Solution Providers Navigate Complex AI PC Landscape Amid Transformative Wave

Solution Providers Are Helping Customers Navigate Complex AI PC Landscape

AI PCs are garnering a lot of enthusiasm and support across the channel, and solution providers will be key to helping customers navigate a complex landscape. Ever since the first AI PCs started shipping last year, major tech vendors ranging from Intel and AMD to Dell Technologies, HP Inc. and Lenovo have vowed that the new device category will represent an inflection point and a new vector of growth for the client computer market. Many solution provider executives share in that optimism, including Future Tech Enterprise CEO Bob Venero, who believes the AI PC category represents a "very big opportunity" for the channel because it will eventually become the norm for personal computing. "AI PCs are going to be the standard. Eventually, you're not going to really have, in my opinion, a delineation between an AI PC and a non-AI PC because AI is going to be integrated into everything that companies are going to be doing," he said. [RELATED: CRN's 2024 AI Special Issue] The emergence of this new device category comes at a critical time: Devices purchased during the pandemic-induced buying flurry of 2020 and 2021 are nearing the end of their expected four-year life cycles, while Microsoft Windows 10 is sunsetting in roughly a year's time, pushing many businesses to upgrade their PCs to support the beefier hardware requirements that Windows 11 brings. Factor in the zeitgeist of this new age of AI and it's no wonder AI PCs are garnering a lot of attention in the channel. But it's still early days, and sales in this nascent market are more complex and a heavier lift than traditional PC sales as solution providers help customers sort out how AI PCs can best benefit them. "The opportunity is there if partners get the nuance that you need to be focusing on what the business outcome of that AI PC is to your customer's environment, and then if you can articulate that, you'll be in good shape," Venero said. First-generation AI PCs have been available for several months, and a second wave of more capable machines are starting to hit the market. While the industry has been creating buzz about AI PCs for roughly a year, analysts have predicted that sales of AI PCs would start to gain limited traction in the second half of this year before picking up steam in earnest in 2025. AI PCs are only expected to "drive modest PC growth" in 2024, according to research firm IDC. But with the PC market only just starting to make a comeback this year, the firm forecasts that AI PCs will go from representing one in five shipments in 2024 to nearly two in three PCs shipped in 2028, bringing higher average selling prices in tow. Solution provider executives believe the AI PC category will become a game-changer for businesses, but they will need to help customers navigate issues that have tempered early adoption, whether it's finding compelling use cases or justifying higher prices. One issue that may have impacted the early adoption of AI PCs was Microsoft's decision to launch its newly branded Copilot+ PCs in June with Qualcomm's Arm-based Snapdragon X processors instead of x86 chips from Intel or AMD. This added another layer of complexity into the product category because solution providers now have to scrutinize the underlying hardware more closely and help customers make a choice: go with the Arm-based Copilot+ PCs available now or wait for wider availability of new x86-based systems that feature chips from Intel and AMD and meet Microsoft's Copilot+ PC performance requirement that AI PCs must have a neural processing unit that performs at least 40 trillion operations per second (TOPS). The x86-based systems released earlier this year do not meet that mark. "The normal message would be, 'Hey, just future-proof your hardware. You're going to buy the hardware today. You'll be future-proofed,'" said an executive at a major U.S. distributor, who asked not to be named in order to speak candidly about the challenges facing broader AI PC adoption. "Then this Microsoft [Copilot+ PC definition] shows up, and it's like, 'Whoa, wait a minute, am I really future-proofed because my notebook won't do a TOPS of 40?'" Microsoft in a recent blog post said, "We have seen incredible energy and momentum for Copilot+ PCs, with customers telling us that the performance and battery life exceeds their expectations." As the new x86 chips hit the market, the question then becomes how many AI PCs will become broadly available in the commercial market in the near term, said Megan Amdahl, senior vice president of client experience and North America COO at Chandler, Ariz.-based Insight Enterprises. To date, according to Amdahl, allocations have been more focused on consumers. "In the next month or two, it'll be interesting to gain clarity on how much will get allocated to consumer versus enterprise," she said. One of the biggest issues critical to early AI PC adoption has been the need to find compelling use cases, including applications that can unlock new capabilities using the new chip architectures that are powering these computers. "One of the things that we're just paying close attention to at Insight is what other software will be forthcoming that really requires the 40 TOPS for the performance. Because Copilot+ is awesome; it's just, how many other [applications] will people be able to take advantage of [it for better productivity, collaboration or graphics]?" said Amdahl. When vendors started talking about the concept of AI PCs last year, there were only a few demonstratable use cases and even fewer applications that people could use. But over the past year, vendors like Intel and HP have been working to enable and showcase ISVs that are developing so-called killer apps. Intel, for instance, has been working with more than 100 ISVs on over 300 AI-enabled features that will take advantage of its Core Ultra processors. The features include AI-powered anti-phishing by BufferZone, accelerated data analysis by Microsoft's Power BI, AI-enhanced presentations by Canvid and AI-powered deepfake detection by McAfee. At a summer event, HP highlighted several ISVs that are making AI PCs -- in the company's own words -- "real and tangible" with apps that take advantage of the computer's underlying chipset. Those apps include business intelligence platform Polymer, photo editor Luminar Neo and presentation maker Beautiful.ai. The expanding software ecosystem underlines how AI PCs represent a "huge opportunity" for channel partners, according to Alex Cho, president of personal systems at Palo Alto, Calif.-based HP. "This is not a device opportunity. This is a device/solution/software/service opportunity," he said. "And so the fact that our partners are able to bring that solution to our customers, that's an expanded shopping basket. That seems very transactional, mechanical, but it means they're going to be able to bring a much broader set of the solution to our customers." While AI PC use cases continue to grow, solution providers have also been watching how pricing evolves since many of the initial devices come with the need to justify higher price tags, Insight's Amdahl said. However, Amdahl said, she has already received indications that pricing could come down to make AI PCs an easier sell for a greater plurality of customers. "To the extent that that actually comes to fruition, once OEMs do the manufacturing, then you [will] get a lot of IT teams saying, 'Well, I'm future-proofing. If I'm choosing between [an AI PC and a regular PC], I could go with the AI PC," she said. To Michael Affeldt, senior vice president of sales at Buffalo Grove, Ill.-based ACP CreativIT, there is another hurdle solution providers have to help potential AI PC customers get over: concerns around data governance and security. "Those are the things that are making people reluctant," said Affeldt. These concerns were underscored by Microsoft's May reveal of Recall, which was supposed to be a key feature for Copilot+ PCs at launch, giving users the ability to search for things they've previously seen on their screen using natural language queries. However, security and privacy experts quickly pointed out that because Recall searches are performed against a central database of screenshots taken of a user's desktop every few seconds, the feature could allow threat actors to "automate scraping everything you've ever looked at within seconds," as one expert laid out. Days before the release of Copilot+ PCs, Microsoft delayed Recall to improve the feature's safety and privacy safeguards by making it opt-in, requiring customers to use Windows Hello authentication to access it and encrypt its search index database. The feature is now expected to be available for external testing in October. "Security continues to be our top priority," Microsoft said in a recent blog post about the issue. As companies navigate the security implications of new AI features, Affeldt views the adoption of AI PCs as inevitable as the incoming PC refresh. "You ever seen surfers out there beyond the break just kind of sitting on their board?" Affeldt said, recalling something he told his team recently. "That's kind of where we are right now. The waves are coming, and those waves are Windows 11-capable devices, AI-capable devices and the [post-COVID] PC refresh."

The AI Transformation 'Tidal Wave' Is Coming

Many solution providers are already seeing a positive impact to their overall business from AI with more 10 percent of solution providers calling the impact "significant," according to a study from IPED, the channel consulting arm of CRN parent The Channel Company. Many solution providers are already seeing a positive impact to their overall business from artificial intelligence, with more than one-tenth of solution providers calling the impact "significant." These results are part of a recent AI study from IPED, the channel consulting arm of CRN parent The Channel Company. The study focused on the commercialization of AI in the channel and how solution providers are also using the technology to improve their own business. The research showed that 62 percent of approximately 250 solution provider respondents are seeing a positive impact today from AI, including 13 percent who classified that impact as significant. [RELATED: CRN's 2024 AI Special Issue] Looking ahead, 78 percent of respondents said they expect to see a positive impact from AI technologies for their business in the next 12 months to 24 months, including 40 percent who classified that expected impact as significant. "The tidal wave is coming, and they see it," said Mark Williams, a senior consultant with IPED. "They're worried about, 'Am I going to catch the wave or am I about to get swamped?'" IPED classifies the 13 percent of solution providers that are already experiencing significant positive impact as "AI pioneers," Williams said. The small share of AI pioneers shows that the field isn't too crowded for solution providers early in their AI practices or still debating whether to start an AI practice. For AI pioneers aggressively investing in AI, the IPED study found that they are putting their money into: * Business process and business transformation advisory services The majority of their early AI business comes from larger customers, infrastructure deployments, and business and data consulting. Solution providers reported "a number of large-scale infrastructure deployments that are multimillions of dollars," Williams said. "There are just not tons of them. But they are happening." Williams said that he sees AI as more akin to the advent of the internet and smartphones than the proliferation of the cloud. "This is going to change everybody's way of doing everything," he said. Cloud "was just a technology refresh, in my opinion. ... The internet changed who we are, what we do and how we actually live. That's what this is." Virtusa, based in Southborough, Mass., and No. 20 on CRN's 2024 Solution Provider 500, is among the solution providers seeing just about every customer experiment with AI, Surajit Bhattacharjee told CRN in a recent interview. The senior vice president and global lead for generative AI at Virtusa said that customers who started heavy experimentation in 2023 have now moved AI products into production. "We have helped clients move upward of 10 solutions into production [since November]," Bhattacharjee said. Customers overall have been emphasizing ROI on AI projects, with more up-front diligence around AI use cases, and working backward from business problems to see if AI can solve them, he said. For Virtusa, an example of an AI use case in production includes transcribing while a customer is on a call with a support agent and then the AI program using the transcript to surface relevant information on the agent's screens -- and then after the call, making recommendations to the agent for follow-up. Another AI solution Virtusa is piloting with a customer leverages AI and drone footage to detect damaged boxes in a warehouse, Bhattacharjee said. AI "is becoming part of every conversation," he said. "Upward of 90 percent of conversations are having AI come up in some form or other." IPED's research also showed that 87 percent of solution providers are largely learning, experimenting and working to identify customer AI use cases and the related products and services they can develop and offer. Bobby Guerra, CEO of Jacksonville, Fla.-based Microsoft partner Axiom, is among the solution providers that are not recommending generative AI tools to every customer. Guerra remains concerned about data governance and the information AI tools aggregate, he told CRN in a recent interview. The risk of an AI user learning about sensitive business data they shouldn't be privy to is too high in his view. "It really unlocks a whole bunch of questions that we just aren't really very comfortable or feel prepared to be able to tell a client, 'This is how it's going to be,'" he said. "In those situations where you've got risk that you're unsure and unclear about how it could be, you tell clients, 'Hey, let's steer away from that for right now until we get some more clarity.'" For customers especially interested in AI adoption, Guerra and his team have said, "You can do some sampling, testing of it, but we're not recommending rolling it out organization wide yet." Solution providers surveyed by IPED overall called cybersecurity and technology accuracy the biggest inhibitors to AI business success. Solution providers described by Williams as "AI skeptics" -- indifferent or not interested in AI -- were more than 50 percent more likely to cite personal and individual privacy and accuracy as top inhibitors to AI success compared with AI pioneers. Larger and AI pioneer solution providers, meanwhile, were 50 percent more likely to cite the rate of change for AI as an inhibitor compared with their counterparts. The technology is changing too quickly to make investment decisions, according to these solution providers. Smaller solution providers were 50 percent more likely to cite cybersecurity as an inhibitor, according to the research. A large share of solution providers told IPED that they believe widespread adoption will come based on effectively establishing proof of value -- not just proof of concept. Law firms, health-care providers, researchers and other potential AI users want to see how they can save time going through large amounts of data and files. In other words, vendors and solution providers need to prove the business value to customers. The technology's capabilities alone won't win over users -- customers are "not going to spend $1 million to get increased productivity of 10 people," Williams said. Many solution providers believe 2024 is the year customers will determine proof of value for AI. The research showed that 56 percent of AI pioneer solution providers believe the cloud will have a significant impact due to hyperscaler partnerships. About 60 percent said they recognize the potential for expanding application and solution development, but only 42 percent said they have application development capabilities today, according to the research. The AI opportunity cuts across a variety of practices, with solution providers potentially seeing positive gains in security, risk management, analytics services, platform and data management services, industry-based offers and infrastructure optimization in the next 12 months, according to the research. The technology could also provide "a forcing function around upgrades in a lot of different areas," including PCs that enable AI use cases, Williams said. AI is "like the internet. You're not going to stop it," Williams said. "The question is, from a solution provider [perspective], how do you make sure you position yourself to help your clients the most?" Williams warned that solution providers shouldn't underestimate the emotional component that comes with businesses deciding to adopt AI compared with other areas of technology transformation. The growth of the cloud and 5G are "way different than, 'We have the ability to have someone think for us,'" he said. He also warned solution providers against getting too absorbed in news around how AI affects vendor stock prices. "Partners make money on revenue, not stock prices," he said. For Williams, a lesson from a prior IPED study on IoT should hold true for AI and should show vendors why investing in their partner program can help with AI adoption by customers. "Trusted adviser status is absolutely critical because there's too much misinformation out there," Williams said. "Clients want to talk to someone that they can trust to give them the right answer ... that's why, in a lot of our IoT work, the majority of the deals were coming from existing customers that had existing relationships with partners in the beginning."