Resellers should appoint 'AI czars', says analyst

Moor Insights and Strategy's Patrick Moorhead talks up AI opportunity for channel following deal between Nvidia and Baidu

An analyst has waxed lyrical on the lucrative opportunity artifical intelligence (AI) offers the channel following the inking of a broad partnership between graphics card maker Nvidia and Chinese cloud giant Baidu.

The deal between the duo touches industries including autonomous cars, cloud computing and consumer technologies, extending the fast-growing market and highlighting opportunities channel partners can act on now.

Solution providers from resellers, to SIs and MSPs will play a role in a booming AI market that IDC analysts in its Worldwide Semiannual Cognitive Artificial Intelligence Systems Spending Guide predict will grow this year by more than 59 per cent over 2016 to $12.5bn and reach $46bn by 2020.

The key will be for channel partners to aggressively develop AI expertise and programmes within their companies and to establish themselves as thought leaders in the market, according to Patrick Moorhead, principal analyst with Moor Insights and Strategy.

"There's opportunity just like the opportunity there is with analytics because, in a way, AI is just another way of doing analytics," Moorhead said. "In the hierarchy of analytics, machine learning is at the top. … Whether it's super-sophisticated SIs all the way to resellers, where there is not a lot of value added there is opportunity for people to make money from AI."

The idea of AI has been around for decades, but it wasn't until the past several years - with growth in parallel computing fueled by more powerful GPUs and CPUs, the massive amounts of data being generated by the rapidly proliferating number of devices and systems and the development of deep learning techniques - that the AI industry and market started expanding quickly.

Now, most top-tier tech vendors, from system makers, like IBM, to software and cloud providers like Microsoft, Google and Amazon, to component makers like Intel and Nvidia are aggressively expanding their AI and machine learning efforts.

The deal between Nvidia and Baidu illustrates the different ways AI can be used in a wide range of markets. Baidu will use Nvidia's next-generation Volta GPUs in its public cloud to deliver a robust deep learning platform to its customers and the chip maker's PX platform for its self-driving car effort and to help Chinese carmakers create autonomous vehicles. Baidu will optimise its PaddlePaddle open-source deep learning framework for the Volta GPUs and make it available to researchers and academic institutions and will add its DuerOS conversational AI system to Nvidia's Shield streaming console.

For channel partners, such a fast-growing and far-reaching industry means a wide-open field for growing their businesses. VARs will be able to resell AI platforms from vendors, such as IBM, Moorhead said. Big Blue offers PowerAI, a toolkit that includes all of the top machine learning platforms, such as TensorFlow, Caffe, Theano and Chainer, running on IBM Power systems.

Partners also can integrate their own AI platforms or those from other companies atop hardware to offer to customers and "establish though leadership with clients or create new accounts", he said.

In addition, Moorhead pointed out that cloud providers like Microsoft Azure and Google Cloud Platform offer channel partners a percentage of the revenue when they bring in customers to use the AI platforms they offer on their public clouds.

SIs and other solution providers also can also help stand up AI clouds and services that customers can leverage, he noted.

There are additional steps channel firms can take to gain traction in the AI space, Moorhead said. Right now, most of the momentum in AI and machine learning is happening with component makers, like Nvidia, Intel and Advanced Micro Devices, and with public cloud providers, Moorhead said. Partners will want to look for how they can add value to what is currently being offered and developed, he added.

Solution providers also need to educate themselves about AI, develop the kinds of applications that customers can use in AI environments and establish relationships with the top vendors in the market.

At the same time, channel firms should appoint someone inside the company to be their "AI czar", who can communicate across the company to "get it ready for the next big wave", he said.

All of this will be important for a company to establish itself as a thought leader in AI, according to Moorhead, who said that instead of simply telling customers that they're learning about AI, channel companies can say that they have programs and applications in place and an executive who knows the market and what customers need.

"You're not just saying, ‘We're thinking about it and researching it' but, ‘We're really doing it'," he said.