Jul 01 2021
Digital Workspace

How GPU Virtualization Helps Universities Ensure Quality Online Education

As postsecondary schools embrace hybrid learning, specialized software from NVIDIA can help optimize learning environments.

Universities and colleges face a looming challenge as the 2021–­2022 academic year approaches: delivering optimal learning and teaching experiences for students and staff during hybrid learning.

This is a far cry from the first few months the pandemic, when universities pivoted as quickly as possible to emergency remote learning. In the upcoming academic year, higher education institutions will be focusing on better hybrid learning solutions. According to a recent survey by the Institute of International Education, 62 percent of American postsecondary schools will offer some combination of remote and in-person learning.

The challenge lies in ensuring institutional IT infrastructures are up for the task. Since students and staff will likely still be dispersed in multiple locations, they will need reliable access and robust network performance that is available anywhere, anytime.

Higher education institutions need solutions that can keep up with current requirements, with the ability to scale up to meet evolving demand.

This is where tools like workspace virtualization through NVIDIA’s virtual GPU (vGPU) software can help. But what does NVIDIA vGPU technology consist of, exactly? And how can virtual desktop infrastructure (VDI) help create quality online learning experiences?

DISCOVER: Learn more about workspace virtualization made for higher education.

What Is NVIDIA vGPU Technology?

When it comes to meeting Gen Z students’ high expectations for user experience and performance, NVIDIA vGPU virtualization technology is key. NVIDIA vGPU software delivers virtual workspaces that are equivalent to physical workstations. It makes server-class performance possible for even the most compute-intensive courses. NVIDIA vGPU software enables IT teams to help remote learners stay productive, whether they are working on sophisticated artificial intelligence applications or collaborating on team projects in real time.

In practice, this means shifting the burden of compute processing away from local physical devices by using VDI, which provides desktop applications and services via virtual machines (VMs). These are delivered to end users over a network connection.

“On a physical device, the GPU is typically installed in the device and handles all the rendering,” says Phoebe Lee, product marketing manager at NVIDIA. “NVIDIA vGPU technology sitting on top of the VDI enables GPUs in the data center to deliver GPU processing. All rendering is done on the server.”           

Universities and colleges can opt for on-premises or cloud-based deployments depending on their needs. For example, some schools prefer on-premises solutions that provide maximum provisioning flexibility. “For this implementation, a lot of schools are using the data center to provision and create VMs to run student and staff workloads in the morning. Then, at night, they use the same GPU resources that are freed up from the morning classes to run high-performance computing tasks,” Lee says.

For higher education institutions that prefer cloud computing, NVIDIA works with major cloud service providers along with smaller and regional ones to provide maximum flexibility.

Lee puts the NVIDIA approach to GPU provision simply: “This is no longer a one-to-one relationship but one-to-many that provides different solutions based on use case and user needs.” 

How Does NVIDIA vGPU Help Universities Deliver Quality Education?

To handle compute-intensive workloads on campus, universities and colleges have historically equipped students and staff workspaces with desktops that each contain a dedicated GPU.

However, online learning and teaching introduced a new challenge: If students couldn’t afford the right technology, they were effectively priced out of an engaging, quality education.

As noted by Lee, NVIDIA’s vGPU solution helps create a “cordless campus” that supports virtualized classrooms and lab sessions along with distance learning programs. “At a high level,” she says, “universities can expand their education anywhere to anyone, and at any time.

For Clemson University, this meant deploying vGPU infrastructure to better support its homegrown high-performance computing cluster, Palmetto. This resulted in reduced IT costs, with fewer onsite servers required. It also improved resource utilization and maximized performance. The University of Colorado School of Dental Medicine, meanwhile, leaned on NVIDIA to carry out a VDI hardware refresh. This helped improve remote student access to critical dental training tools.

How Can Universities Effectively Implement NVIDIA vGPU on Campus?

If a university or college already has an established VDI structure, implementation is easy — vGPU simply leverages existing frameworks to deliver on-demand compute resources. For higher education institutions that don’t have their own data center or VDI infrastructure, NVIDIA works with both large and small cloud providers to offer VM solutions.

With vGPU, customization is standard. “There are multiple ways to provision resources,” says Lee. “Schools can slice out a single GPU into multiple, smaller GPUs. For some schools with compute-heavy workloads, we see them bonding multiple GPUs together to provide more power. It really depends on the workload.”

In addition, vGPU solutions do not require physical device upgrades to meet evolving software or hardware requirements.

This is the fundamental value proposition of NVIDIA’s vGPU: Depending on the workloads of students, faculty and staff, resources can be allocated on demand to enhance both in-person and online classroom experiences.

Brought to you by:

FatCamera/Getty Images

Become an Insider

Unlock white papers, personalized recommendations and other premium content for an in-depth look at evolving IT