• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

AI-Capable PCs Forecast to Make Up 40% of Global PC Shipments in 2025

AleksandarK

News Editor
Staff member
Joined
Aug 19, 2017
Messages
2,260 (0.92/day)
Canalys' latest forecast predicts that an estimated 48 million AI-capable PCs will ship worldwide in 2024, representing 18% of total PC shipments. But this is just the start of a major market transition, with AI-capable PC shipments projected to surpass 100 million in 2025, 40% of all PC shipments. In 2028, Canalys expects vendors to ship 205 million AI-capable PCs, representing a staggering compound annual growth rate of 44% between 2024 and 2028.

These PCs, integrating dedicated AI accelerators, such as Neural Processing Units (NPUs), will unlock new capabilities for productivity, personalization and power efficiency, disrupting the PC market and delivering significant value gains to vendors and their partners.



"The wider availability of AI-accelerating silicon in personal computing will be transformative, leading to over 150 million AI-capable PCs shipping through to the end of 2025," said Ishan Dutt, Principal Analyst at Canalys. "PCs with dedicated on-device AI capabilities will enable new and improved user experiences, driving productivity gains and personalizing devices at scale while offering better power efficiency, stronger security and reduced costs associated with running AI workloads. This emerging PC category opens new frontiers for both software developers and hardware vendors to innovate and deliver compelling use cases to customers across consumer, commercial and education scenarios."

The rapid uptake of AI-capable PCs will drive a moderate increase in the TAM value of the broader PC market. "The enhanced capabilities of this new category will create momentum toward premiumization, particularly in the commercial sector," said Canalys Analyst Kieren Jessop. "In the short term, Canalys expects a 10% to 15% price premium on AI-capable PCs compared with similarly specified PCs without NPU integration. With the spike in adoption, by the end of 2025, over half of PCs priced at US$800 and above will be AI-capable, with this share increasing to over 80% by 2028. As a result, PC shipments in this price range will grow to account for more than half the market in just four years. This will help boost the overall value of PC shipments from US$225 billion in 2024 to over US$270 billion in 2028."

Canalys' latest forecasts underscore the massive potential and wide-ranging impact AI-capable PCs will have on the industry over the next five years and beyond. Vendors that can deliver innovative, differentiated AI-accelerated experiences stand to gain significant advantage as this disruptive new category goes mainstream.

Canalys' "Now and next for AI-capable PCs" report provides a comprehensive overview of this future computing paradigm. It includes initial category definitions and forecasts, technical considerations, market opportunities and potential industry challenges. For more insights into the transformative effects of AI on other sectors of the technology industry and channel, explore here.

View at TechPowerUp Main Site | Source
 
Joined
Nov 13, 2007
Messages
10,256 (1.70/day)
Location
Austin Texas
Processor 13700KF Undervolted @ 5.6/ 5.5, 4.8Ghz Ring 200W PL1
Motherboard MSI 690-I PRO
Cooling Thermalright Peerless Assassin 120 w/ Arctic P12 Fans
Memory 48 GB DDR5 7600 MHZ CL36
Video Card(s) RTX 4090 FE
Storage 2x 2TB WDC SN850, 1TB Samsung 960 prr
Display(s) Alienware 32" 4k 240hz OLED
Case SLIGER S620
Audio Device(s) Yes
Power Supply Corsair SF750
Mouse Xlite V2
Keyboard RoyalAxe
Software Windows 11
Benchmark Scores They're pretty good, nothing crazy.
I’m confused- it takes months to train an llm on a room full of servers… then that service is served to the public that can be consumed by any digital device. What exactly does “an ai capable” 7w chip do?

apple has had an npu since the m1, still no idea what it does and when.
 
Joined
Feb 20, 2020
Messages
9,340 (6.02/day)
Location
Louisiana
System Name Ghetto Rigs z490|x99|Acer 17 Nitro 7840hs/ 5600c40-2x16/ 4060/ 1tb acer stock m.2/ 4tb sn850x
Processor 10900k w/Optimus Foundation | 5930k w/Black Noctua D15
Motherboard z490 Maximus XII Apex | x99 Sabertooth
Cooling oCool D5 res-combo/280 GTX/ Optimus Foundation/ gpu water block | Blk D15
Memory Trident-Z Royal 4000c16 2x16gb | Trident-Z 3200c14 4x8gb
Video Card(s) Titan Xp-water | evga 980ti gaming-w/ air
Storage 970evo+500gb & sn850x 4tb | 860 pro 256gb | Acer m.2 1tb/ sn850x 4tb| Many2.5" sata's ssd 3.5hdd's
Display(s) 1-AOC G2460PG 24"G-Sync 144Hz/ 2nd 1-ASUS VG248QE 24"/ 3rd LG 43" series
Case D450 | Cherry Entertainment center on Test bench
Audio Device(s) Built in Realtek x2 with 2-Insignia 2.0 sound bars & 1-LG sound bar
Power Supply EVGA 1000P2 with APC AX1500 | 850P2 with CyberPower-GX1325U
Mouse Redragon 901 Perdition x3
Keyboard G710+x3
Software Win-7 pro x3 and win-10 & 11pro x3
Benchmark Scores Are in the benchmark section
Hi,
Surprised the marking didn't say 100% by then
Damn they're slipping hehe
 

Fourstaff

Moderator
Staff member
Joined
Nov 29, 2009
Messages
10,043 (1.90/day)
Location
Home
System Name Orange! // ItchyHands
Processor 3570K // 10400F
Motherboard ASRock z77 Extreme4 // TUF Gaming B460M-Plus
Cooling Stock // Stock
Memory 2x4Gb 1600Mhz CL9 Corsair XMS3 // 2x8Gb 3200 Mhz XPG D41
Video Card(s) Sapphire Nitro+ RX 570 // Asus TUF RTX 2070
Storage Samsung 840 250Gb // SX8200 480GB
Display(s) LG 22EA53VQ // Philips 275M QHD
Case NZXT Phantom 410 Black/Orange // Tecware Forge M
Power Supply Corsair CXM500w // CM MWE 600w
I’m confused- it takes months to train an llm on a room full of servers… then that service is served to the public that can be consumed by any digital device. What exactly does “an ai capable” 7w chip do?

apple has had an npu since the m1, still no idea what it does and when.
Training takes a lot of effort, running the model takes a lot less (but still a lot of effort). What's going to happen is that we will be running other workloads such as image processing and other data analytics instead of ChatGPT.
 

MacZ

New Member
Joined
Jun 29, 2023
Messages
28 (0.09/day)
It makes absolutely no sense ...

Why would you need to have a local model running on your CPU, which will be severely limited by your RAM and your 'NPU' capabilities when having a distant model on a much more powerful and capable server will probably get you a faster and better answer ?

I understand the point of view of those that sell the hardware, but from a customer point of view, I search the logic when actual AI products that you can run locally on a GPU for example, are not that common.

Now, if it is not for the benefit of said customer ...
 
Joined
Jan 3, 2021
Messages
2,764 (2.24/day)
Location
Slovenia
Processor i5-6600K
Motherboard Asus Z170A
Cooling some cheap Cooler Master Hyper 103 or similar
Memory 16GB DDR4-2400
Video Card(s) IGP
Storage Samsung 850 EVO 250GB
Display(s) 2x Oldell 24" 1920x1200
Case Bitfenix Nova white windowless non-mesh
Audio Device(s) E-mu 1212m PCI
Power Supply Seasonic G-360
Mouse Logitech Marble trackball, never had a mouse
Keyboard Key Tronic KT2000, no Win key because 1994
Software Oldwin
How will we able to buy an AI-incapable PC in 2025?
 
Joined
May 3, 2018
Messages
2,358 (1.07/day)
Well duh. You won't even get Microsoft approval for Windows 12 if the PC doesn't have a 50 TOPS NPU and 16GB RAM minimum. So it's not consumer demand driving this, it's rapacious clown company desperate for shiny baubles to attract the moths to the light. Companies will have no choice but to build these sort of AI infested PC's.
 
Last edited:
Top