Tuesday, April 5 

S6699 – Opening Keynote

Jen-Hsun Huang CEO & Co-Founder, NVIDIA
packed-key

The Keynote hall was packed.  Here are some of the important highlights from Jensen’s 2-hour Keynote which set the stage for the rest of the conference:

5-miraclesFirst and foremost of interest to professionals and eventually to gamers, Pascal was released as the Nvidia Tesla P100 GPU – the most advanced hyperscale data center accelerator ever built. P100 is the latest addition to the NVIDIA Tesla Accelerated Computing Platformp100-introThe Tesla P100 enables a new class of servers that can deliver the performance of hundreds of CPU server nodes.  It is a massive 610mm chip that uses 3D stacking and HBM2 that is made possible by what Jensen called “5 miracles”.  We were also able to attend the deep dive session  which we will share with our readers in this wrap up.  Unfortunately, there is no Pascal white paper yet available.

250-serversJensen also announced Nvidia’s DGX-1 – the world’s first deep learning supercomputer which was developed to meet the intense computing demands of artificial intelligence. It features eight Tesla P100 GPU accelerators that deliver up to 170 teraflops of half-precision (FP16) peak performance.

The DGX-1 is not inexpensive at $129,000, but it will save a lot of money over the millions of dollars it would cost to buy and set up comparable CPU servers.  And the energy costs will be far lower for Nvidia’s DGX-1 system at just over 3,000W compared with the electricity usage for comparably performing CPU servers.  The first systems will go out to researchers next month, and by Q1 of next year they will be available to OEMs for general purchase.

DrivePX-AI-DLL-carsMuch emphasis was placed on future cars that will also use Deep Learning, and the World’s first autonomous race car was announced.  Not just one, but all 20 cars in the ROBORACE Formula 1 2016/2017 E circuit will use Nvidia’s DRIVE PX 2 as the “brains” of their autonomous platform.

Nvidia believes that AI for Medicine is critically important to all of us.  This year Nvidia has partnered with Massachusetts General to advance healthcare by applying AI to improve the quality of treatment and the management of diseases.

mars2033Jensen’s keynote also held the first public demo of Mars 2030 , an outer space VR simulation experience created by FUSION that uses satellite imagery and data provided from NASA. On stage, Jensen introduced the VR experience to PC pioneer Steve “Woz” Wozniak over a video link powered by Cisco TelePresence.   Jensen and Woz were joking when Woz commented that he “felt a little dizzy” by the VR experience, to which Jensen replied, “that was not a helpful comment”, much to the delight of the audience.

The full Mars 2030 experience will transport its viewer to Mars in VR. The terrain, lighting and gravity models, and even the habitats its viewer encounters in this VR simulation are grounded by scientific data. And Nvidia’s VRWorks technologies will make the experience richer by using Multi-resolution Shading and even VR SLI support.VR

Some GTC attendees were able to experience the first part of Mars 2030 which is a VR exploration of an enormous Martian lava tube where the first explorers are actually expected to live.  This experience is expected to be commercially available in the Autumn.  Unfortunately, the lines were long and we missed this experience.  We did get to check out the Oculus Rift VR demo of EVE: Valkyrie in Nvidia’s “VR Village”.

Iray-VRJensen also announced Iray VR which is Photorealistic VR rendering that lets architects and design professionals simulate their creations accurately.  Of course, photorealistic VR is computationally expensive, so there also is a scaled-down “lite” version of it available for smartphones that sacrifices little.

Jensen also announced Nvidia’s Unified SDK which helps developers create solutions for deep learning, accelerated computing, self-driving cars, design visualization, autonomous machines, robots, gaming, VR, and more.

After Jensen’s two hour keynote ended at 11 AM, we ran for the press conference.  Here is Nvidia’s Panel as moderated by their VP of Corporate Communications, Bob Sherbin:

Press-Conference-panelPress-Conference

Unfortunately, we learned very little new from this half-hour session as so many of the questions were directed at unreleased products that Nvidia’s representatives could not comment on.  We did note that Nvidia is expecting rapid growth in automotive and in deep learning applications and that their solutions are scalable.

After we grabbed a quick lunch and a brief look at the exhibits, remembering to schedule our VR experience for the next morning, we headed to the Deep Dive into Pascal architecture inside the main ballroom which was packed.

S6176 – Inside PascalUntitled-1

deep-diver-presentersMark Harris Chief Technologist, GPU Computing Software, NVIDIA
Lars Nyland Senior Architect, NVIDIA

Level: All
Type: Featured Presentation
Tags: Supercomputing & HPC

This was a fascinating look at Pascal architecture compared with Maxwell and Kepler as it applies to the new mega-chip, the P100.  Here are the slides from the presentation:

Untitled-22Untitled-21Untitled-20Untitled-19Untitled-18Untitled-17Untitled-16Untitled-15Untitled-14Untitled-13Untitled-12

Now we look at the differences between the Kepler and the Pascal cores.  Pascal cores are half the size of Kepler, but they do more work.Untitled-11

Untitled-8

Untitled-7

Untitled-6

Now here’s a look at the way the P100 is laid out:Untitled-5
Interestingly, the P100 is not the full chip as 60 SMs are possible, but in the case of the P100, 56 are enabled.

Untitled-4

Untitled-3

The above are the charts from the deep dive but here is the link to the entire presentation.

We returned to the exhibit room, grabbed a bite to eat and then headed back to our room to prepare for another full day on Wednesday.