Over the past several years, progress in the area of artificial intelligence (AI) has reached a fever pitch. You can find AI on the agenda of nearly every large tech company in the world, and experts in the field are in high demand. Autonomous-driving technology is seeing similar demand, with companies racing to be the first out of the gate with a viable and profitable self-driving-vehicle program.
One way companies are increasing their likelihood of success in these nascent fields is through partnerships with other tech companies. This approach allows them to pool their resources, mitigate the risks, and share best practices and discoveries.
Graphics-processing pioneer NVIDIA Corporation (NVDA 1.02%) and Chinese search giant Baidu, Inc. (BIDU 0.97%) this week announced a far-reaching AI partnership that includes cloud computing, autonomous driving, consumer devices, and Baidu's open-source AI framework PaddlePaddle. The scope of this agreement is a testament to the massive size of the opportunity AI presents.
NVIDIA Telsa V100 data center GPU. Image source: NVIDIA.
The deal
Baidu announced that NVIDIA's DRIVE PX 2 AI supercomputer will be the foundation of its autonomous-vehicle platform. This news follows the revelation in April that Baidu will open-source its self-driving car technology, in a project titled Apollo. Baidu announced at its developers' conference this week that it has garnered 50 partners to commercialize its autonomous-driving technology, including such big-name partners as Microsoft, Intel, and Ford, as well as top Chinese automakers and universities.
NVIDIA's leading-edge Volta V100 GPUs and Tesla P4 GPU accelerators, which were designed specifically for AI training and inference in data centers, will be deployed to the Baidu cloud. In addition, the companies will collaborate to optimize Baidu's open-source PaddlePaddle deep-learning AI framework, which will be broadly available to academics and researchers, allowing them to develop new AI applications.
Shield TV, NVIDIA's set-top box, is designed for streaming, gaming, and smart home applications. The device will be rolled out to Chinese consumers later this year, equipped with voice command capabilities courtesy of Baidu's DuerOS AI virtual assistant.
Baidu self-driving car demonstration at the Consumer Electronics Show Asia. Image source: NVIDIA.
Everybody wins
The two companies have much to gain from the partnership. It gives Baidu the opportunity to leverage the gains NVIDIA has made in the field of accelerating AI development. By applying these gains across a broad swath of its applications, Baidu has a better shot at maintaining the lead on its home turf, and potentially gaining on its U.S. tech rivals. For its part, NVIDIA gains a firm foothold in the quickly evolving Chinese AI market.
Baidu has been working to develop new revenue sources in the face of slowing growth of advertising in its core search. In its most recent quarter, Baidu's revenue increased to $2.45 billion, up 6.8% year over year, while its earnings of $258 million fell 10.6% from the prior-year quarter. The company has worked to improve the quality of its advertisers in the face of stricter government regulations. Baidu's online customers fell by 23%, though its remaining customers spent 27% more.
For its part, NVIDIA has an ongoing and potentially lucrative market for its most advanced GPUs. The company has been reaping the rewards of its early move into the field of AI. In its most recent quarter, NVIDIA grew its revenue to $1.9 billion, a 48% year-over-year increase. The most impressive results were the product of AI growth in the company's data-center segment, which grew to $409 million, up 186% over the prior-year quarter.
The AI gravy train
Market intelligence firm IDC estimates that AI revenue will top $12.5 billion in 2017, up 59% over 2016, and accelerate to more than $46 billion by 2020. This growth rate provides companies with a powerful incentive to adopt the technology as quickly as possible.
This is a big win for both companies and one of the biggest partnerships among AI players to date, covering the most important areas of AI research. With so much at stake, don't be surprised to see more such pair-ups in the future.
