Here’s something interesting you might not know:
The graphics processors, or GPUs, that make possible the eye-poppingly realistic graphics of games like “Quantum Break” are also really well-suited to powering artificial intelligence and other high-intensity tasks.
The world of high-performance computing measures power in “FLOPS,” or “floating point operations per second.”
It turns out that as video game graphics have gotten better, the hardware used to produce them is increasingly well-suited to powering the AI future envisioned by companies like Google and Facebook.
“[After] 2007, all the big advances in FLOPS came from gaming video cards designed for high speed real time 3D rendering, and as an incredibly beneficial side effect, they also turn out to be crazily fast at machine learning tasks,” wrote Stack Overflow founder Jeff Atwood in a March 2016 blog entry.
In fact, when the Google DeepMind AI won its history-making Go series against Lee Sedol, it was sporting 1,202 CPUs, or traditional processors, and 176 Nvidia GPUs under the hood.
Nvidia and Google are actually partners on artificial intelligence, dating back to the Google Brain image recognition system, as detailed in an Nvidia blog entry. Long story short, Google Brain needed 2,000 CPUs, plus all of the server infrastructure to support them. That’s a tall order. But they found that 12 Nvidia GPUs could deliver “the deep-learning performance of 2,000 CPUs.”
Which is to say, depending on how the DeepMind team set all of those chips up in the real world, the 176 Nvidia GPUs in DeepMind could well have been as powerful for this specific task as 29,333 regular old computer processors — insanely efficient.
Video games are the future
What this means for you, the non-AI, non-high performance computing expert, is that whenever you buy a new video game console, or upgrade your PC with a new graphics card, you’re subsidizing Nvidia and the other graphics card companies.
With those advancements, those companies and their customers can apply the technology in all kinds of interesting ways. For Google, that’s a great way to boost its AI ambitions. Nvidia itself is pitching an on-board computer for self-driving cars that it claims is as powerful as 150 MacBook Pros.
I recently spoke to Todd Mostak, CEO of MapD, a startup backed by Google Ventures and Nvidia. MapD uses the ludicrous performance of these GPUs to analyze immense amounts of data, like political campaign contributions in a geographic area, and display it on a maps or charts in real-time, interactively.
Mostak told me that the technology would simply not be possible at the scales it’s dealing with if it weren’t for the fact that gamers, ever-hungrier for better graphics, were driving massive demand for cheaper, more powerful GPUs. That translates into better performance and better margins for MapD.
“They wouldn’t have gotten there if gamers hadn’t wanted to game on bigger screens at higher resolutions,” Mostak says. “We definitely have people playing Quake to thank for most of the technology.”
This entry passed through the Full-Text RSS service – if this is your content and you’re reading it on someone else’s site, please read the FAQ at fivefilters.org/content-only/faq.php#publishers.