INSANE: Intel demos Project Endgame, lets you borrow compute from nearby devices

    Overnight, Intel held their Vision 2022 Day 1 Keynote and unveiled something truly remarkable – Project Endgame.

    As we gear up for future experiences in the Metaverse, Intel are trying to solve the issue of devices not having the necessary hardware to power a great experience.

    In the demo, we see a fairly decent gaming laptop running a real-time simulation of a real city, using the brand new and gorgeous Unreal Engine 5. Problem is, the experience is just a couple of frames per second, so how do we get to buttery smooth, photo-realistic graphics in a laptop? You could wait 10 years for the hardware to get good enough, or you could use Intel’s Project Endgame.

    This phenomenal experience is referred to as ‘Continual compute infrastructure’ and works by looking for compute power in the vicinity of your device, within the latency thresholds set by in Endgame.

    In the demo, has a nearby computer (imagine this is in your home office, while you have your laptop on the couch in the lounge room). This computer is an Alienware Aurora Gaming desktop and while they don’t list the specs, it likely houses the latest Intel 12th-gen Core-i9 processor, lots of RAM and an Nvidia RTX 3080 (or above).

    After enabling Project Endgame on the laptop, the laptop radically improves in performance, with the gameplay now smooth, despite the endless city in the environment. The laptop is borrowing compute, from a nearby PC.

    That’s incredible and has some dramatic consequences.

    Essentially what’s happening here is that the two devices are able to communicate over a high-speed, low latency network (I imagine WiFi 6e etc). The rendering tasks from the laptop are essentially streamed or outsourced to the nearby PC, and when those frames are rendered, they are returned (streamed back) to the laptop all without the user seeing any difference.

    If we take this idea and imagine it in a corporate situation, there are a lot of computers in an office, working on Office documents, which may only take the processor and graphics at 20%. What if the guy in the corner working on AutoCAD could leverage the spare compute in the other devices on the network to build simulations or render videos much faster than otherwise possible. This would enable businesses to delay hardware upgrades, and increase the value for money of the devices they own.

    If we take this to the extreme, there is a chance that payment could be built into the service. This would enable you to sell spare compute resources depending on the application. Not everything would have real-time RPG level requirements, so potentially the latency on something like 5G could be low enough to leverage resources much further away than the building you’re in.

    We’ve certainly seen cloud gaming taking off, with services from Nvidia, Microsoft and others, but this is a totally new way to think about maximising the hardware we own and the bits we don’t.

    This could potentially mean you buy a fairly run of the mill laptop, but on Friday nights, when you play games online with your mates, you pay $10 and borrow some serious compute from the guy down the road.

    If Intel can take this concept and make it a full-blown supported service, I think they’re on to something big here. If the Metaverse, a fully immersive virtual reality experience Zuckerberg keeps talking about actually arrives, its technology like this that may help make it a reality and to enable it everywhere.

    Jason Cartwright
    Jason Cartwright
    Creator of techAU, Jason has spent the dozen+ years covering technology in Australia and around the world. Bringing a background in multimedia and passion for technology to the job, Cartwright delivers detailed product reviews, event coverage and industry news on a daily basis. Disclaimer: Tesla Shareholder from 20/01/2021

    Leave a Reply


    Latest posts


    Related articles