Thank you very much. Thank you, everybody. Very nice to talk to you today. I'm with Denis, and today we will comment you the results that we announced yesterday evening. Let me share the agenda for discussion today. As usually, we will start by sharing a high-level summary of what we do for those who do not know as much, and then we will go through all the announcements that we made yesterday that will include the announcement in term of revenue. We will also comment the one contract that we signed, the major media customer. I will come back on the announcement of our new generation product called K2, and also say a word about a couple of funding projects that we signed during the last period.
After this comment about last announcement, I would like to spend some time to also give a little more background and context about what you've certainly seen in the past few months in terms of awareness about the importance of AI in our domain, and also in terms of edge computing. I would like to explain our view, and also why this has a very positive impact on Kalray, and why Kalray is definitely at the heart of our vision now in the industry. I will come back on the objectives that we set for 2023, and we should have some time, of course, to answer any questions you may have. Let me start with a very rapid summary of what we do.
As you know, we are a semiconductor company, a typical deep tech company with a very, very interesting technology, which is result of an intensive investment in the past years, but technology which is now fully mature. What we provide to industry is a new type of processor. This processor are called DPU. I think the industry now realized that there were three type of processor that will be at the heart of our end screen in the future. The first one is the traditional one, it's called a CPU, the one you have in your mobile and your PC, and with company like Intel and Arm providing the CPU. The second one is called a GPU. GPU, as you know, with NVIDIA, especially for graphic and intensive calculation.
The industry has defined a new term called DPU for that type of unit. It happens that Kalray is one of the pioneer and one of the leaders in the DPU space. We designed this DPU, and we built in the last few years solutions around these processors in order to meet the demands of a growing market, the market of data-intensive applications and of next-generation data centers. In terms of company, we're about 20 people, and quarters in Grenoble, with teams in UK, in US, and a few partners in Asia. As you certainly remember, we've got support of very strong and well-known industry investors, including NXP, Renesas, Mitsubishi, Safran, Mediaip. A couple of very important company who've been following us and supporting us for now many, many years. Just to restate our vision, our mission.
Our vision is pretty simple. In our modern societies, it will be more and more the case in the coming years and decades, data is key. Any company in this world, to be successful, we have to treat and to process the data available in a faster and better way as possible. To do so, and this is our mission, we want to become a global leader, providing the hardware and software technology solutions for these data-intensive application, and to help this vision, to help our customers to be able to get faster and better insights of this data. What do we provide to the market right now? As you understood, of course, our core product is our processor, and we've got a family of processor called the MPPA family.
We believe in the last few years, complete solutions, so that we have some a solution ready to use for our customer. It goes from what we call cards that can be integrated into servers, but also software, and we invest massively in the software, because software is very, very important to help our customer integrating our technology. Software, including also two very important product that we actually acquired from acquisition of a company early last year called pixitmedia . Now we've got a complete solution, as you know, to process data smarter, more efficiently, and more securely, work wherever the data is. Let me now really focus on what we announced yesterday. You will see a lot of announcements showing that we made a lot of progress in that six months.
2022 was really a year which we changed to mention. I would tend to say in summary, that H1 is really moving forward with the same dynamic and with the same rhythm. Starting with the revenue which is very growing very fast. We announced yesterday that our revenue for H1 was EUR 15.3 million, which is definitely a huge growth compared to the EUR 4.8 million H1 last year. If you compare to the total consolidated revenue of last year, so in 2022, the total revenue was EUR 16.4 million. You see that we nearly reached the same level of consolidated revenue of 2022 in 1 semester.
Which is, of course, we show the dynamic that we have. The dynamic also is noticeable also in our pipe. We've got a very strong pipe, which make us very confident that this growth and this acceleration will continue in H2, as you will see later in the presentation. This is especially due to a strong increase in revenue due to our what we call solution sales. We see a large success of our solution sales. What we mean by solution sales is the combination of both our hardware and software solution. That we've got a package solution, easy to use, very efficient, combining hardware and software, that we sell on the market, especially with partner like Dell, and we'll come back on that. Huge growth in terms of revenue.
If we zoom a little bit on this revenue, on this EUR 15.3 million, you will see that the balance between hardware and software is very similar to what we had in 2022, which once again is very, I mean, very logical since that what I mentioned is that we are selling complete solution, which include hardware and software. That means that you, we find the same type of splitting between the hardware and software parts. In term of regions, we see a very strong demand, especially US. You can see that actually the share of revenue from US has grown. We are about 50%, even still very well balanced between US and Europe, and in Europe, we include UK. Very limited revenue right now outside these two regions.
Main reason is because we focus mainly on US and Europe. We really do by little try to extend our business in Asia, but obviously this is not the priority as we see a huge demand and huge opportunity in this region where we are already growing very fast. In terms of application, we are addressing all the application which are requiring data-intensive workloads and heavy loads. AI, and we come back on that, is definitely a key driver. Health, media, research, all these verticals are requiring data-intensive technology and solutions, and that's exactly what we provide to these verticals. As part of this growth in terms of business, it's important that we wanted to highlight that we especially signed a major deal with a major media player.
Let me explain a little more what is behind this deal and behind this announcement. The customer is a very large US-based media company, and the type of product that this media company was looking for is very much in line with what I just described, in a way to manage and process a lot of media data, wherever this data are, whatever the workload they have, in a very efficient way. What the solution we provide to this customer is what we call ngenea2. ngenea2 is a overall platform for efficient data management for data-intensive workloads. It includes both our software solution for engineer, as well as our hardware data-intensive storage solution based on our DPU.
This contract is not what we call a jumbo contract in the sense not EUR tens of millions every year, as the one we announced a few months ago, but it's a very significant contract where EUR several millions are expected in 2023. A part of it has been already recognized, actually in H1, and with potential upselling in the next years. A very interesting contract, which once again, is a very good proof point of our capability to sign large contract and to offer a very interesting value to our customers, and of course, is one amongst them.
Just a contract I wanted to share with you because once again, it show the traction we have, and it show that whatever the industry, a lot of demand in the data-intensive application, we have very, very valuable solution for these clients. What about the Dell partnership? The Dell partnership, as you remember, has been announced and closed last year, end of H2. It's a very important collaboration for us, and I just wanted to share with you that this collaboration is progressing very well. As you remember, as a reminder, we are part of a program called the Titanium program, which is reserved by Dell true core partners, about 10 key partners worldwide in this program.
We've got a deal with Dell, so that Dell storage divisions, we sell our solution and promote our solution to their customers. Which is, of course, very important for us because once again, if you understand our strategy right now, a part of our strategy is to sell complete solution, especially to a large enterprise. I think the support and a recently agreement with company like Dell is the dealer, in the storage industry worldwide with over 20%, is extremely interesting for us as a huge sales force and huge capability to deploy our business.
We made last year a couple of million EUR already with Dell, we grew this business already in H1, and we expect a very significant growth of this collaboration and the revenue coming from this collaboration in H2. Very intensive activity with Dell. Huge demand, a huge opportunity, and this partnership is moving really in the right direction. Let me say a word about Coolidge. You may have seen that we announced the launch of what we call the tape-out of Coolidge 2. Let me give you a little more background. First, Coolidge 2 is a new generation of our processor. The one which is already sell today and which is actually part of our revenue, is called Coolidge one.
That been tape out in 2020. We've been working in last two, three years in a new generation called Coolidge 2, with a pretty large, and you will see next slide, improvement in terms of performance, in terms of power consumption. What we announce has been the tape out. What is the tape out? Tape out in our industry mean that the time where you basically push the button and send all the files to manufacturing, to our partners, TSMC, so that they can start manufacturing. Of course, it's very important at the time, because that's where you completely finalize the design. Of course, when you press the button, you need to be 100% sure that the design is right, it will not have issue.
That's really the outcome of very intensive work of the team, and very proud of what we've done. Once you press the button, then the manufacturing process start, and we will receive a sample. The first sample about three months later. We expect the first sample from TSMC manufacturing by beginning of September. What is very interesting with Coolidge two, actually there are two main improvement or main benefits of Coolidge two. The first one, as mentioned, is more performance, less power consumption, so improvement of what we had already. The second one is also in term of extension, new features, and the capability for us to address new market, to better address market we are already in.
On one side, you know that we're focusing on the storage market, Coolidge two include and integrate a number of additional features to be even more relevant on the storage market. It includes, for example, functions to be more efficient in terms of protection of the data, in terms of encryption, compression. They're very important features for storage market, which will make us a really be storage for the GPU storage solution in the marketplace. Then the second also very interesting additional features we have in Coolidge two, are features to address a new market we've been working on for, like, two years, which is called the edge computing market.
Why we will be able to address this market with Coolidge two, because we extended massively the performance of Coolidge, especially in term of AI, which make us extremely relevant, as you will see the next slide, for the two verticals that we focus on for edge computing, which are machine vision, you know that we already signed a very large contract for machine vision, and 5G and telecom. Let me give you just an example of the type of improvement we're talking about. This is one very specific example. It's benchmark. The benchmark is actually running a specific AI network called U-Net, which is very used, extensively used in the smart vision industry, in edge computing. Here you can see the type of improvement we reach comparing V1 and V2.
Let me explain it to you more with a table. There is a metric which is very important in our industry, which is called FPS. FPS means frame per second. That means basically the number of frame you can process per second, and frame is usually an image or a segment of video. You can see that for V1, for example, we were able to manage 175 frames per second in a data format which is called FP16. FP means floating point, FP16 is a very usual deployed format that is used to run algorithm. Okay, there is second format which is called INT8, integer 8, you can see that Coolidge two V1 was able to process 250 frames per second in INT8 format.
If you compare with V2, you can see that for the same type of format, same type of algorithm, same type of frame, rather than 175, we reach 1,400, and for INT8 formats, we go from 250 to 2,100. Which is an improvement between 8 and 10, which is massive. If you go to the metrics, which are actually more relevant than performance, which are the one used by the industry, which are the frames per second per watt and frames per second per dollar. What do we mean here? Frames per second per watt means the amount of frame you can process for 1 watt. Frames per second per dollar is the amount of frame you can process per dollar.
These are obviously the metrics which are used by the customer. It's about how efficient you are in terms of power consumption and how efficient you are in terms of dollar, in terms of prompts, in terms of cost. You can see here that if you compare data for V1 to V2, we typically go... For if you look at V1 frame per second per watt, we go 3.9-13.5, which is an improvement here again, between 8-10 in term of power consumption. If you look at in term of frame per second per dollar, that means that we can go from 0.2-2.1, here more than 10 improvement. Okay. Here again, a huge improvement.
If you remember when I described a few months ago, the reason why this large customer with what we call the jumbo contract, selected us, you remember the key metrics for that was especially frames per second per dollar. You can see why, I mean, they selected us because we were able to provide a unique performance in frames per second per dollar. You may remember that at the time I shared with you that the existing solution market, including solutions that use NVIDIA for edge computing, and that the reason why they selected us was because we were between 5 times better than the best competitor. Okay, you can see why, especially thanks to improvement, but we managed to reach with Coolidge™2 V2. Coolidge™2 V2 definitely a huge announcement for us.
The result of very, very significant effort from our team for like up to 3 years, and which allow to further extend our business in the storage industry, and to now be very relevant and one of the best solution for edge computing, especially in smart vision and 5G, which of course, will lead the signature of a new contract in the future. The last announcement we shared yesterday is about the IP-CUBE and also the Selection as France 2030. These are news that we announced a couple of weeks ago, but I think it was also important to re-summarize what we've done and what we're doing to be able to fund our project in a non-dilutive way for sure.
As you've seen in the past announcement, we've been selected, and we have been granted a project called IP-CUBE. IP-CUBE has been an intensive effort on our side to show to French administration how important is what we do, and to also show and to get the support for next generation of processor called Dolomite. We're very happy to be selected as part of this project, and to have a funding of about 10 million EUR that will be spread over the next three to four years to take part of spending of our next generation processor called Dolomite.
We have been also selected as part of France 2030, there again, I think it's a very good evidence of the support we get and the visibility we get from the French, but also European administration. Why is it important for us? It's important for two things. First one, of course, obviously, for us, is to get extra funding in a non-dilutive way, which is, of course, very important. The second one is in terms of visibility and the capability to collaborate with major players in industry. It brings us additional visibility and additional credibility on our industry, which is, of course, very interesting.
I believe this is important to show you the effort we do here and the support we get, and of course, we have other project in the pipe, and we'll continue having the strategy to try to get as much funding as possible in a non-dilutive way. I think I hope that you are more than welcome to comment and give more detail about this announcement. I think these announcements are great and show how fast we go and how unique player we are becoming in this very high growth market of data-intensive application. I would like to step back a little bit and share a word about what you have seen in the last few months.
I believe that you are more than aware of the importance of AI, so I wanted to spend a couple of minutes explaining what our vision about this wave of AI and how Kalray actually play a crucial role in this evolution of our industry. First, let me explain my vision, our vision about this surge in AI, I would like to say this burst on AI. Okay. For us, it's not a burst. I mean, we've been working in the AI domain for more than 10 years. Already when we listed the company back in 2018, we mentioned how AI was important and adding processor to process. Incidentally, that was important. Let me explain why it's so important and why Kalray is at this heart of this evolution.
The first thing to re-understand that AI is a way to analyze data. It's a new way to run algorithm. Why is this important? The reason why AI is important is because we are generating more and more data, AI is one of the rare way to actually be efficient in analyzing massive amounts of data. Okay. The reason for that is that the usual way to analyze data was to have functional algorithm, functional algorithm were designed by engineer, and assumed that the engineer understood the overall data there to take into consideration. With the rise of this huge amount of data, it's getting more and more complex to understand the rationale of the data you get, you need to find another way to process the data.
AI is very, very actually efficient way to process data, especially data where there is a lot of unstructured data, where it's very difficult to get basically, a very clear understanding of what is in the data. First thing, yes, we are generating a lot of data, and AI is a new algorithm, very efficient, is an algorithm to treat a huge amount of data. Okay. Actually, it's not so new, actually. I mean, for example, when I was an engineer student, I studied AI and AI algorithm, and at that time it was already very similar to what we do right now. What was the difference? Now we're talking about AI, whereas we were not talking so much about AI 30 years ago. The reason actually is about semiconductor.
The semiconductor industry made so much progress in the last 30 years that now we are capable of running such a amount of parameter, such a large network, that it starts to be very, very efficient. That's what you realize if you play with ChatGPT. The reason why ChatGPT is very efficient is because it's just such a large, actually network of parameter that have been trained on such a lot of data, that now the answer of ChatGPT starts being relevant. Okay, once again, ChatGPT... What is ChatGPT? It's an algorithm, and the algorithm based on AI, okay, predictive AI. Why it's powerful? Because it crunch a huge amount of data, especially the data available on the wide.
ChatGPT is capable of basically predicting outside the set of sentences you put as an input to the algorithm, predicting what is the next sentence supposed to say. Okay. It's an algorithm crunching a huge amount of data, and which is enabled by a new type of processor. I hope you understand why it's so interesting for Kalray and why it's we are really at the heart of this market, and why our products will be at the heart of the market, because you need to have efficient processor to manage that. Now, let me go into more detail about what it means for Kalray and for our business.
Basically, I mean, on our side, we decide to focus on as you remember, on a certain number of verticals, we explained that just before, these vertical are storage and edge computing. The main reason why we focused on the vertical, once again, is because we are very relevant on this vertical, and because we didn't want it also to face a fierce competition with the other verticals. Typically, just to make it clear, we are not providing our processor for ChatGPT itself, okay? NVIDIA is definitely leading the pace. They invested massively here. It would be completely silly to fight headward on this domain. We actually have invested in very interesting domain, which are also growing very fast, and storage is one of them, which is very interesting.
In any ChatGPT implementation, you need to have very fast storage. One, because you need to feed the GPU, you need to feed the AI processor very, very fast with data. What you need to do so is intensive storage, and that's what we do, and that's what we need, okay? For us, the adoption of this new AI model is extremely interesting because that's a huge driver of growth for selling our solution for storage. When I start on edge computing, for edge computing, we decided to focus on edge computing because we have very unique position compared to our competitors. If you look at machine vision, I think you understood why we are very strong in machine vision with the previous slide I explained to you.
The reason is that the industry is moving to traditional algorithm, to AI-based algorithm, and we are extremely strong in building this algorithm. The key metrics here is about performance per dollar, performance per watt. That's exactly what I described with Coolidge two just before. The second vertical we focus on for edge computing is 5G and telecom. Why? Because this industry is also adopting more and more AI algorithm and need to merge in the same solution, standard algorithm and AI-based algorithm, and we have a unique solution to provide that to our customer. You see, these three segments are growing very massively thanks to AI, but are very, very well defined for Kalray because we've got a very unique position, key selling points versus competition.
I hope this helps you understand what happened on AI, sharing our view about AI and why it's so important, and why what you see is not a burst, it's really massive, and we are at the heart of the revolution. Before moving to any question you may have, let's remind you what our objectives for in 2023 rapidly, and then, of course, we'll open the floor to any question you may have. The first, in terms of 2023 key objectives, of course, it's about business growth. We mentioned that our target was 40 million EUR for this year. I confirm. That our target for this year is 40 million EUR. As you've seen in H1, we generate already 15.3 million EUR, which has been already a very significant growth compared to previous year.
We see that we will accelerate, and we're comfortable we will reach this EUR 40 million in 2023. In addition to business growth, we are continuing working on strategic contract. We, of course, continue working on the German contract we announced last November. We are completely on track with what we announced. We are already in discussion to duplicate type of contract and have also opportunities of high growth with very large contract. Second objective is to continue building the team, and we made some very important recruitment in H1. We continue in H2, both in terms of exec team and also the team for our sales, for operations, and for also developing our next generation of DPU called Dolomite. Third objective for this year is to continue improving EBITDA.
As you will see, we'll share all our detailed result in September. Our EBITDA definitely is under improvement, we will see number in H1, and we will continue for the whole year. As a conclusion, I think as a conclusion, we initiated a change of dimension last year. As you can see, we continue, and we're moving fast. We're moving rapidly. The goal is to be a European leader in a domain, in a strategic market, which is this market about data-intensive processing for data-intensive application and next generation of data center. The market is now booming. Our product is mature. Our product and technology are matures, as you can see with our revenue.
We've got strong partnership in place. We are capable also, in addition to the strong partnership and this growth in the revenue, in putting in place major contract with large player. We believe that, we are on track to become the European leader in our domain and one of the global leader. Now very open to answering question during myself. Thank you very much.
Ladies and gentlemen, if you would ask to ask a question, please click the Raise Your Hand button on your computer or press star six on your phone keypad. Mr. Le Bihan, you can now speak.
Yes, thank you, Eric. Thank you, Denis, for the presentation. I had just a question about your confidence in achieving your 2023 guidance. I was wondering if it evolved during with the new contracts you've seen during the first semester, and if you are more confident in achieving the EUR 40 million revenue?
Okay. Thank you. Of course, a very important question. As I mentioned, you've seen H1 is already a very large growth. What we see based on offer now, based on the commercial traction, that H2 should be much higher. Definitely, yes, we are very confident we will make it. Despite the fact that the industry may a little bit slow down, what we see is anyway a very strong demand. Also, as I mentioned to you, we've got a couple of partnership, including Dell partnership, which should really give most of the benefits in H2. Combined with contract we signed directly, especially in the media space, we are very confident we will reach EUR 40 million.
Okay. Thank you very much.
Welcome. Any other question? No. Maybe for those who were attending to that yesterday, we had already a lot of questions, maybe we're done with questions. Anyway, I want to thank you very much for following us and for supporting us. I wish you an excellent summer, excellent vacation for those who are looking to vacation. The next synchronization point will be on the 21st of September.
September.
21st of September. Sorry about that. Looking forward to our next call. Thank you. Have a good day. Bye-bye.