Project Eagle - CS50 Fair 2025 at Harvard

Hi everyone, I'm here with Presley, a member of the class of 29 here at Harvard College who's in Stoughton Hall, and he's here today to talk about his project Eagle. So can you just tell us a little bit about what your project is? Yeah, so, um, Project Eagle is a, it's a data center power efficiency calculator, and what it does is basically helps you to figure out how much, uh, data that you are using efficiently to use for like in particular context of the data center. So basically you, um. A data center uses multiple different uh sections, and power is being utilized in the IT infrastructure mainly and what you want to do is that you want to focus on that IT infrastructure and forget all the other cooling and um power like other power usage aspects of it. Yeah, that's awesome. Can we see your project in action? Yeah, of course, great. So, um, there's an about page we can start with an about page which is the project eagle Base. You don't need to log in or register for this, um, and basically it just gives you like a quick run. Rdown of my project. So I'll be working in the winter with, um, uh, a group. It's called the Okaria Data Center, and this is a completely 100% geothermal data center and, uh, this is compared to like the usual data at Diesel, um, it's just a quick rundown about me and the fact that like uh how much power it uses and everything. So this is me and about myself. So but let's go up to like the mean bone of things, right? Let's go to the efficiency calculator, um. Basically what it does, right, you input the amount of total amount of power used in the data center. So usually we use around 0. 2 megawatts, let's say per day, right? And the IT load, which is the amount of power used by the IT equipment only, which should be around, uh, efficiently, you should use around 0. 1, but normally it's usually around 0. 7, 07, and cost efficient efficiency, efficiency like 0. 05 business days. Otherwise, let's just do that. And in this case we want to use geothermal just like as a case in point. And what it does is once we calculate, it should give you results based off of that and it gives us a power usage effectiveness, which is an index of the power use in the IT infrastructure versus the entire data center. And we also have the estimated CO2 savings, a daily operational cost. We have a green data center. Simulator as well, which kind of also shows us how much power is being used by the overhead cooling alone so we input our data source as well, that power source as well, and the computational load we use the same 0. 02, uh, 0. 2, sorry, um, cooling method. Let's use liquid cooling and we should be able to get a simulation results, right? And we get a total cost per day, uh, around $400 per day. 8% of the total amount of power being used by the data center is being used for the overhead cooling, and we have our carbon footprint to show a profitability score of the data center. So we're able to actively see if you want to pitch to organizations, we can actively see the profitability of the data center. And lastly, if I log in here, so I already have my login credentials available already, so I'll just log in really quickly. And I can access field audit and audit logs tools. And basically what I'll be able to do is as a researcher, I go into the data center and I plug in that my information. So where the data center is, the temperature, average temperature of the data center, the power drawn, the efficiency that we have calculated from the green data center stimulator previously, and I already have predetermined information. uh, so if I go to audit logs, like I've already, uh, put information there, right? I have a history of. All the information I've put in with all the IDs I've deleted a few, um, I have Harvard, Earia, Northwest, just, uh, just to show us like a average roughly, um, and you're able to see the temperature and time, uh, temperature change over time and the power drawn over time as well, right? Because you're, if you have one particular data sensor, uh, we're using the time stamps, you're able to show, actively show a customer how much you are saving, drawing, and. What is the temperature at that average time and in order to show them the efficiency compared to either a diesel-powered, um, data center, solar-powered or hydroelectric, and this is, yeah, it just goes to show how like the real demo, you're able to delete records, add records, log out, and that's the general functionality of it, yeah. Yeah, that's so cool, Presley. Um, and can you just tell me a little bit more about what Project Eagle does and what your affiliation with it is? Oh, great. So, um, I. I mean it's, it's more on this project, but basically I will be working with Project Eagle in the winter, right? It's a, it's a green data center, 100% geothermal, uh, energy, and what I will be doing, right, is like I will be pitching ideas to customers, potential clients. So far we've already amassed around a billion dollars in investing from Microsoft Corporation, right? Um, it's a pipeline project with also a company called G42 from, uh, Saudi Arabia. And we, uh, this project is done in like an environmental consultancy style, right? Because I'm trying to push for like more funding to show the stark difference between

um, using like diesel-powered or like carbon-powered, uh, data centers versus renewable-powered data centers and how the different renewable uh match with each other because I've shown geothermal, solar, hydro. Electric and diesel powered data centers, and I want to show the stark difference between diesel and geothermal, right? And since we are working on a 100% geothermal data center, then we're kind of trying to show like why the funding that we're asking for is important, right? And like why what we're trying to do is important, how it'll help in the long run. So Super cool, incredible project. Thank you for sharing with us. This was CS 50.