NVIDIA Deep Learning Institute and GPU Audio speak on adapting to the future, the GTC Conference, and opportunities for audio
GPU Audio is collaborating with NVIDIA’s Deep Learning Institute to host workshops and training labs on utilizing NVIDIA GPUs for low latency, accelerated audio computing.
NVIDIA recently held its 2022 GTC Conference. We spoke with Andy Cook, Senior Manager at NVIDIA’s Deep Learning Institute (DLI), about how the company is driving the development of the metaverse, how the future will change, and the partners that are helping them along this journey.
For DLI, it can be a challenge to keep up with the rapid acceleration of change in the world of Artificial Intelligence (AI) and GPUs. “NVIDIA is working with researchers, developers, and partners to design and implement cutting technology,” Andy says. “DLI started with a focus around deep learning and accelerated computing, but as our team, focus, and company have grown, we’ve expanded into other areas like data science, robotics, and the metaverse.”
The GTC Conference presents opportunities to address demand from researchers, data scientists, IT professionals, developers, students and early adopters for learning in these areas. “Doing things like hosting a session with GPU Audio at GTC is how we offer more comprehensive learning experiences,” Andy adds.
Cook himself has spent a big part of his career in audio: coming out of college, he worked at Digidesign, developing Pro Tools, which Avid later acquired. “Seeing this transition from analog workflows to digital [ones] and the impact, new techniques, and opportunities that came out of that transformation was really exciting,” he says. “Along the way, you also recognize that things like the Convolution Reverb are so demanding — the DSP horsepower required for really putting that into practice was very limited.”
Moving to NVIDIA, Andy noticed that on one side, there was a lot of demand for audio processing power, and incredible processing power on the other — so, he was waiting for the bridge that could connect the two.
“When GPU Audio came along, it was a moment to see groundbreaking innovation opportunities connecting. I think it’s really exciting for the future,” Andy concludes.
With GPU Audio’s workshop session at the GTC conference, Cook hoped that it would help convince people. Many recognize the myriad of opportunities that GPUs theoretically present for audio but have shied away because of the challenges associated with harnessing that power. “GPU Audio has done some amazing things in this area,” Andy continues. “One of the things I’m hoping [is that] by having this hands-on training through DLI, people are actually rolling up their sleeves and getting into the nitty-gritty of it.”
One of the most significant possibilities opening up is the one that AI is offering. For Cook, artificial intelligence allows for extended experimentation: more specifically, trying ideas in a way that you can actually afford based on the timeline and the creative work environment. If before you could only try one alternative because of the time, AI now removes this time-sensitive limitation from audio production. “AI for music generation — initially serving limited scope that has a financially driven reward for it — is going to be eclipsed by all of the creative components that will follow,” he predicts. “It’s just a matter of time, but also the tools and techniques evolving to catch up to the technology and connect that process with these new capabilities.”
As Andy explains it, the processing power relative to the community and use cases of audio, seems to be in a bit of a “leapfrog” relationship. For instance, you start getting more power out of your CPU as new models are released. Audio then grabs that and suddenly does more with it. “I can now go from doing 16 tracks to doing 24 tracks! But then enter higher sample rates, bit depth, surround sound, and suddenly your sheer system is brought down to your knees again,” Cook elaborates, saying that you learn to limit yourself along the way. “You put your blinders on because you can’t go there — your system doesn’t have that power [– so you have to try workarounds]. And so when new tech like [GPU Audio] hits, you can reprogram yourself: ‘now, I gotta take my blinders off because I can do all these things!’”
But beyond accelerating audio production workflows, perhaps the most exciting areas of long term development are in reimagining and reshaping what audio is and what it’s used for, in a more fundamental sense. For example, there are a lot of possibilities emerging in NVIDIA Omniverse that may have never occurred. As an example, Cook talks about how you can use audio to generate facial expressions, drive gestures, and new implementations beyond audio production itself.
“As a medium, audio is very sophisticated — it’s the core building block of society, so there are going to be even more new techniques and implementations as the power starts to take hold.”
Advancing further in broadcast has been a relatively new area of focus for NVIDIA. “A lot has been done in a very short amount of time [and] audio is a huge component for that,” Andy shares, saying that there is so much that GPU technology and AI techniques can bring to audio production. “I think we’re at the very beginnings of it. A lot of it will be in hands of our industry partners and software developers; we provide the foundation for amazing development and tools to be created and put into the hands of the customers — it’s going to be up to who’s paying attention, who’s harnessing that, and who’s pushing it forward and developing [those tools].”
The NVIDIA Deep Learning Institute offers resources for diverse learning needs — from learning materials to self-paced and live training to educator programs — so individuals, teams, organizations, educators, and students get what they need to advance their knowledge in AI, accelerated computing, accelerated data science, graphics and simulation, and more.
Check out a list of self-paced courses or browse through the list of instructor-led workshops available to enterprise companies.