Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
There weren’t many silver linings during the dark times of 2020, but if there had to be one, it was the robust sales momentum in the PC market that only now has begun to taper off. Yet PC makers remained confident, as a new chapter in computing devices unfolded, driven by artificial intelligence (AI) – or what some called AI PCs and others Copilot+ PCs.
Apple announced refreshes for its iMac, Mac mini and MacBook Pro product lines this week, showcasing the Apple Intelligence suite built on the neural processing capabilities of their new M4 chips.
In Q3 2024, research firm IDC reported that global PC shipments declined 2.4% year-over-year (YoY) to 68.8 million units. However, as Arnold Su, vice president for Consumer & Gaming PCs at the System Business Group for Asus India, explained to HT, the PC ecosystem’s emerging multilayered definition was proving to be its strength.
“If you follow the general bigger definition of AI PC, business now is between 5% or 10% of the overall PC market,” said Su. He referred to the various AI PC definitions currently circulating in the industry.
Industry experts expected newer AI PCs – including Copilot+ PCs powered by Qualcomm, Intel and AMD’s chips, as well as Apple’s M4-based Macs – to drive premium sales in the coming months.
In early 2024, Intel’s newer generation chips introduced neural processing units (NPUs) rated at 10 TOPS (trillion operations per second). By summer, that baseline evolved to 45 TOPS, with Qualcomm and AMD’s new chips introducing even more powerful NPUs.
NPUs are crucial for on-device AI computing. Common AI use-cases included chatbots, writing assistants for documents or email, summarization tools for mails or PDFs, and generative media. While Intel’s new Lunar Lake chips achieved 45 TOPS performance, Apple’s benchmarks might prove difficult to match.
“With the world’s fastest CPU core, immensely more powerful GPUs, and the fastest Neural Engine ever, the power-efficient performance and capabilities of the M4 family extend its lead as the most advanced lineup of chips in the industry,” said Johny Srouji, Apple’s senior vice president of Hardware Technologies.
Microsoft opted for the term Copilot+ PC rather than AI PC, with significant stakes in the game as their Copilot assistant became more deeply integrated within Windows 11 and Microsoft’s services.
“Every Copilot+ PC comes with your personal powerful AI agent. Copilot will now have the full application experience customers have been asking for,” said Yusuf Mehdi, executive vice president, consumer chief marketing officer at Microsoft.
Copilot utilises advanced AI models from OpenAI.
Despite varying branding approaches, PC makers were moving in a unified direction. HP, Asus and MSI had recently expanded their AI PC portfolios. “We’re at the beginning of a new era of AI PCs that will redefine what a personal computer can do,” said Ipsita Dasgupta, senior vice president and managing director, HP India.
Eric Kuo, MSI’s executive vice president and general manager of NB Business Unit, described their portfolio, including Stealth 18 and Creator 16, as “the industry’s most comprehensive AI+ PC lineup.”
“The debate has moved from speculating which PCs might include AI functionality, to the expectation that most PCs will eventually integrate AI NPU capabilities,” said Ranjit Atwal, research director at Gartner.
PC makers’ confidence stemmed from the fact that more users would increasingly find relevance for AI and therefore optimised computing devices for workflows. Software companies including Adobe, Blackmagic Design that made DaVinci Resolve Studio video editor, and Canva subsidiary Serif which built Affinity editing software, had added NPU support.
There are differing predictions on when this transformation would take effect. Research firm Gartner estimated worldwide shipments of AI PCs will touch 114 million units in 2025, an increase of 165.5% from 2024’s 43 million shipments.
Linn Huang, IDC’s research vice president of devices and displays, believed it will be 2026 before we saw a substantial shift. “While we expect AI to reach ubiquity at some point at the end of this decade, the ramp up towards mass market will take longer than expected. The next year will be largely about developing software, use cases, and target audiences for this AI-enabled hardware,” he said.
The pace of change will also depend on price tags. At this time, the most inexpensive Copilot+ PC from HP cost around ₹1,20,000. It was similar for Asus’ portfolio, with pricing defined by high-end AI chips in use. Consumers and businesses might not find true value until that generational premium deflates.