tokenized data for AI and I think India as a country generates a lot of data. So we continue to believe very, very strongly in the long-term India story of AI factories being built here and India becoming the AI factory of the world.
Now with that, I would like to kind of hand over for more numbers driven financial highlights talk to Nitin and then we’ll open it up for questions-and- answers. Nitin, over to you.
Nitin Jain — Chief Financial Officer
Thank you, Tarun. Good evening, everyone. Thank you for joining us today. I will walk you through our financial performance for the fourth quarter and full year for FY ’26 Q4 FY ’26 again witnessed a good quarter for E2E Networks. We have demonstrated our ability to scale AI infrastructure rapidly, achieve high utilization across GPU cluster, and convert capacity into strong revenue growth. Most importantly, we have been able to drive our EBITDA margin in Q4. So let me start with the quarterly highlights for the Q4. Revenue stood at INR956 million, which is up 186% year-on-year and 37% quarter-on-quarter. EBITDA stood at INR581 million with EBITDA margin expanding up to 60.7%. Profit before tax turned positive at INR86 million compared to a loss of INR75 million in Q3. PAT stands at INR64 million. This performance reflects INR120 million swing in profitability sequentially driven by operating leverage and a strong execution.
EBITDA margin expanded by 413 basis points sequentially. At the same time, depreciation increased to INR513 million in Q4 reflecting ongoing infrastructure investments. Despite this, we achieved positive EBIT and positive PBT. For the full year FY 2026, revenue stands at INR2456 million, which is up 50% year-on-year. EBITDA crossed INR1263 million, which is up 30.6%. However, we reported a PAT loss of INR156 million, which is driven entirely by the depreciation on our GPU infrastructure investment. Our core business remains strongly cash positive, operationally profitable at EBITDA level. As utilization continues to ramp up, revenue will progressively outpace depreciation, improving profit — reported profitability. To conclude, Q4 demonstrate our infrastructure investments are translating into revenue and we are firmly on a path for growth.
Thank you for your time. Now I would open the floor for the questions.
Questions and Answers:
Operator
Thank you, sir. Ladies and gentlemen, we will now begin with the question-and-answer session. [Operator Instructions] First question comes from the line of Bhavya Gandhi from Bajaj Alternate Investment Management Ltd. Please go ahead.
Bhavya Gandhi
Yes, hi. Thanks for the opportunity. So first question is regarding the asset light model that we are looking out for with L&T. Could you throw some light how will the asset light partnership fructify what sort of arrangement we have?
Tarun Dua
First of all, of course we have a MoU with L&T to monetize the GPU infrastructure that they are building. So now that is still in exploratory stage. So as the GPUs get deployed and we start into — start the work into monetization and as we know more and more, we will talk more and more about that. That being said, this is not an exclusive arrangement so we will continue to operate at an arm’s length with L&T and we will continue to explore other partnerships of similar or different nature as well. So there is a long gap between say pure debt versus pure equity. So there are a number of structured possibilities in terms of figuring out that how to finance the GPU in partnership with a lot of different types of partners with differing objectives. And we’ll continue to explore all those partnerships and as and when we kind of reach a conclusion on any particular partnership, we’ll obviously keep everyone informed.
Bhavya Gandhi
Sure. But will it be like margin accretive or because the capex is going to be on other partners as well?
Tarun Dua
It’s too early to say how would the numbers apply. So obviously it’s very very early to say how those things would look like. So unless the structure is finalized, these would — see, obviously we all work for a profit so we’ll not do anything which doesn’t earn us the profit. So that being said, so how to put numbers to that is not something we can do today.
Bhavya Gandhi
Got it, sir. And on the MRR, would you like to guide for the next year because you’ve been doing this capex from your own pocket also this year so if you can?
Tarun Dua
So we’ll continue to kind of focus on growth just the same as rest of the market is doing. We want to be very, very growth focused. So that being said that, I think it would not do justice to do a MRR guidance for one year in the future. I think the future is changing very very rapidly week on week so it is obviously impossible to give a week on week guidance on MRR and it would be better to kind of watch it quarter on quarter than kind of predicting four quarters in advance.
Bhavya Gandhi
If you can just throw some light on the asset turn for the capex that we are doing at least some sort of understanding we can get in terms of capex.
Tarun Dua
See, I think we want to get the framing away from things like asset turns. So essentially don’t look at us as an asset monetization business, look at us as a technology business. So I think there was a recent conversation around the tokens becoming more valuable. So same set of tokens that were being generated by say open source AI or closed force AI like what used to produce say x amount of value, we are increasingly seeing that as the accuracy efficiency capabilities of AI increases, then the generated tokens also become more valuable for businesses who have figured out how to utilize those tokens in their business. So which means that kind of setting up ourselves for looking at the value of our outcomes based on the size of our assets doesn’t do justice to the business. So ultimately we want to get away from saying that okay, this infrastructure produces only X amount of value. So we will continue to explore options for how to produce more value from the same infrastructure by figuring out where the tokens are more valuable.
Bhavya Gandhi
Got it. Fair enough. Okay, sir. That’s it from my end. I’ll get back in the queue. Thank you so much.
Operator
Thank you. Our next question comes from the line of Keshav from Niveshaay. Please go ahead.
Keshav Sureka
Yes. Thanks for the presentation and congrats on the great set of numbers. Hi sir, am I audible?
Tarun Dua
Yes, yes, please go ahead.
Keshav Sureka
Sir, like we procured the Blackwells way back in the quarter and no, they just did not lie. So what’s the reason behind the delay? Are we facing some kind of procurement issues or something from the Nvidia?
Tarun Dua
See, global supply chains have been impacted somewhat so it’s always — the delays are always like for want of a horseshoe nail. So kind of we have been working very diligently towards making sure that everything is planned out and everything gets delivered on time, but then you can’t control every single component and sometimes some components can be the most critical ones from the point of view of getting the entire deployment done. So we are targeting the first deployment to go live before mid of May. So keeping our fingers crossed over there.
Keshav Sureka
Got it. So this quarter that GPU demand has been strong globally and we heard that all the GPUs are sold out, including the old ones. So how are you seeing this trend in the domestic market? So are you also experiencing the same or is it different situation in India?
Tarun Dua
Okay. I don’t think I have understood your question, but let me try to answer what I have understood. So see, obviously it has turned into — somewhat into a much better market than it was a couple of quarters back especially for people who have the GPU infrastructure in place. That being said, as a company — as an Indian company, we are very, very focused on India first. So wherever we get an opportunity to support Indian companies, we definitely like to do that. Now with that in mind, basically we always prioritize India first and then we are also happy to support the global infrastructure needs today. I hope the answer is in the right direction towards what you were asking.
Keshav Sureka
Yes, definitely. I was just trying to ask like demand has been strong globally that the GPU goods got sold out. So I was just trying to understand…
Tarun Dua
There is fresh we are seeing demand both India as well as globally. It is strong both ways and we definitely try to prioritize India. But then wherever the demand comes from, eventually like we are happy to fulfill it.
Keshav Sureka
Got it. And one last question from my side. That employee cost increase this quarter so if you could give some color on that like what’s the driver behind this rise?
Tarun Dua
See, as we grow, we kind of start figuring out more and more interesting problems that need to be solved in this space. Obviously our goal is to go toward what you call higher value tokens and that obviously requires the application of quite a high level of talent. So as we obviously grow, the base effect would still be there. But on the other hand, we want to maintain a balance of not losing out on future opportunities because we didn’t invest today on what was required 12 months later or 18 months later or two years later.
Keshav Sureka
Got it. And one last thing, if you could also provide the GPU utilization level slate if possible for this quarter?
Tarun Dua
See, broadly across our entire infrastructure not just GPUs, I think we are looking at 80% plus utilization. So certainly across the quarter now I would say only in the March month, we are looking at like 80% plus kind of utilization, certainly less than 85%. So there is elasticity in terms of basically how we can increase the utilization further over there.
Keshav Sureka
Got it. So I will get back into queue. Thank you so much and come back again.
Operator
Thank you. Our next question comes from the line of Bharat Gulati from Dalal & Broacha. Please go ahead.
Bharat Gulati
Yes. Hi sir. Just wanted to get a breakup of what kind — what is our MRR breakup currently? Can you give a split between India AI Mission Enterprise and also a split between what would the GPU contribution of that MRR be as compared to CPU?
Tarun Dua
Yes. See, I think broadly the way things are progressing like in another couple of quarters, I think GPU contribution would exceed closer — would be closer to 85%, 90% over coming quarters. That being said, all these numbers are fairly dynamic. So in the sense that they vary very, very rapidly from a week to another week so I wouldn’t like to kind of — and the sample size today is fairly small. So what’s the split is something that we don’t want to worry about today. So as we grow the cluster sizes to a much larger number, then it would start making sense to kind of talk about okay, Hoppers, you’ve got a couple of thousand of them; then Blackwells, we’ve got a couple of thousand of them. That’s the point at which it would start making sense to say okay, what’s happening on Hoppers, what’s happening on Blackwells? Today, the entire universe of GPU, CPU, storage; all of that stuff that we have is fairly, fairly small to kind of put a very, very strong split around those numbers today.
Bharat Gulati
Sir, but could you just give a MRR split between the India AI mission and our enterprise or our SME clients? Just trying to get an idea of what kind of visibility do we have in that INR374 million MRR. And also just to add on to that before that, our MRR for the previous two months if we average it would come to like a INR290 million kind of MRR. So just trying to understand that spike that came in the last month, what was the reason for that spike and also how sustainable is this INR374 million?
Tarun Dua
It’s broadly an increase in overall utilization that has led to the increase in the spike in the March month. From the perspective of like overall split, I think the NDIA kind of — so basically the overall government business across this quarter has not exceeded more than say 35%, 40%.
Bharat Gulati
All right. So sir, could you — is there some particular reason for that? Because ideally our missions were — our India AI mission was supposed to start in the month of Jan. We are in March as of this quarter’s report reported numbers. So what kind of ramp up are we seeing in that? And just any kind of timelines and what kind of revenues in terms of MRR do we expect from that?
Tarun Dua
Where we have always looked at India AI mission is that look, we are trying to obviously, like we said, we are supporting India, we are supporting India AI mission, we are supporting Indian companies, and we definitely give a preference to all those workloads. But that being said, whatever capacity is not uptaken so we are not trying to kind of force anyone to take any capacity from us. So we are very, very happy to sell it to outside the scope of either the AI mission or the government workloads. So we have enough customers across multiple segments to kind of solely rely on one revenue driver growth. But that being said, we continue to work very closely with the AI mission for the coming generations for GPUs as well and we are hopeful that we will continue to collaborate, continue to work together, continue to solve problems.
Bharat Gulati
So just trying to understand that India AI mission was almost all our Hopper — Hopper series GPUs were given to India AI mission. We’re seeing 80% utilization, but India…
Tarun Dua
No, we have not scaled up that to that extent. So majorly whatever was the scale required by the allottees of AI mission, we have provided that scale, but we have not insisted on them having to increase their scale without having the need for that. So we are quite okay with that.
Bharat Gulati
So in terms of our revenue currently, what kind of clients are we currently catering to and what’s the pipeline? You said mid May we expect Blackwell to get deployed. Do we have some firm orders?
Tarun Dua
See again, this is a very, very small base. So it’s like kind of like okay, there are GPU customers everywhere in the world. There are GPU customers in India, there are GPU customers outside India. Every possible segment that you can think of whether it is enterprise, whether it is BSSI, whether it is education; everyone needs GPUs. So it’s practically like — basically like almost first come, first serve today. So the nature and profile and putting a stat on okay, this much of our business comes from here, this much of a business comes from there is kind of like very, very premature. I think we need to grow the base of GPUs to a couple of tens of thousands before these numbers start to make sense. So today at a very, very small base, none of these numbers would make sense to kind of track and then kind of say that okay, what’s the kind of number this quarter? It will change next quarter.
Bharat Gulati
So would it be fair to say that until we don’t build a huge base of GPUs, our MRRs will be lumpy in nature? Is that what you’re trying to iterate at?
Tarun Dua
No, I think the demand has secularly changed in the GPU world. So from the era in which the lumpiness was there because our base was even smaller, I think we have come a long way over there. That being said, the lumpiness will continue to decrease as we continue to build more and more volume of GPUs under our management.
Bharat Gulati
So sir, out of the INR374 million, what exactly would be long-term MRR that at least we are seeing the visibility for the next eight to 10 months or a year or so?
Tarun Dua
We are not very, very focused on — so we have a balance of basically thinking about okay, what needs to be six months, one year, month on month, which is pre-emptible. So we are doing kind of a whole series of them. So it’s a week on week effort to kind of figure out that okay, what is the capacity you want to sell hourly? What is the capacity you want to sell weekly pre-emptible? What is the capacity you want to sell monthly? What is the capacity you want to sell yearly? That being said, wherever larger customers are involved, we try to do at least six months to one year visibility with our customers. And increasingly for even larger customers, we are trying to look at even longer visibility to at least two years to even going up to three years. But that being said, our view is that as tokens become more expensive to generate and kind of they generate more value for people, it is better to kind of not put all the eggs in the long-term basket, but have like a judicious mix in very short-term contract and medium-term contract and some degree of long-term contract.
Bharat Gulati
Got it, sir. And just one last thing, sir, in terms of the Blackwell that we expect to get deployed in May, do you — can you give any idea in terms of who will be the customer? Will it be an enterprise India AI mission and SME? And when would that customer’s revenue start to come in? Would it come in from May or will there be a lag?
Tarun Dua
We will announce that after closing that particular deal instead of saying anything speculative today.
Bharat Gulati
Fair enough, sir, Fair enough. Thank you. That’s it for my side, sir.
Operator
Thank you. Our next question come from the line of Nishant Joshi from Equisense Advisors Private Limited. Please go ahead.
Nishant Joshi
Good evening. I have one question. As you said, company is building computer assets using…
Operator
I’m sorry to interrupt you Mr. Joshi, but your voice is not audible.
Nishant Joshi
Is it audible now?
Operator
Yes, please proceed.
Nishant Joshi
I want to say that as company is planning to build assets using debt as well as equity and subsequently intend to go for the asset light model also. So will this change our business model means instead of providing hardware, we would be providing more of service? We’ll be offering our TIR platforms. So will this lead to a change in our revenue number?
Tarun Dua
Think of this as the expansion of the number of business models without shutting down any of the existing business models. So we will continue to own, continue to operate GPUs, we will continue to acquire new GPUs, we will continue to expand the partnerships to kind of bring increasing number of GPUs under our management available through our TIR stack platform which operates at multiple layers where it is the choice of the customer to what layers they want to buy. And kind of nothing is changing from the point of view that okay, we are not going to do this or we are not going to do that. We are simply saying that we are going to do additional number of things to expand the universe of our thinking to do a lot more than what we are doing today.
Nishant Joshi
So sir, it would be an ideally additional line of revenue you mean to say?
Tarun Dua
Absolutely.
Nishant Joshi
That was my query, sir. Thank you.
Operator
Thank you. [Operator Instructions] Our next question come from the line of Deepak Poddar from Sapphire Capital. Please go ahead.
Deepak Poddar
Yes. I’m audible, sir?
Tarun Dua
Yes, you are.
Deepak Poddar
Okay. Thank you very much, sir, for this opportunity. Sir, just wanted to understand now at a March MMR — MRR of INR27.4 crores, so what is our capacity utilization and on what capacity?
Tarun Dua
So this is closer to overall capacity utilization of around 80% in the March month.
Deepak Poddar
And on a capacity base of 3,900?
Tarun Dua
Yes. So this is like GPU, CPU, storage, all capacities put together. So the utilization is closer to 80%.
Deepak Poddar
Closer to 80%. And I think we have around 2050 lined up in next, what, six to 12 months, right?
Tarun Dua
2048 plus some shares that we are planning to deploy this financial year starting from May. The first lot of 1024 is expected to go live.
Deepak Poddar
So by FY ’27 end, we would be targeting around 6000 of capacity.
Tarun Dua
I wouldn’t want to place a limiting number over there, but you could say that that’s the minimum number.
Deepak Poddar
At least, at least, at least that. Okay, understood.
Tarun Dua
Yes, that part is already visible. So that’s what we’ve already spoken about.
Deepak Poddar
And you mentioned around 80%, 85% utilization by March. I mean you’re talking about by March ’27 of this expanded capacity.
Tarun Dua
No, this is the previous bond by March.
Deepak Poddar
Previous March only. Okay, understood. And my second question is on your fixed assets. We have got around INR1500 crores order fixed asset, right, and last two years capex has been around close to that only. I mean so this entire is a depreciable asset. I mean would that be a fair assumption?
Tarun Dua
I would let Nitin put a perspective on that.
Nitin Jain
Yes. For the entire debt, fixed block is a depreciable asset which constitute of the new B200 which is currently displayed as CWIP.
Deepak Poddar
Correct. And what is the amortization schedule for this? I mean is it three to four years?
Nitin Jain
Six years. Six years.
Deepak Poddar
Okay. And what would be FY ’27 capex target?
Tarun Dua
Like we said, we are already planning to deploy 2048 B200. So that is already in place as far as our loan plans are concerned. That being said, we are not limiting ourselves to those plans. So we continue to deploy capital under various business models judiciously to expand the GPU footprint like fairly rapidly.
Deepak Poddar
Correct. So any number you have, I mean what is — I mean I think FY ’26 was close to…
Tarun Dua
I think look back — a look back is more important than putting a number in today.
Deepak Poddar
Okay, understood. That would be it from my side. Wish you all the best.
Operator
Thank you. Our next question comes from the line of Varun Gandhi from Finavenue Growth Fund. Please go ahead.
Varun Gandhi
Hi Tarun. My question is on GPU rental prices and are we seeing pressure on them? And I ask this from two aspects. The first aspect is technology is advancing rapidly so we see the Ruben architecture has been disclosed by Nvidia, which is much more cost efficient for inference compared to your — even your Blackwell architecture. And number two is the competition that’s been ramping up across domestic competitors, right? So are we witnessing any sort of pressure on GPU rentals right now or do you foresee that in the near term?
Tarun Dua
See, I think broadly the trend today is that there are not enough GPUs in the world that people want to buy and deploy. So that’s the current trend. Now whether this trend remains for how long is hard to predict. But the broad secular trend has been there that basically there are AI believers and there were AI non-believers. I think the proportion of non-believers is slowly dwindling and it’s quite clear that basically — so we are barely scratching the surface in the terms of utilization of AI. We have looked at so many of our customers, so many of the organizations we work with, including our very own so where we say that okay, is it the case that our AI utilization is to an extent that we want it today? The answer for ourselves that we increasingly get is no, no, this is going to be more like 10x, 20x, 50x of where we are today in terms of how we are utilizing AI. So AI is not coming out of the IT budget. AI is essentially coming out of your budget for the capabilities that you’re building for the business. So that being said, I don’t think the demand environment in the foreseeable future is going to be changing negatively for long periods of time. I don’t think that’s going to be the case from the visibility we have today. So we don’t see that there is any negative pressure on the pricing today. So in fact there are I think good set of tailwinds which are slowly inching up the prices rather than decrease in pricing.
Varun Gandhi
Got you. So from a demand standpoint, we aren’t seeing any sort of negative. But from a technology advancement standpoint, let’s say the Rubin architecture comes in live this year and it should, would that…
Tarun Dua
See, there is a lot of workloads that will go into production where they will continue to operate on the targeted architecture. So typically it’s, what you call, like a funnel kind of a growth where the next layer of the funnel is bigger than the previous layer of the funnel. So which means that the demand for existing GPUs will continue to remain very, very strong. So demand for even the NPO series has remained strong, Hopper series has remained strong.
Varun Gandhi
Tarun, I understand the demand has remained strong. But let’s say two years ago, the GPU rental prices for a Hopper H200s, are they same today or they are lower — 30%, 40% lower than what they were quoting at two years ago?
Tarun Dua
See, I think taking the case of two years ago is slightly anomalous from the point of view that I think that was one particular year in which the GPU demand went from say 1 to 10. So I think that capability driven shortages where the capability to print that many GPUs had to be built very rapidly so that created I think a local maximum of pricing. But broadly we are seeing that there is stability in the pricing for GPUs over extended periods of time. Typically we feel that it is possible to kind of definitely use and utilize the GPUs over a seven, eight year period. So we don’t think that the GPUs are like a two year story or a three year story and then kind of there is something available at a much lower cost than you’re able to shift and those shifting costs are like very reasonable. So that’s not the case. Your entire test debug cycle for doing AI on a new chipset is like a whole other set of expenses that you don’t want to do if your job is getting done by a GPU, which is available at a very reasonable price.
Varun Gandhi
Got you. Now my second question is at the beginning of the call, you touched upon how you’ve significantly improved your software architecture. Could you just give me a brief — if you could just highlight a brief example of how you’ve done that. You iterated that you’re now trying to capture high value tokens. Is there some sort of example that you could showcase?
Tarun Dua
So this is broadly from the point of view is that the tokens themselves today are more valuable. It’s just a point of choosing the right set of tokens that people are going to utilize. So it’s not kind of like a strategy statement to go and capture one particular set of high value tokens today. It’s a broad futuristic outlook that look, ultimately we have to go and figure out which are the high value tokens. Now that being said, from a platform improvement perspective, we have been working on things which are required by people running training on larger clusters, people who are running inference how to support them, people who want to build rapidly the pipelines reliably across larger number of GPUs and understanding how what is it that the AI data scientists do probably been there done that perspective. We have made improvements in all those areas for productization of our TIR stack.
Varun Gandhi
Correct. So just as an extension to this question, see we’re very well positioned to capture the sovereign AI tailwind and my understanding is that BFSI companies would be a key client base over there. Is there any sort of strategic development that you could share where you’re partnering with some sort of SaaS companies or other software companies and trying to offer bundle of software plus compute? Is there something you could help us with?
Tarun Dua
Yes. So as we do something, we’ll definitely share that with everyone over here. So as things happen, we’ll do a look back and say that okay, this is what we’ve done.
Varun Gandhi
Okay, But there’s nothing that you can share in terms of your thought process over here.
Tarun Dua
See, we are — we continue to drown in opportunity. What we capitalize on, we will come and inform the market.
Varun Gandhi
All right. Thank you very much.
Operator
Thank you. [Operator Instructions] Our next question comes from the line of Abhishek from InCred Equities. Please go ahead.
Abhishek Shindadkar
Yeah. Hi sir. Thanks for the opportunity and congrats on a great set of numbers. Sir, my first question is about the guidance commentary. If we summarize your comments on the demand, it appears to be really strong. We have a visibility of 2048 GPUs and then we said that we don’t want to get fixated on the guidance. So just trying to understand that the purpose here was not to get fixated on the number rather than any worries in your mind about the current macro. Is the understanding right?
Tarun Dua
Yes, we are not worried about whatever numbers we give whether we will be able to meet them or not. I think it’s like we don’t want to give a underwhelming number without first exploring over — week after week over like 52 weeks what we are capable of doing. So it will become sort of a limiter for us ourselves to say that okay, we are only to meet this number. So that’s why we are not putting a number today.
Abhishek Shindadkar
Perfect. That’s very helpful, sir. The second question is on the pricing. So contrary to the prior question, the recent articles are suggesting that the GPU spot rates for April have gone up substantially higher. So for an H100, the prices are up anywhere by 25% to 30%. Can you just…
Tarun Dua
We are seeing that, we are seeing that. We are looking at that. That being said, obviously we tried to capture some of that. But then that being said, we are also trying to support the Indian companies and India as a country where we want to have a balance between long term and short terms. So short term of course we want to make more money and long term of course we want to work with the kind of customers who will sustain and grow on our infrastructure over a long period of time.
Abhishek Shindadkar
Perfect. So just last one data point from Nitin sir. So the depreciation for the next quarter should be up by around INR25 crores a quarter, right? Is that assumption right?
Tarun Dua
I’ll let Nitin answer the questions around the depreciation. Nitin, over to you.
Nitin Jain
Sorry, could you repeat again?
Abhishek Shindadkar
Sir, the depreciation for the next quarter should be up by — should go up incrementally by INR25 crores a quarter. Is the assumption right, sir?
Nitin Jain
Why do you expect that?
Abhishek Shindadkar
Because of the B200 cluster.
Nitin Jain
B200 cluster would be that we are saying the timeline of mid May so that’s what. There would be an increase in that depreciation, but not to that significant effect what you are telling us.
Abhishek Shindadkar
Okay. Perfect sir. That is very helpful. Thank you for taking my question and best wishes for the next year.
Operator
Thank you. Our next question comes from the line of Neil Munot from PICO Capital. Please go ahead.
Neil Munot
Yes. Hi. Thank you for the opportunity. Sir, my first question is building on the previous participant. The capacity utilization is…
Operator
Sorry to interrupt you, Neil, but your voice is breaking.
Neil Munot
Is it better now?
Operator
Yes, please proceed.
Neil Munot
Yes. So the first question is basically building on participant’s question is that the 80% utilization that we’re talking about, I’m assuming this is excluding the new B200.
Tarun Dua
Yes, that is excluding the B200 obviously.
Neil Munot
Yes. So when we talk about the 80% utilization with respect to the remaining capacity available, have we kept these aside for POCs or demand is still ramping up in that sense?
Tarun Dua
No, the demand is still ramping up over there. So it’s like all sorts of different sorts of capacity. So they all work in different combinations. So an ideal utilization number could be kind of like it depends on basically where do you end up, but that’s not a limiting factor on the MRR from a very straight line perspective. So for instance like the same particular piece of hardware could under different circumstances we build at wildly varying rates.
Neil Munot
So what is my assumption?
Tarun Dua
These are all averages essentially. So these are all important from a look back perspective. So what — this is our assumption on that. Look, we are doing what is sort of like an 80% capability. Now 80% could be 80%. 80% could be maybe 85%, 80% could be even 75%, maybe even 70%.
Neil Munot
Okay. Understood. And sir, with respect to the demand scenario in E200, have we closed India AI mission?
Tarun Dua
We have not closed any deals on the Blackwell capacity which is going up as yet. Obviously once we close the deal, we will kind of inform everyone over here.
Neil Munot
So the reason for this is the reason ramp up or is the reason that we are still closing POCs or that’s it?
Tarun Dua
So there is a lot of interest in the Blackwell capacity that we are building. So kind of we are working towards with multiple customers to figure out what sort of problems they need us to solve for them and what is it that they are looking at from us. And we obviously want to work with the customers that are best placed to utilize what we have to offer.
Operator
Neil, please rejoin the queue if you have more question. Thank you. Next question come from the line of Hardik Gandhi from HPMG Share & Securities. Please go ahead.
Hardik Gandhi
Hi sir. Thank you for the opportunity and congratulations on a set of good numbers. Just wanted to push on a question asked by the previous participant. I think so majority of the investor group just wants to know whether the revenue uptake is due to any lumpy nature or are you seeing any sustainable growth in the long run? Because in the last December we suddenly dropped — sorry, but we suddenly dropped the numbers. We stated that the startups were no longer wanting our ecosystem, right? So we just want to know from a long-term perspective or at least from a…
Tarun Dua
Sure. Let me try to answer that question as best as possible. So see the lumpiness is a part of the nature of the business. So the lumpiness will keep going down as we keep increasing our GPU base. Now that being said, we are currently seeing the kind of customers who are buying these GPUs have sustainable long-term workload. So these are not of the short-term nature for most part. So we do expect to have sustainability for these workloads over the medium term. So I don’t think there is going to be a massive amount of lumpiness in the near term. Medium term of course as the number of GPUs grow, I think the effect of lumpiness would be far more muted than it has shown up in the past in the early days of us building the capacity.
Hardik Gandhi
Right, right. And the second question from my side was to understand again on this part itself that what percentage of revenue are we allocating to a long term MRR — sorry, long-term contracts so that our MRR…
Tarun Dua
It’s of a dynamic nature. So basically it is not possible to predict.
Hardik Gandhi
Yes, we are not predicting. I’m just saying that from a…
Tarun Dua
We are not setting any hard lines over there. There are I think far more number of variables to consider than it would be possible to describe over here that what to pick up when in terms of customer interest. So I don’t think it would be — I would be able to give like a very fair picked answer to that question. So we are definitely interested in having some percentage of long-term contract. What that percentage will come out to be is something we need to look at it on a look back basis when the size of — the size and scale of the GPU installation is sufficient. So I think in the short term there is no clear cut answer to that.
Hardik Gandhi
Right. Just a small data point. What percentage of revenue did come from customers outside India versus domestic demand for the last quarter?
Tarun Dua
I’ll let Nitin take a stab at this question.
Nitin Jain
So the international revenue for the last quarter is roughly around 35% odd, 35% to 37% is the international customer revenue.
Hardik Gandhi
Okay. Thank you. That’s really helpful. Appreciate the time.
Operator
Thank you. Our next question come from the line of Srinivas K from Trust Investment Advisors Private Limited. Please go ahead.
Srinivas K
Hi sir. Am I audible?
Tarun Dua
Yes, loud and clear.
Srinivas K
Yes. So sir, what percentage of Q4 revenue came from inference workloads versus training and what would be the future split up of this mix at least directionally sir?
Tarun Dua
So I think there is a lot of confusion about what consists of inference revenue versus training revenue. I think these are not super hardline. Sometimes the same set of GPUs are used by the customers for kind of bursting up the inference workloads when the inference workloads are high and at a time when the inference workloads are low, they get converted to training. So the workload management is far more flexible today to kind of point to okay, what percentage of inference, what percentage of training today. And I think again at a much larger scale like these numbers would become clearer. So today of course it’s hardly possible to kind of super differentiate between okay, who is using that 1 GPU, 4 GPU, 8 GPU, 16 GPU cluster either for training or for inference. But that being said, majority of revenue is still closer to training than to inference.
Srinivas K
Okay, sir. And you said 6,000 GPUs by FY ’27 at least minimum, right? So what would be the upside sir? Is it — could you deploy 8,000 on upside — 8,000 to 10,000 GPU?
Tarun Dua
I don’t want to put any numbers right now. Let’s look at those numbers as a look back over coming period of time rather than putting out a number today.
Srinivas K
Okay. Thank you and best of luck.
Operator
Thank you. Our next question comes from the line of Rohan Nagpal from Helios Capital Management. Please go ahead.
Rohan Nagpal
Thanks for the opportunity and congrats on a good set of results. My first question was so you said that there is a variety in terms of your current MRR being short term, medium term, slightly long term. Could you provide some directional commentary on how much of that MRR is six months plus or one year plus?
Tarun Dua
I think one is we are still on a very small base of GPUs so kind of we are not doing any hard split in terms of where we want to be or where we are today and it’s fairly dynamic and rapidly changing both with new capacity coming up online. So all these numbers are subject to very rapid and very wide amount of change. So these do not add to any understanding for anyone today.
Rohan Nagpal
I mean if there is a 20% to 25% increase in the spot price, then capacity that is contracted out for six months or one year will not be in a position to take advantage of any increase in realization. So I think it would be somewhat beneficial for someone looking at it from the outside to get a sense of some amount of revenue uplift or exposure to increased realizations.
Tarun Dua
So that’s a very granular aspect of understanding that you are seeking on a very small size of the infrastructure installation base compared to the global peers. So I think we would like to avoid this conversation for today and as we grow in size, then we kind of talk more about that. Now these numbers actually mean something and they are making sense to someone on the street. So today, these numbers would not make much sense.
Rohan Nagpal
Okay, sure. Thanks a lot. And then the second question was. So you talked about higher value tokens so could you provide some color on which industries are sort of taking advantage of higher value tokens from your vantage point?
Tarun Dua
Not really. I think it is more — it’s more closer home than that in terms of okay, what’s the higher value token? So asking a question like okay, what is closer to a search query is like a reasonably low value token. Someone using the tokens to solve this problem that is paying the bills for a company is like a completely different value conception for those tokens. So answering a basic transactional support query is like a reasonably low value token. Being able to answer a more complicated query or being able to confirm a more complicated transaction with a customer on call is like a higher value token. So there are infinite number of variations of how people use AI and the key is to find those customers who are utilizing the higher value tokens where they are not worried about okay, what’s my ultimate per hour GPU price?
But what’s the cost of my solving a particular problem and generating a higher ROI? So again we are not directly involved all the time in identification of higher value tokens and kind of being able to strategize in a very, very static manner that we should go and sell to this customer, they’ll probably pay higher for this particular GPU. It’s a broad industry trend where we are seeing increasingly the demand coming from customers who are not price sensitive because of their ability to generate or consume higher value tokens through a combination of their own software or proprietary or open source software that is available extant in the market today.
Rohan Nagpal
So in that case, what is E2E’s role in this high value tokenization? I mean that would just be an outcome of if the person who is renting a GPU or taking on GPU capacity is able to extract higher value, they’ll be willing to pay more, right? So the price discovery could just take care of who has — who is generating higher value tokens. I wasn’t too clear on how E2E is sort of working towards higher value tokens.
Tarun Dua
See, how you are able to quickly get to that point of being able to generate and utilize those tokens for your business. It’s fundamentally how to build on various parts of the software and hardware stack to make it available quicker. I think that’s where the role is played by E2E.
Rohan Nagpal
Understood. Okay. That’s it from my side. Thank you very much.
Operator
Thank you. Ladies and gentlemen, due to the time constraint, that was the last question for today. I would like to hand the conference over to the management for the closing remarks. Over to you team.
Tarun Dua
Yes. Thank you to all our investors, all our team members, all our customers, all our ecosystem partners for all the support you have all extended to E2E network over last many years. It’s been an overwhelming amount of love for us that we are seeing in the community and overall community. And we wish you all a great day ahead and thank you for patiently listening to our call. And we would like to kind of once more thank everyone and end this call.
Nitin Jain
Thank you.
Operator
[Operator Closing Remarks]