Mercedes and Nvidia together for the robot car, which is updated as a mobile phone

SUPPLIERS OF the 3rd MILLENNIUM – A few days after the announcement of the suspension of the collaboration between BMW and Mercedes, which at the beginning of 2019 had signed an agreement for the development of technologies in the field of autonomous guide, the latter has not lost time and has sigliato an agreement with the giant of microprocessors Nvidia. And it is a true sign of how it is changing the automotive industry: a manufacturer with more than 100 years of history, is an alliance with a company founded in 1999, with all other interests: computer graphics for gaming and advanced modeling. The turn of Nvidia to the automotive sector took place already for years and the prototype Sony saw at CES (here to learn more) is just one of many examples of the commitment of this name in the Hi-Tech.

SOFTWARE INSTEAD OF the METALMercedes and Nvidia will work together to create a software-defined computing architecture. No pistons, turbochargers and valves, but an infrastructure composed by hardware and software. Everything is aimed to create cars of the new generation that can be updated via the internet and possess artificial intelligence, and driving ability alone. The fruits of this agreement will see the light in 2024, and the two partners aim to develop the it architecture of most advanced and sophisticated ever implemented in a car. The platform not only accentrerà vehicle management, today fragmented into many “control units”, but will give the car the ability to drive from the sun along the routes stated. In addition, there will be many future security applications and convenience of use.

TRILLION OF TRANSACTIONS – THE owners can purchase and add new features, applications, and subscription-based services via software upgrade, with the prospect of going to the autonomous guide of level 4 and even 5. This flexibility will, for example, to upgrade the vehicles to adapt them to new regulatory frameworks and technological (such as the future standard of data networks). The chip at the base of this ductility is the most powerful Orin, a monster that integrates 17 billion transistors to perform 200 trillion (thousand billion) operations per second. It is not a device that is low cost, but since it is programmable, open source tools (available for free), the necessary investment will be amortized through multiple generations of products: the ability of the calculation are, in fact, to think of an obsolescence that is anything but quick.