NVIDIA (NASDAQ:NVDA) made yet another graphics card announcement last week (making it the fourth in just two months). The Titan X card is built on the company's new Pascal architecture, and packs an unparalleled amount of processing power, with 12 GB of memory, 3,584 cores running at 1.53 GHZ, and an amazing 11 teraflops of performance.
NVIDIA said the new Titan X its the largest graphics processing unit (GPU) it has ever built and is 60% faster than the previous version.
The company's CEO, Jen-Hsun Huang, unveiled the card at a meeting of artificial intelligence (AI) experts at Stanford University, which is a nod to how NVIDIA's gaming GPUs are converging with AI. Huang debuted the new chip alongside Andrew Ng, who is Baidu's chief scientist and an associate professor at Stanford focusing on AI and deep learning.
The new Titan X -- along with the company's new GTX 1060, 1070, and 1080 graphics cards -- shows the company's laser-focus on drawing in even more gaming revenue, while building out a strong AI segment at the same time.
NVIDIA's strong position
NVIDIA earned $687 million from its gaming segment in fiscal Q1 2017, which was a 17% year-over-year increase. Gaming is the comapany's largest revenue driver, and it holds a dominant position in the space, taking more than 70% of the discrete GPU desktop market market share.
The company has proved that it's able to increase revenue from PC graphics cards even while sales of PCs have slowed down over the past few years. And the Titan X could keep the trend going.
Titan X is positioned the top of the high-end market with a $1,200 price tag, but its performance could be enough to convince gamers that it's worth the cost. It is, after all, NVIDIA's highest-performing gaming card to date, and gamers are more than just a little partial to the company's technology.
And the new card, along with the recent release of the $599 GTX 1080 card, together give gamers little reason to hand their money over to NVIDIA's rival Advanced Micro Devices (NASDAQ:AMD).
AMD made some market share gains against NVIDIA in the first quarter of this year, jumping up to 29.4% of the discrete GPU market share (a 3.2-percentage-point increase from the previous quarter). And the company was hoping to gain more traction with the recent debut of new GPUs from its Radeon R9 series.
But the new Titan X and other recently released cards should easily keep NVIDIA in the lead in the gaming market -- and help it expand further into AI, as well.
Building more opportunities
The new Titan X obviously has the ability to help NVIDIA to continually build on its gaming revenue, but I also think it's a way for the company to transition even further into deep learning and AI performance chips.
The Titan X comes with its own deep learning inferencing instruction (to allow it to learn certain things without having to be programmed). NVIDIA said it's giving Ng one of the first new Titan X's for his research. Ng said, "It's hard to think of an industry that will not be transformed by AI in the next decade," and added, "If you're a machine learning researcher, having access to a machine that is 2 times as fast means that you are 2 times as productive as a researcher."
NVIDIA's Tesla chips are already being used in machine learning (a type of AI) for Facebook, Amazon, and Google. The new Titan X isn't in the same class as the Tesla GPUs, but Titan does allow the company to expand its AI footprint for researchers and others looking for powerful GPUs.
NVIDIA's expansion of using its GPUs for AI can't be overstated. The biggest tech companies, researchers, automotive and enterprise companies are focusing more and more attention on artificial intelligence, and they'll need powerful processors to make that happen. GPUs that were once relegated to the gaming market are now seen as a resource for AI computers. That's pushing NVIDIA beyond its gaming roots and further into the deep learning market, which is expected to be worth $10 billion by 2024.