Neal Stephenson first coined the now popular term ‘metaverse’ in his 1992 science fiction novel Snow Crash. The dystopian storyline painted a bleak future ripe with big brother surveillance, racism, sexism, and hypercapitalism. Fast forward to today and everyone is talking about the metaverse but its hype revolves around brighter, more opportunistic ideas. These include virtual worlds where social networks are merged into one, environments that are powered by artificial intelligence, decentralized governance systems, and high-performance computing (HPC) through Web3 interfaces. Metaverses are just getting started and high-performance computing is not only useful for faster processing, but it is also becoming a requirement.
Nobody could have known what computing power requirements were needed to power today’s industries. Weather simulations, protein folding, fraud detection in finance, personalized health care, and smart energy grids are the tip of the iceberg when it comes to examples that require high-performance computing. The metaverse is currently being developed in all shapes and sizes but the question remains: It has been 30 years since Stephenson released his vision of the metaverse but what will the metaverse look like in the next 30 years?
Will we continue to have multiple decentralized metaverses created in silos and will they somehow interoperate between each other? Or will there will be one dominant centralized metaverse that the majority of the world will plug into? The former looks like the current landscape where sandbox-type metaverses dominate the blockchain gaming industry such as The Sandbox, Axie Infinity, Decentraland, and so on. The latter is what Meta seems to be working on that resembles a virtual realization of Facebook of sorts. Though we won’t be able to predict what the future of the metaverse will look like, we can bet that we’ll need enormous computing power to drive it.
In the future, will we be depending on our graphics cards and processors to provide a seamless and enjoyable metaverse experience? Decentralized players often run into performance issues such as dropped frames, stuttering, or sluggish movements due to issues with their dedicated graphics card. A quick update to the driver should address the issue but what about upcoming triple-A games?
Should you need to upgrade your graphics card every time there’s a new metaverse game? This sounds a lot like the PC game market where you constantly have to pay to upgrade your systems. We’re already seeing cloud-based gaming taking off in the console world where digital downloads are becoming the norm. Though Stadia, Google’s cloud gaming platform, didn’t exactly take off, the attempt at revolutionizing gaming through cloud-based HPC was a valiant one nonetheless. However, it should still give us insight into the future of metaverses and where the leaders in silicon valley are taking us.
As an executive of Intel’s Accelerated Computing Systems and Graphics Group, Raja Koduri made some predictions about Web3 and the metaverse saying, “The metaverse may be the next major platform in computing after the world wide web and mobile.” He goes on to describe how we will have a long ways to go and that “our computing, storage and networking infrastructure today is simply not enough to enable this vision.” To create a truly immersive metaverse experience at scale which is accessible by billions of people in real-time, he envisions a 1000x increase in computational efficiency from today’s technological state.
It’s all about creating a simulation of the world in cyberspace. When Super Mario Brothers was released on Nintendo, the world got hooked instantly. Today, we smile at the nostalgia that the 8-bit graphics brings us because of the generational leap of graphical prowess and computing power required by current generation titles like Half Life: Alyx or Star Atlas. Just as console games evolved from 8-bit to virtual reality, metaverses will evolve and we are currently seeing its evolution in computing requirements.
We need high-performance computing to power the metaverse of the future because it will require real-time data processing and lots of it. This includes integrated artificial intelligence in environment design, rendering 3D environments and character movements, creating predictive weather models, and so on. This is why Meta teamed up Nvidia to build an AI supercomputer just to power the metaverse. Metaverses are a hot topic right now and there are plenty to plug into. Which will survive over the next few decades? No one knows but one thing is for sure, we will continue to depend on high-performance computing to power these metaverses.