Nvidia GPU Technology Conference (GTC) 2020 kicked off on October 5 and will continue through October 9. The virtual conference had CEO Jensen Huang present a keynote on some of the implications of Artificial Intelligence (AI) advancements. There were a whole bunch of announcements during the virtual event including Nvidia Maxine – AI-powered tools that improve the quality of video calls, Ampere-based RTX A6000 and A40 graphics cards, an AI model that predicts oxygen needs in COVID patients, new NVIDIA Studio program, and more. Huang also shared that the Nvidia GeForce RTX 3080 and GeForce RTX 3090 shortages are expected to last till next year.
Here are the top six things you need to know about first day announcements at Nvidia GPU Technology Conference 2020
- Nvidia Maxine is a GPU-accelerated AI platform that can “vastly improve streaming quality” and offers features like super resolution, noise cancellation, gaze correction, and live captions, as per the company. It significantly reduces the bandwidth, has face alignment features to have people talk face to face, has animated avatars, and more. Live captions help in transcribing spoken words from a different language into text in your preferred language in real time.
- Nvidia has decided to switch over from its Quadro naming scheme and is calling the new graphics cards, Nvidia RTX A6000 and Nvidia A40. They are not dedicated gaming graphics cards but are designed to take care of enterprise workloads. The GPUs are built on Ampere architecture just like the gaming GeForce RTX 3000 series cards, and come with new RT Cores, Tensor Cores, and CUDA cores. The Nvidia RTX A6000 is said to deliver 2x to 6x performance gains over the previous generation. Both the RTX A6000 and A40 claim up to 2x the throughput with second-generation RT Cores, up to 5x the throughput with third-generation Tensor Cores, and have 48GB of GPU memory. The GPUs will be available starting mid-December.
- Nvidia has also announced that it plans to build UK’s fastest supercomputer that can been used by healthcare workers. It will spend over $50 million on the project. The supercomputer will be called Cambridge-1 and it will consist of 80 Nvidia systems and is expected to be online by the end of the year.
- The company has worked with Massachusetts General Brigham Hospital to develop an AI model that “determines whether a person showing up in the emergency room with COVID-19 symptoms will need supplemental oxygen hours or even days after an initial exam.” The team is calling it the EXAM (EMR CXR AI Model) initiative. Huang also explained that AI can speed up drug discovery by using simulations.
- Further, the Nvidia Studio programme gives more options to creators with Omniverse 3D simulation and Nvidia Broadcast. Omniverse is a an RTX-based simulation, collaboration, and rendering platform for 3D workflows while the Nvidia Broadcast app gets AI-powered noise removal, auto frame features, and virtual backgrounds.
- Another important piece of information that Huang shared was that the shortages of Nvidia GeForce RTX 3080 and GeForce RTX 3090 graphics cards will last till 2021. The company expects “crushing shortages” of these graphics cards to persist till the end of this year. “Even before the holiday season, we were doing incredibly well, and then you add on top of it the ‘Ampere factor,’ and then you add on top of that the ‘Ampere holiday factor,’ and we’re going to have a really really big Q4 season,” Huang added.