NVIDIA (NASDAQ:NVDA) turned in robust fiscal third-quarter 2018 results earlier this month. The graphics-chip specialist's revenue jumped 32%, GAAP earnings per share surged 60%, and adjusted EPS soared 41%.
On the Q3 analyst conference call, talk centered on competition. That's because other companies have, or are exploring, approaches to artificial intelligence (AI) to rival NVIDIA's graphics processing unit (GPU) based approach to deep learning. In addition, news out earlier this month that Intel poached Advanced Micro Devices' former graphics head strongly suggests that the former will enter NVIDIA's turf, the discrete GPU space. The chip behemoth surely wants a piece of the AI business that's been propelling NVIDIA's data-center platform to torrid triple-digit year-over-year revenue growth in recent quarters.
CEO Jensen Huang outlined two of NVIDIA's key competitive advantages:
1. Focus on one architecture
From Huang's remarks:
The fact that we are singularly focused and completely dedicated to this one architecture [CUDA]... allows everybody to trust us and know that we will support it for as long as we shall live. ... When you have four or five different architectures to support ... and you ask them [customers] to pick the one that they like the best, you're essentially saying that you're not sure which one is the best. And we all know that nobody's going to be able to support five architectures forever. ... [S]omething has to give, and it would be really unfortunate for a customer to have chosen the wrong one.
Huang commented further on this topic:
And that's the reason why NVIDIA could be an 11,000-people company and arguably, performing at a level that is 10 [times] that. [W]e have one singular architecture that's ... accruing benefits over time instead of three, four, five different architectures where [a] software organization is broken up into all these different small subcritical mass pieces.
CUDA is NVIDIA's parallel computing platform and application programming interface. Thanks to CUDA, NVIDIA's GPUs can be used for general-purpose processing.
The benefits Huang outlined stemming from NVIDIA's having just one architecture, as opposed to multiple ones like its primary competitors, make good sense. The flip side of having all its eggs in one basket, of course, is that NVIDIA would run into huge trouble if its sole architecture fell out of favor for some reason. At this point, it looks like clear sailing at least through the intermediate term for NVIDIA's CUDA-enabled GPUs in AI and in high-performance computing.
2. Seven-year head start in GPU-based approach to deep learning
Huang believes that NVIDIA's seven-year head start in deep learning -- a category of AI that essentially trains a machine to think like we humans do -- is a significant competitive advantage. His below remark is in response to a question about Intel's expected entrance into the discrete GPU business:
They're [NVIDIA CUDA-enabled GPUs] the most complex processors built by anybody on the planet today. And that's the reason why IBM uses our processors for the world's largest supercomputers. That's the reason why every ... major cloud [provider and] every major server maker in the world has adopted NVIDIA GPUs. ... The amount of software engineering that goes on top of it is significant as well. And so if you look at the way we do things, we plan a road map about five years out. It takes about three years to build a new generation. ... And on top of that, there are some 5,000 engineers working on systems software and numerics libraries and solvers and compilers and graph analytics and cloud platforms and virtualization stacks in order to make this computing architecture useful to all of the people that we serve. ... And that's the reason why we're able to speed up applications by a factor of 100. [Emphasis mine.] You don't walk in and have a new widget and a few transistors and all of a sudden speed up applications by a factor of 100 or 50 or 20.
Intel has deep pockets and it now has a graphics head with deep experience at AMD, NVIDIA's arch-rival in discrete GPUs, but I agree with Huang that NVIDIA's big head start gives it a powerful competitive advantage.
The italicized sentence: The reference is to NVIDIA's newest GPU architecture, Volta, released earlier this year, as being 100 times faster than Kepler, its GPU architecture from four years ago, according to the company. Huang has previously said that Kepler was already 10 times faster than central processing units.
Lastly, NVIDIA has another important competitive advantage worth mentioning: It's run by a founder-CEO. Success for a founder is more than about money, so it's not surprising that a growing number of studies show that founder-led companies outperform in the stock market.