GTC Europe: Busy—and Rewarding
Here we are again, back for Day 2 of GTC—the NVIDIA GPU Technology Conference—in Amsterdam. Yesterday was a busy day but, refreshed by one or more of Amsterdam’s famous beers yesterday evening, we’re ready to go again. In fact: by the time you read this, I’ll have already delivered my presentation (see picture above) on the use of GPU technology in military ground vehicles. There was standing room only—which was very gratifying.
Yesterday’s show started with the keynote speech—delivered to a packed house, as is traditional at GTC—from NVIDIA’s CEO and co-founder Jen-Hsun Huang, who highlighted the evolution of the GPU from the PC and Internet in 1995 through to mobile/cloud in 2006 and to the dawn of the AI revolution in 2016.
There were some exciting product announcements, with the P4/P40 AI-dedicated processor targeted specifically at the AI deep learning markets. The later P4 will, apparently, come in at an astonishing 50 watts.
Possibly even more exciting was the announcement of the new SoM codenamed Xavier. This 8-core ARM 64-bit CPU will be paired with a 512-core Volta GPU and is expected sometime in 2017. It packs an impressive 7 billion transistors in a 16nm package. This processor could be ideal for embedded applications in the mil/aero sector in both avionics and in ground vehicles.
If you’re interested to know more of the detail of what Jen-Hsun Huang said, you may like to read the NVIDIA blog on day two of GTC Europe here.
You may also want to take a look at a new white paper we've just published on the use of GPU technology to enable the development of autonomous vehicles such as driverless cars.
Meanwhile, back on the Abaco stand, we had a great response to our TX1 embedded graphics and computer vision concept (see our press release here), with a lot of really positive comments to digest before we go to market with the final qualified product.