In the next part of the article that I was not able to fit into the excerpt it says...
According to Gartner data released in August, AI PCs will account for 31% of worldwide PC market by the end of the year. The research firm also shared these forecasts:
— AI PCs will be become the Norm by 2029 — AI PC shipments will total 77 million in 2025 — AI PCs market share will climb to 55% in 2026
After that there are descriptions and links to other articles. Starting with:
Related AI PC coverage:
AI PCs to surge, claiming over half the market by 2026 August 28, 2025: Despite some market hurdles, AI PC adoption is taking hold, with about 143 million units expected to ship next year. according to Gartner research.
IT buyers are investing in AI PCs — with no clue how to use them July 24, 2025: Eyeing the upcoming end of support for Windows 10 in October, many IT buyers are snapping up Windows 11 AI PCs despite having no clear understanding of how to use them in their infrastructures. “The use cases around AI PCs are not fully fleshed out just yet,” said Jitesh Ubrani, research manager at IDC. “Companies are struggling to figure out if AI PCs are worth the extra investment or not, especially when there’s a lot of AI that can be done in the cloud, as opposed to on the device itself.”
Linux AI PCs?
Get a computer soon if you want to avoid AI.
Just like most people who are constantly looking into computers and possible upgrades... I am confused about what benefit the latest generations of CPUs, GPUs and APUs will actually be for me. There ae people here who have greater insight into this than I do. Hopefully a few of them will contribute.
Here is what Grok says about Linux on the latest generation AI enhanced processors:
“Can Linux take advantage of the latest AI enhanced processors and computers?
“Yes, Linux can take advantage of the latest AI-enhanced processors and computers, as it supports a wide range of hardware architectures and has robust software ecosystems for AI workloads. Here’s a concise breakdown:
### Hardware Support
- **AI-Enhanced Processors**: Linux supports modern AI-optimized hardware like Intel’s Xeon with AMX (Advanced Matrix Extensions), AMD’s Ryzen AI and EPYC with NPUs (Neural Processing Units), NVIDIA GPUs (CUDA/ROCm), and specialized AI accelerators (e.g., Google TPUs, Intel Gaudi, Graphcore IPUs).
- **Driver Availability**: Open-source drivers and proprietary frameworks (e.g., NVIDIA CUDA, AMD ROCm) are well-integrated into Linux distributions, enabling efficient use of AI hardware for tasks like matrix multiplication and neural network inference.
- **NPUs and TPUs**: Linux supports NPUs and TPUs through frameworks like TensorFlow, PyTorch, and ONNX, with kernel-level optimizations for accelerators in distributions like Ubuntu, Fedora, or specialized AI distros.
### Software Ecosystem
- **AI Frameworks**: Linux is the primary platform for AI development, supporting TensorFlow, PyTorch, JAX, and others, which are optimized for AI hardware via libraries like cuDNN, oneAPI, or OpenVINO.
- **Kernel Optimizations**: The Linux kernel includes features like heterogeneous computing support (e.g., for CPUs, GPUs, NPUs) and scheduling enhancements for AI workloads, such as improved power management for AI chips.
- **Community and Tools**: Linux’s open-source community actively develops tools like OpenCL, Vulkan, and ROCm for AI acceleration, ensuring compatibility with cutting-edge processors.
### Practical Examples
- **NVIDIA GPUs**: Widely used in Linux for AI training/inference with CUDA and cuDNN, supported by major distros.
- **Intel AMX**: Linux kernels (5.16+) support AMX for AI workloads, used in data centers and edge devices.
- **AMD NPUs**: ROCm on Linux enables AI acceleration on AMD’s Instinct accelerators and Ryzen AI chips.
- **Edge AI**: Linux powers AI on edge devices (e.g., Raspberry Pi with Coral TPUs) with lightweight distros and frameworks.
### Limitations
- **Driver Maturity**: Some newer AI chips (e.g., AMD NPUs) may have less mature Linux support compared to NVIDIA’s ecosystem.
- **Proprietary Software**: Certain AI hardware relies on proprietary drivers, which may lag in Linux support or require manual setup.
- **Configuration Complexity**: Optimizing Linux for specific AI hardware may require kernel tuning or custom builds, which can be complex for beginners.
### Conclusion
Linux is well-positioned to leverage AI-enhanced processors due to its flexibility, open-source nature, and strong AI software ecosystem. For optimal performance, users should choose a distro with good hardware support (e.g., Ubuntu, CentOS) and ensure drivers/frameworks are updated. If you’re targeting specific AI hardware, I can search for the latest driver compatibility or provide setup guidance—let me know!”
Misleading article...I really doubt that that more than 0.01% of PC/Mac users are running & using LLM’s (”AI”) locally on their home PC...when using MS Co-pilot, OpenAI, Grock, etc. you are using the “cloud” version on their HW/servers.
No such thing as an “AI PC”...just PC’s with loads of RAM and high power GPU’s. (Like a “Gamer” PC).
If interested in running “language models” (”AI”) locally it can be done using lower power home computers by using smaller language models.
YT video on running smaller language models on Macs:
https://youtu.be/bpIjlMJJ0I0?si=_HubaayyuXjX_Ul3
(Reminds me of the days when they were promoting new “Digital TV” antennas to get you to buy a new TV antenna! DTV rides on an analog carrier...your old antenna/rabbit ears are fine!)
AI is not creative.
All you get is what the original programmer
did and what data bases are available.
I have a BS in Computer science and I
can create AI machines/engines.
I’m also an electronics engineer
as well as technician.
My point is this;
AI cannot deal with the un-expected
or random nature of our lives.
It is algorithmic, it cannot handle
random events like an intelligent person can.
So we replace all our lower intelligence
people with AI,
so what do we do with them?
They need to eat, drink and survive in
a society that can no longer use them.
Ask your AI that question!
"I'm sorry, Dave. I can't let you do that."