 Good afternoon, fabulous hardware community, and welcome back to Denver, Colorado. We're here on day three of supercomputing 2023. My name is Savannah Peterson, joined by my fabulous co-host and analyst, David Nicholson. David, how are you feeling? End of day three, still feels like the energy level's up. Is your energy level up? Doesn't feel like the end of the day to me. Feels great, yeah, feels great. I'm very excited about our next guest, so yeah. I am very excited about our next guest. Please welcome Happy to the show. Happy, thank you so much for joining us. Thank you. You came all the way from South Africa, yeah? Yes, 17 hour flight plus five to get here with some layovers, so it's a long trip. Yeah, that is a commitment to participate in the community, we appreciate that. You would never know looking at you, they've traveled that far. Are you off by 10 hours, 11 hours? Yeah, yeah, off by 10 hours. Yeah, okay. South Africa is 10 hours ahead of 10 hours. Yeah, yeah, yeah, okay, so we'll be gentle. Yeah, yeah. Casual 2 AM broadcast for you, no big deal. We'll just turn up the lights. That's awesome. So you work for the National Integrated Cyber Infrastructure Center, correct? Can you tell us what you guys do? Yeah, we look at the National Integrated Cyber Infrastructure, it's a government funded entity, funded of course by the South African government. And this is to ensure that we can provide cyber infrastructure services and of course infrastructure to the research community in South Africa. That will be your high performance computing, make sure that we have also the broadband connectivity. So this will be equivalent to what you have as Internet 2 here in the US. But the other third component, which is quite important within the cyber infrastructure, is the data management. What do you do with the data? How do people move the data? So that's what we do and to some extent we provide those services to our industry also to make sure that we can get our industry to be competitive. But we extend this services also to other parts of Africa to build capacity in Africa. So yeah, that's what we do. It's awesome. You mentioned to me earlier that you were the only center in the continent, correct? You're the only center doing this in Africa? That is correct. And we started with this in 2007. So you could imagine before 2007, there were no supercomputers in the continent. So we started off with a system, 2.5 teraflops in that time in 2007. If you look in the top 500 calf in 2007, 2.5 teraflops was not really that significant, but very important for Africa. That's where we started. Today, we are at like a capacity of 1.6 teraflops and we grow it, yeah. That is quite the journey and in not a very long time. You also mentioned, and I thought this was a great point I want to bring up, you're our first government guest, also our first African guest on the show for supercomputing 23. Governments need a lot of evidence to make an investment in HPC. What was this journey like for you guys? Look, it's not an easy thing to convince. No. The government to put money into what they could see as like, I mean, toys for scientists. So the first thing is to look at what can you be able to solve in terms of tangible things in the country? So we started off from there with some seed funding and looking at that, in South Africa, we looked at what are some of the strategic initiatives in the country? What are the pain points? And can we be able to take those pain points and use high performance computing to accelerate them? So that's where the journey started. And looking at that, for an example, we look at challenges like climate change and look at it and say, can we be able to contribute in that? Some of the things is industry competitiveness. Can we be able to accelerate some of the industries? But more important to make sure that the research outputs can be competitive. So that's why we started and demonstrated that with now a little bit of investment, we can have some production and then from there then, the government has this long-term investment. So it's important to talk to your stakeholders' needs and demonstrate that we can contribute to that. Yeah, and get that most affordable. I hate using the term cheap in our industry because it's very expensive to do anything with that most affordable proof of concept over the line to continue to get the funding. Tell us a little bit about some of the projects. We talked, there's a lot of research being done. I know there's some very exciting astronomy projects that you're leading. Look, by the time when we started investing in high-performance computing, we were at a stage where South Africa wanted to convince the world that there will be the right host for such big radio astronomy project. So at that time, there was Australia and South Africa competing to host this. So apart from saying we are geographically situated and we have got that geographical competitive advantage to hold such a project, one of the things is for such project, you will be producing a lot of data. You need to be able to share it with the rest of the world. You need to show that you have got the right tools which are cyber infrastructure tools for that. So in 2007, we focused a lot in that to push that case. And today, that has been announced. It's now, we're already starting with the first phase of that astronomy project because we demonstrated that we can do it. We have got the computing facilities. We can grow them. We've got the capabilities and that is very important. And we will be getting the first phase of, I mean, the second phase in 2025 of the SKA project. Exciting. So what have you seen? Did you attend Supercomput last year by any chance? Last year I did not attend but I have been following up on the news on that. You remember that 2019 was the last physical event and then things quite off. And then, so last year I could not attend because I did not update my visa. But fortunately now, at some point my passport expired and I wasn't aware because there was no traveling. I've been in the same circumstance. Yeah, yeah, yeah. So I lost that. Different travel. But I was looking at the technological advancements. Things start moving and I think at that time that's when things like AI, HPC turning into AI. This is the question, yes. Gaining traction. Things like quantum computing now being seriously, not just being on the sidelines but getting into the mainstream of our discussion of HPC. So do you feel that that makes it potentially easier to get that attention and funding on projects like this because it is sort of in the forefront. Government officials who are in the process of making decisions about how that money is allocated, if they've experienced something like as simple as chat GPT, then at least they've got a sense of AI. And if you're talking about supporting AI with what you're doing, is that helpful? Or does it confuse the issue? Look, one of the things that you have to look at is that in the adoption of your generative AI we were just coming out of the COVID-19. One thing that was there was everybody started not being interested in the data and the data analytics. And so generative AI came in at the right time because people started looking at the importance of high performance computing in the data analytics. So it was no longer just for your big science projects or for engineers and scientists. Even my grandmother now could follow HPC, understand that it is important for a lot of data to be processed to be making decisions. So it is at that time that everybody is convinced that, look, we need to invest in this. So I think it has made its case and everybody is backing on this development. Of course, the challenge is always for us, traditional HPC to say, we've been left out now on the sidelines and is this everything is going to be just focusing on this generative AI? They actually need it. So much foam all right now. Yeah, exactly. You can actually get a rubber stamp of the letters A and I and you can add them to all of the things that you've been doing. It's like, oh no, no, it's AI now, it's AI now. Certainly everyone tried to claim that for sure, yeah. We were also talking earlier about this idea that we generally accept that technological improvements are going to deliver universal advantages to humanity. You mentioned that you are unique in the entire African continent. How do you look at the way that these technologies can be leveraged to help more than just the scientists that are working on the projects themselves? Is that kind of a directive guidance from the government that this meet? We talk about democratization of this. I think in my mind what it means is the ability to leverage these tools in a way where more people take advantage or more people gain advantages from them. How has that looked at? Or is this purely scientific research and that's for policy makers? No, look, it's important. First thing is that you should understand for governments to invest in this. Governments have got a short lifespan, right? In the U.S., how long do they stay? Four years. In South Africa, it's five years. So whatever that I have to convince them, I have to make sure that in five years they can see success. Because they need to invest in something that can also make them look good. Make that story look good, yeah. So it's important to understand what are the pain points. So HPC, Generative AI has to answer those things, has to try and improve humanity and remove the challenges that people have. So that's the only thing that will make us successful. And it has to be able to talk to that. But the other thing, you're looking at the whole thing about as the technology advances, are we living the community away? So when you talk about democratization, it's that we should not be living part of the society out of there. So in terms of skilling up people. AI, for an example now, in most part, I think in the U.S., they will be having those challenges to say, is it going to take up our jobs? Am I going to? So looking at that is how do we bring in people to make sure that we can reskill them from where they are to follow where now the technology is going so that they're not left behind? So we've got a lot of work to do to make sure that we can answer those questions. You have such, I mean, you play a really important role. I know that education is core to your being. You're actually the educator of the year in the internal audit community last year, which is very exciting about that. And you're all about accelerating discoveries in Africa. I'm assuming you see quite a variety of different projects. Tell us about some of the projects that excite you both from the student level all the way up to enterprise that you get to engage with. Look, the story here for me, and when I looked at HPC is that if you look at a lot of people who are in the HPC space, we got in there because we had a problem to solve. This was not taught anywhere in universities. So a lot of people, it was self-taught. There were no any sort of courses in any curriculum. So like for an example, I'm not a computer scientist. I am a physicist and I had a physics or material science problem to solve and I needed a computer. So I taught myself Linux and all this thing. So it is that point where now we're working on how do we make sure that this is integrated in the curriculum right from the beginning. And one of the big projects that is going on around the world is getting students to be able to understand high performance computing at undergraduate level. Not only to start looking at it at postgraduate when you're doing your research and things like your student-luster challenge. It's one of the projects. Even here at SC, you'll see the students. Going building this, I mean from undergraduate level. So we've got that program in South Africa. We started it in 2013 and of course with support of some of the OEMs and other centers here in the US. I can tell you we took students who couldn't even spell HPC. But those students managed to go in be able to build a small cluster and they competed with the rest of the world. And we had South Africa competing quite well. Winning gold at the international supercomputing for a number of years. So this has been a sustained intervention to make sure that we can upskill these kids from undergraduate. Of course, we're not just looking at computing. We're also now looking at data analytics and the live cyber security. So it's building those skills. What do you think it means for the younger community or folks just learning about the space to your point? You're taking in students who can't spell HPC. What do you think it means for them to see you on this stage to see South Africa playing in the big leagues, winning and on that global scale? Do you think that's a great catalyst for the next generation of technologists? We have to really make sure that they can see the future in this. Because the kids today, they look at what is cool. That's the name, the word that they use. Is this cool? Am I going to have a successful career in this? So we need to be able to demonstrate that and show them that it's important to invest in that. They have to follow it. So we have got a big task. And today there are so many careers that people don't have to work a lot in terms of science and engineering. So that's a lot of competition to be able to do that. So we have to show that these things are important. And you know, something like Generative AI, almost everybody wants to say, I want to be able to do programming. I need to understand HPC. So it's a good advertisement for skills development. Well, I'm just going to go out on the limit and say I think you're cool, happy. I think HPC is very cool. And I'm excited to see more cool projects coming out of the center. I know that you think HPC is cool as well. Thank you both for being here on this segment with me. And thank all of you for tuning in to learn more about the exciting projects and research going on on the continent of Africa. My name's Savannah Peterson, coming to you live here from Denver at Supercomputing 2023. You're watching theCUBE, the leading source for emerging tech news.