BlueXP is now NetApp Console
Monitor and run hybrid cloud data services
Hello everyone. >> How are you doing? Welcome.Yay.Hope you're having a great netup insight event. I'm really enjoying myself here and thank you very much for joining our presentations. Also, since we know that it's lunchtime, so thank you for sticking through. Uh I'm Vladimir Esther. I'm the CTO and one of the co-founders of Cluster Power. And if you've been here through the rest of the day, I think that you already know this. However, we won't be sharing any confidential information. And as a matter of fact, please share it around with everyone you know. Uh reach out, tell everybody about what we're doing. And I hope that you will enjoy it. So today you'll discover clusters power's vision to help technology companies accelerate their scaling and also their go to market with customized state-of-the-art AI infrastructure. We will discuss a bit where it all began with us. What it means to have an AI factory. How does it help our customers? what it's all about, the infrastructure and the underlying technologies. And last but not least, we will be walking through a use case of a real customer that will amaze you with their AI potential and AI technologies that are making humans reach the use of AI in their day to day-to-day life. But before we start, let me know if you know who is this guy. Anyone?Well, he is Tomos II, the son of Greek general Alexander the Great that established the Tomic Empire with the capital in Alexandria. Do you know what is most known for? Well, he was an innovator and also a solution architect of his time with the Alexandriancourt being at its height during his rule. He was also one of the first CEOs of multinational companies with his empire greatness and richness being based largely on trade and knowledge exchange between Europe, Asia and Africa. But most of all, he was a visionary and a champion of human progress as he erected and promoted the Library of Alexandria. So at this point in time you may ask yourself what is the library of Alexandria has to do with AI and also the future of mankind. Well history tends to repeat itself. So at a close look the library of Alexandria was somehow the blueprint of that period's AI factories. It was a hypers scale to store the whole knowledge within the reach of a of the empire.It had the largest connectivity exchange more or less similar to an internet service provider exchange of the times with people traveling by ships or caravans to bring scrolls. It was distributed and with built-in backup systems as originals were copied and only copies could be read by scholars.It was a one-sizefits-all still tailored to any industry or business sector with the information covered from philosophy to science, from trades to history, from manufacturing to the art of war. And most of all it was democratic. Everybody had access to it from workers to scholars. So fast forwarding two millennas during which the human knowledge had its ups and downs, we get to the year of 2019 when we started to develop cluster power and bal. When we established the company, we integrated vertically two core activities, technology infrastructure and energy productions, hence the name cluster and power. One year after the covid kicked in and remote working, cloud collaboration and ultimately the use of AI accelerated and all of a sudden we found ourselves at the forefront of two tidal shifts that are shaping our society's work and governing.It's the AI revolution on one side and the green transformation on energy on the other side. We built in Romania and operate the first hyperscale technology compass in the region bringing a novel design that uses onremise power production facilitiesin order to run its primarily energy source. And we established a lot of firsts for that part of the world. And not only we are talking about 200 megawatts of hypers scale capacities that can host up to 5,000 racks of high density equipments.We have the first hydrogen ready power production facility that runs the data center and one of the most efficient data center operations in the world with a power usage effectiveness of just 1.1. We have uptime certification from um from uptime institute as well as a very strong partnership with Nvidia as we are a DJX ready data center facility which allows us to host and operate large supercomputing platforms as well as a cloud service provider that provides its customers with artificial intelligence infrastructure as a service. With all the hype around the AI, hundreds of companies have joined the race to change the way that we live, work, and interact with each other. It's a market where speed is key. That's very clear. With cluster power, you get access to a state-of-the-art AI infrastructure built on Nvidia H100 GPU technology. From super scalable parallel AI training to massive LLM inference that requires hundreds of GPUs working in sync. We provide the right infrastructurefor your workloads to accelerate time to market as well as global expansion. We do it in partnership with leading tech innovators such as NetApp, Nvidia or Cisco. And our validated setup is based on HGX with H100 80 GB of memory with SXM5 version running in a back end with connectivity on Infiniban 400 gigs NDR technology on a non-blocking fabric. with the Cisco ACI network fabric and of course the storage is powered up by the NetUP data fabric and NetApps ONTAP AI orchestration.The scaleout units can be interconnected to form multi-node date deep learning and HPC dedicated platforms of up to tens of thousands of GPUs. So at the end of the day, what why NetUP? So the process of operating AI requires massive amounts of data to flow unhindered through a five-stage process from ingestion through up until archiving of data. As enterprises of all types embrace AI technologies, they face data bottlenecks from edge to core to cloud. Netup delivers AI solutions that remove bottlenecksand are able to accelerate AI workloads and make it smoother cloud integration. This is why you we use in our AI factories, NetApp's data fabric, a unified data management environment spanning across edge devices, data centers, and one of the more public clouds. So your AI data can be ingested, collected, stored, and protected no matter where it is. Only then you can optimally train AI drive machine learning and empower your deep learning algorithms necessary to bring your AI projects to life. But the future is now. The success of tomorrow needs action today. Whoever models and trains their LLM faster, whoever proves their business model scalability and raises their next investment, whoever accelerates growth and go to market to get more market share, those AI companies will be the victors. We are here to help them accelerate with dedicated state-of-the-art infrastructure, tailor projects, and flexible financing solutions.Together, we want to foster economic growth, societal development, as well as knowledge democratization through AI. I know that for some of you will say that these are only marketing big words. But to prove them wrong, I have the pleasure to invite on stage my friend and dear partner from Humans AI, Bart Vinman. Thank you. >> Thank you. Can you hear me well? >> Yes. Perfect. Happy to hear. My name is Bart uh from Amsterdam, Netherlands. Uh very proud to be here uh for showing you um the why behind uh having a good uh data structure and data infrastructure. Um I work for humans.AI for several years. Um I'm aAI enthusiast um andinnovator uh by the heart and I'm very happy thatwe have this collaboration with cluster and net app because itenables us tobring the things that we want to bring to society. Yesterday we heard as well we need to embrace technology. We need to embrace uh AI andspeak about it how it can help society instead of fighting it andpushing it away. And that's our aim. That'sour goal is to make AI happen and make AI uh work for humans. Um we are a global provider of AI solutions. Um we deliver we started as a synthetic media company uh creating thesecool synthetic avatars uh speaking in different languages uh voice cloning. Um but we noticed thatthelandscape is so fragmented. we need to bring it together and we need to enable people tomake the right decisions in using the right technology. Um we delivered the first AI counselor in the world uh supporting a local uh or national government. Uh we developed the first blockchain for eyes inthe world. Um world. Um and like um Mr. Kuran said yesterday um datadriven businesses uh they need to be AI ready. uh and this is what we enable people or businesses to do. It's important to understand what it takes tobe a leader in AI and what it takes to incorporate it into your enterprise. And it's more than just data. It's more than just having somecool use case and uses. It's a complete um uh ecosystem you need to develop andinclude. Um so it starts with the right technical framework. Um on top you can build the AI core but it involves code, trainers, data, hardware, users of course and developers. It includes so much. We want to make it tangible andaccessible forevery kind of business uh in the world. So we developed the AI platform works towards AGI. So we're working towards the future by applying AI what is available today into one platform. And this is all about creating trust and increase the pace of innovation. And creating this trust and p thisa pace of innovationis enabled by our collaboration with cluster power and netup. And what we've built is an open collaborative ecosystem. So within this ecosystem, we bring together uh developers, scientists, users. It's a complete ecosystem which is umnot just one solution for one company. No, we bring in all the different stakeholders intoone. By doing this, we enable companies to develop their own narrow AI applications. So think about real tangible use cases having AI support your business case using large language models of course using the right data but your own data not just the openly available data but we want to work with your data your company data and we've seen some great examples already during these days but it's important if we enable people to have conversational AI if we enable people to have humanlike interaction it's also important to Give an identity to your data, identity to your company. If you want to have a conversation, I we believe you need to be able to talk to humans, but that should be scalable. So, for example, give it a face, add different languages. Dear friends at NetApp Insight 2023, it is a pleasure to be here with you >> and be able to interact with the data through a human person. Although it's synthesized, yes, but it's real time and you can interact with the data. Of course, this should be uh device agnostic. Doesn't matter where you are uh whichtool you need to use. Uh you need to be able to access that data as a user. Like I said, it's your data, it's your hardware, and we offer uh our clients the solution together with cluster power um that fits their needs. So either if you're a small company with small data set or a large one and of course we're flexible uh it can be custom as well. Last element important element for us is the blockchain side. So we bring together the right data infrastructure with our own blockchain for AIS to enable traceability transparency of the data as well. With this you can develop incredible use cases. So either it's a retail, it's sports, it's healthcare. Uh we also see a lot uh oftraction within the public sector. So one of the examples I already mentioned before is the first AI counselor in the world serving a governmentwhich literally is a robot we've put in the u theparliament of the government he really got a job and he serves the parliament in Romania by having a face a voice integrated within the mirror data added uh from society and you can interact with the data so The prime minister here is able to talk with the people. So you don't have a vote once every four years. You can send your input to the government directly using the right channels. So you can talk um it'sin app as well and it's in person. So there are different solutions we've brought together because it needs to be device agnostic. We've seen incredible uh pickup by uh society because the interactions from day one were immense. Also, we got a lot of developers that wanted to work for us. They want to do good, create AI for good and tobuild a bridge between the public and um andthe government. So, um I need to wrap up. I'll look at the time. Um it doesn't matter what your use case is. You have data, you have the right data infrastructure, you have the speed which is critical and you need to have the right audience. Bring it together, build the bridge and by then either it's public sector, private sector, we can serve in applying the right AI solutions uh at the right pace. Thank you very much. I give the word back to you. >> Thank you very much Mark. >> [applause] >> Thank you. So, we're trying to wrap up the session. Uh, and I think that the keytakeaways here is that AI development is all about speed. Is not only about the technology, but it'sabout how you implement the use cases and how you get direct outputs that make a real impact in our society. AI factories are here are the factories of the present and also of the future that are needed in order to expand the use cases to expand the reach of technology into society. And of course we don't have to forget that people and businesses are all about and are at the core of AI. We cannot do it withoutengineers, without freethinkers, without business people, without users. It would be all in vain. So, thank you very much again for attending the session. Uh we will be here again tomorrow and but you can also find us around here. We are present on uh LinkedIn and social media either at clusterpower.comor humans.ai or humans.ai or humans.ai to where you can learn more about our solutions and you can also get in contact. So again thank you very much for attending our session. >> Let's hear for cluster power humans AI. Isn't that fantastic? I tell you guys, uh, very humbling and my marketing, uh, department here canattest. My slides look like caveman scrawlings, uh, in comparison to how beautiful that artwork was up there. That was absolutely f absolutely astounding. Uh, I just want to say, you know, before welet you guys go, um, what you guys have done isleaned in on what requirements for tomorrow are going to be. the way you have set up these factories uh as sustainability becomes more at the forefront you're removing that very scary element because of how power intensive you know AI requirements are you're already thinking two to three steps ahead to help customers come in it's like if you're going to climb Mount Everest don't go by yourself have a sherpa there that's done it several times with you so I want to applaud you guys and I found that so fascinating as we were reviewing uh the candidacy for the award so we'll find out tomorrow main stage. Uh if it's going to be cluster power, is it going to be Garage Tech? Is it going to be HCL? Gentlemen, thank you so much uh for what you do. Thank you very much for the partnership. We really appreciate it. Hear it again.
With the advent of LLM and generative AI, machine learning has moved from academia and research to home use and mainstream applications.