This repository includes utilities to build and run NVIDIA Docker images. The full documentation is available on the repository wiki. A good place to start is to understand why NVIDIA Docker is needed in the first place. A signed copy of the Contributor License Agreement needs to be provided to email@example.com before any change can be accepted.
NVIDIA isn't just showing off its Titan RTX GPU and some clever AI demos -- it also has big news for anyone interested in more realistic computer physics. The company is releasing its hardware-accelerated PhysX simulation engine as an open source project, making it accessible to virtually everyone. It's a recognition that the technology is useful for more than just convincing game physics, NVIDIA said. PhysX can help with more accurate AI and robotics simulations, including self-driving car technology. You could see vehicles and bots that are better-prepared for real-world conditions.
Nvidia has announced a new partnership with Bosch to sell its Drive PX 2 driver-assist platform to automakers. In effect, the deal gives Nvidia a go-to-market strategy for its self-driving hardware and software platform. Bosch joins ZF as the two so-called tier-one suppliers that will sell Nvidia's technology to automakers. Nvidia's technology uses "deep learning" artificial intelligence, which is a fancy way of saying its computer brain learns like a human does: instead of needing to be programmed for every possible driving scenario, it learns what the appropriate behavior is, even for unexpected situations. Theoretically, a car company looking to make its car capable of autonomous driving will also be able to go to Bosch or ZF and buy that technology to integrate into their cars, and sell those to consumers.
The work we've seen Google and Uber put into self-driving cars is impressive, but short of Tesla's autopilot mode, consumer applications of autonomous driving are few and far between. Speaking at NVIDIA's GPU Technology Conference, company CEO and Founder Jensen Huang announced that Toyota has chosen the NVIDIA Drive PX platform as the heart of it future autonomous vehicles. Huang stopped just short of announcing an actual vehicle or a release date, but says NVIDIA's engineers have been working closely with Toyota to build a vehicle that will hopefully hit the market sometime in the "next few years." When it does, however, it should have all the autonomous drive benefits of NVIDIA's self driving test car. That, Huang could show -- including standard features like autopilot mode, a co-pilot feature that can take over the wheel on familiar routes, such as a commute, and safety features that can prevent a driver from passing into an intersection if it detects traffic running a red light.