The Hot Chips technology conference isn’t usually that fascinating for gamers. This year it might be a little different, thanks to a trio of deep-dives into next-gen graphics.
The official schedule from the technology symposium, which typically focuses on integrated circuits and microprocessors, has a real interesting trio of panels to bookend the first day. On August 17, Nvidia will be giving a talk titled “NVIDIA’s Next Generation GPU: Performance and Innovation for GPU Computing”, an obvious look into their Ampere next-gen architecture.
But they won’t be alone. Right after that, Intel will be talking about the architecture behind their upcoming Xe discrete GPUs. That’s being given by Intel’s David Blythe, their head of graphics architecture. And if that wasn’t enough, Nvidia and Intel are being directly followed by a look at the system architecture for the Xbox Series X.
It’s the only part of the conference that has some direct parallel with gaming. AMD and Intel will be talking about their next-gen mobile APUs and CPUs as well, but it’s likely to be a more technical, high level breakdown than anything relevant for consumers. As an example, AMD’s Hot Chips keynote last year focused on supercomputers, HPC systems and the server room.
So while most of the show is looking forward at the future of computing, it’ll be nice to get more of a technical breakdown into how the Xbox Series X, Nvidia’s Ampere GPUs and Intel’s XE work. Something to keep an eye on here for Xbox is machine learning. It’s a big topic of interest for the Hot Chips conference, and it’s something supported by the Xbox Series X through DirectML (a component of DirectX).
Here’s a note from the official Xbox Series X glossary:
DirectML leverages unprecedented hardware performance in a console, benefiting from over 24 TFLOPS of 16-bit float performance and over 97 TOPS (trillion operations per second) of 4-bit integer performance on Xbox Series X. Machine Learning can improve a wide range of areas, such as making NPCs much smarter, providing vastly more lifelike animation, and greatly improving visual quality.
Nvidia already has a strong track record using AI and machine learning for video games. Intel’s Xe, meanwhile, is designed to be a scalable GPU architecture that works just as well for laptops and server PCs as it does for enthusiast gaming rigs. How well that works in practice is another matter. But it’s always interesting to hear what the engineers think the future holds.
For more info on the Hot Chips conference, head to the official website.