Cadastre-se agora para um orçamento mais personalizado!

Nvidia-powered AI supercomputer gives students the same tools as big tech engineers

Apr, 10, 2024 Hi-network.com
A lecture hall full of students
Getty Images/Cultura RM Exclusive/Peter Muller

Do you want to play with a supercomputer? Some engineering students are about to get the chance. 

On Tuesday, the Georgia Institute of Technology announced their new AI Makerspace, an "artificial intelligence supercomputer hub dedicated exclusively to teaching students". 

Also: Tech giants hatch a plan for AI job losses: Reskill 95 million in 10 years

The Makerspace was created in collaboration with Nvidia and is powered by the tech giant's AI Enterprise software. The press release describes the hub as providing a "high performance computing environment not unlike those used by researchers in higher education and the workplace." 

Similarly to most professional industries, higher education is responding to the AI boom. Schools including the University of Pennsylvania and the University of Texas at Austin have rushed to create AI-specific degrees, for example. 

In January, the University of Texas At San Antonio announced the creation of a new college for AI and related disciplines. In the same month, OpenAI partnered with Arizona State University, so students can experiment with ChatGPT Enterprise as a research and learning tool. But Nvidia and Georgia Tech appear to be going a step further, giving students more direct access to open-ended computational power.

Also: How to avoid the headaches of AI skills development

Georgia Tech's partnership with Nvidia is focused on the undergrad engineering program. Students in a machine-learning fundamentals course are the first to access the Makerspace, which they can use to prototype their ideas at scale. 

For college students to have access to this kind of computational power is significant. Aside from, maybe, Satori, the computing cluster derived from IBM's Summit (once the world's fastest supercomputer, now ranked#7 by Top 500) and donated by the company to MIT in 2019, undergrads rarely get to interface with such powerful tech. 

That Georgia Tech students can create, iterate, and experiment at scale with cutting-edge computing this early in their careers could mean major AI innovations and a next generation of builders better prepared to shape the field. The partnership marks a direct investment by Nvidia in the future of AI professionals. 

This first iteration of Makerspace is powered by 20 Nvidia HGX H100 systems, running 160 of the company's H100 Tensor Core graphics processing units (GPUs). The systems support advanced machine learning and AI and are linked to Nvidia's Quantum-2 InfiniBand platform for in-network computing. Accessible to students online, the software itself lives on infrastructure designed by Penguin Solutions. 

Also: AI won't replace your job yet because it's still too expensive, according to MIT

Unlike the Nvidia processors in your gaming laptop, these GPUs are enterprise-level tech that individuals can't normally access. The release puts this jump more plainly: "it would take a single NVIDIA H100 GPU one second to come up with a multiplication operation that would take Georgia Tech's 50,000 students 22 years to achieve." 

In addition to the technical nuts and bolts, Georgia Tech added that "students and faculty will also receive support through Nvidia Deep Learning Institute resources, including faculty-run Nvidia workshops, certifications, a university ambassador program, curriculum-aided teaching kits, and a developer community network."

The goal of providing the Makerspace is to "democratize access to computing resources typically prioritized for research to deepen Georgia Tech students' AI skills and shape the future generation of AI systems professionals," the university said in the release. 

Though only initially available to engineering students, the university intends for all six Georgia Tech colleges to use the hub. Access will expand this fall to the entire College of Engineering. By next year, all undergraduate and graduate engineering students will be able to use the hub. 

Also: AI will have a big impact on jobs this year. Here's why that could be good news

Nvidia processors

Nvidia systems at Georgia Tech.

Georgia Tech/

The initiative doesn't stop there: in 2026, Georgia Tech plans to expand the cluster into "the AI Makerspace Omniverse, a sandbox for augmented reality (AR) and virtual reality (VR)." Based on Nvidia Omniverse, the sandbox will be available to all students. 

Non-engineering sectors at the university using the technology could mean exciting interdisciplinary possibilities for the future of Makerspace. This integrated approach is increasingly important for emerging technology, as evidenced by organizations like Partnership on AI. 

"To meet the needs of tomorrow's innovation, we need a diverse workforce proficient in utilizing AI across all levels," said Arijit Raychowdhury, a professor in the College of Engineering, in the release.

Of course, the other benefit for companies like Nvidia in putting their tech in universities is getting the next generation of engineers familiar with their platform -- likely in the hope that they'll develop a preference that they'll carry into the workforce. Nevertheless, it's still a win for students. The Makerspace gives them the opportunity to work with the same level of sophistication as engineers and researchers at big tech companies during a crucial time in AI development. 

tag-icon Tags quentes : Inovação

Copyright © 2014-2024 Hi-Network.com | HAILIAN TECHNOLOGY CO., LIMITED | All Rights Reserved.