Last week NVIDIA hosted its annual GPU Technology Conference in San Jose and I attended the event to learn about what technologies and innovations the company was planning for 2018 and beyond. NVIDIA outlined its advancements in many of its growing markets including artificial intelligence, machine learning, and autonomous driving. We even saw new announcements around NVIDIA-powered robotics platforms and development capabilities. Though we were missing information on new GPU architectures or products aimed at the gaming community, there was a lot of news to take in from the show.
I came away from the week impressed with the execution of NVIDIA and its engineering teams as well as the executive leadership that I got to speak with. CEO Jensen Huang was as energetic and lively as I have ever seen him on stage and he maintained that during analyst and media briefings, mixing humor, excitement, and pride in equal doses. It is one of the areas of impact that a show like GTC can have that doesn’t make it to headlines or press releases but for the audience that the show caters to, it’s critical.

GTC remains one of the last standing developer focused conferences from a major technology hardware company. Though NVIDIA will tell you (as it did me) that it considers itself as much a software company as chip company, the fact is that NVIDIA has the ability to leverage its software expertise because of the technological advantages its current hardware lineup provides. While events like Facebook F8, Cisco DevNet, and Microsoft BUILD continue to be showcases for those organizations, hardware developer conferences have dwindled. Intel no longer holds its Intel Developer Forum, AMD has had no developer focused show for several years, and giants like Qualcomm and Broadcom are lacking as well.
GTC is has grown into a significant force of change for NVIDIA. Over the 10+ years of its existence, attendance has increased by more than 10x from initial numbers. The 2018 iteration included more than 8,500 attendees that included developers, researchers, startups, high-level executives from numerous companies, and a healthy dose of media.
NVIDIA utilizes GTC to reach the audience of people that are truly developing the future. Software developers are a crucial piece and the ability to instill them with information about tool sets, SDKs, and best practices turns into better applications and more usage models applied to GPU technology. The educational segment is impressive to see in person, even after many years of attendance. I find myself wandering through the rows and rows of poster boards describing projects that include everything from medical diagnosis advancements to better utilization of memory for ray tracing, all of course built on GPU processing. It’s a reminder that there are real problems to solve and that much of the work is still done by these small groups of students, not by billion-dollar companies.
Of course, there is a benefit to NVIDIA. The more familiar these developers and researchers are with the technology and tools it provides, both in hardware and software, the better the long-term future for NVIDIA in the space. Technology leaders know that leading in technology itself is only part of the equation. You need to convince the right people that your better product is indeed better and provide the proof to back it up. Getting traction with development groups and fostering them with guides and information during the early stages of technological shifts is what helped created CUDA and cement it as the GPU compute language of choice for the better part of a decade.
NVIDIA wants the same to occur for machine learning and AI.
The GPU Technology Conference is the public facing outreach program that NVIDIA spends a tremendous amount of money hosting. The beginnings of the show were bare and had equal parts gaming and compute, but the growth and redirection to focus on it as a professional development event prove that it has paid dividends for the company. Just look at the dominance that NVIDIA has in the AI and ML spaces that it was previously a non-contender in; that is owed at least in part to the emphasis and money pumped into an event that produces great PR and great ideas.
As for other developer events, the cupboard is getting bare for hardware companies. Intel cancelled the Intel Developer Forum a couple of years back. In hindsight, this looks like an act of hubris, that Intel believed it was big and important enough that it no longer needed to covet developers and convince them to use its tech.
Now that Intel is attempting to regain a leadership position in these growing markets that companies like NVIDIA and Google have staked ground in, such as autonomous driving, artificial intelligence, and 5G, the company would absolutely benefit from a return of IDF. Whether or not the leadership at Intel recognizes the value that the event holds to developers (and media/analyst groups) has yet to be seen. And more importantly, does that leadership understand the value it can and should provide to Intel’s growing product groups?
There are times when companies spend money on events and marketing for frivolous and unnecessary reasons. But proving to the market (both of developers and Wall Street) that you are serious about a technology space is not one of them. NVIDIA GTC proves that you can accomplish a lot of good with this and I think the success that it has seen in areas of machine learning prove its value. What started out as an event that many thought NVIDIA created as hubris has turned into one the best outward signs of being able to predict and create the future.
There is some nice and utilitarian information on this site.
A big thank you for your blog.Really looking forward to read more. Want more.
You have noted very interesting details! ps decent web site.
Good article with great ideas! Thank you for this important article. Thank you very much for this wonderful information.