Early in 2018, NVIDIA unveiled the Windows beta version of their GeForce Now game-streaming service, which lets users pay a monthly subscription fee for on-demand access to a library of video games in the cloud. More importantly, you don’t need an expensive, high-end PC to run the gaming client. NVIDIA’s minimum recommendations are to use a PC with a 3.1GHz Core i3 processor and 4GB RAM, along with either Intel HD 2000, GeForce 600 series or Radeon HD 3000 graphics. These are all specs you’d find in a computer from 2013.
What is remarkable about NVIDIA’s streaming service for gaming is the reliance it will place on data centers in an industry that requires no latency. Different from Netflix or Hulu, who are pushing content to your TV or device. A video game has to respond to your keystrokes, meaning the commands have to travel back-and-forth across the network to be processed by NVIDIAs data centers fast enough so that you feel like the game responds to each keyboard and mouse in real time.
The feeling of real-time immersion is what makes gaming so engaging, and the risk of streaming games from the cloud is latency. Milliseconds of latency can create a slight delay in the response of the game to your mouse click. If the gaming data center is too far away or the network connections are not robust enough, the user experience will suffer dramatically. As a result, NVIDIA (and other game-streaming services that are sure to follow) are placing a big bet on edge computing. Latency can kill a user experience, and the only way to ensure latency is minimal is to place the computing and processing power of the gaming data centers as close as possible to the user.
While this is a direct challenge for the gaming industry, it also has more broad impacts on verticals like higher education. A Pew Research Center survey found that 65% of college students reported playing video and online games regularly. If gaming were made more accessible by enabling streaming of video games on low-cost PCs, it is possible that this number could increase. Streaming services and online video have long created network challenges on college campuses – our customers have struggled for years with the question of removing traditional cable TV connections from residence halls and investing instead in higher density wireless solutions. A streaming gaming service would add to the challenge to build and maintain a data center and network infrastructure to support a campus. College students treat the network as a utility that should always be available and high-performing for whatever educational or non-educational application use.
Many IT organizations in higher education are adopting aggressive cloud migration strategies. For several applications, this makes sense. However, the risk in developing aggressive strategies that collapse data centers too quickly is in not planning for the impact the edge computing requirements that streaming video-game services and the Internet of Things. As new services like NVIDIAs streaming video games launch and new network-connected devices come online in intelligent buildings, it will be critical to have computing and processing as close to the user or device as possible. This may not require a large, single consolidated data center, but it will require multiple smaller edge data centers across a campus.
The rise of the Internet of Things has already increased the need for long-range network and data center planning. The launch of a service like NVIDIA’s cloud-based gaming reminds data center operators – especially those in industries like higher education – that the need to deliver data center and network capacity on-premise remains critically important.