AI CPUs are specialized computer processors that integrate a neural processing unit (NPU). Designed to help you complete AI tasks on your local device, AI processors are appearing in more and more devices and are required to run AI assistants like Copilot and Apple Intelligence.

So, with all the AI CPUs appearing on the market, what should you buy?

AI CPU Comparison

Intel, AMD, Apple, and Qualcomm have announced new SoC (System on Chip) designs for their latest mobile processors. These new processors integrate a combination of CPU, GPU, and NPU in one chip to provide efficient AI compute capabilities. Though some of these new SoCs are still awaiting release in 2024, official announcements, design specifications, and a mix of self-reported and independent benchmarks can help us determine whether these upcoming processors are worth the wait, or if you buy an AI laptop right now.

To help you decide which AI processor to buy, here’s the latest development on AI processors from Intel, AMD, Apple, and Qualcomm.

Intel Core Ultra 200V (Lunar Lake)

​​​​​

Intel announced its new Lunar Lake processors during the Computex 2024 convention. This new line of mobile processors provides several improvements from its last design, mainly focusing on thermals, power efficiency, better GPU, and AI compute capability while still using the x86 architecture. Notable SoC design features include:

Unified Memory Architecture: Intel Lunar Lake processors now integrate LPDDR5 RAM as part of its SoC design. This allows for higher bandwidth and lower power consumption while transferring data between RAM and processor.

3nm Process:With a 3nm process, Intel packs more transistors into Lunar Lake, increasing its performance and power efficiency.

Integrated NPU: Lunar Lake SoC utilizes six NPU compute engines, providing up to 40 TOPS (Tera Operations per second) of AI computing capabilities at INT8 precision.

Disabled Hyperthreading: All eight cores (four performance cores and four efficiency cores) have hyperthreading disabled in favor of better battery life over performance.

With this new SoC design, Intel Lunar Lake processors are expected to have 3x the AI performance, up to 1.5x faster graphics processing, and around 40% more power efficiency when compared to the previous Meteor Lake processors.

AMD Ryzen AI 300 (Strix Point)

In contrast to Intel’s power efficient approach in handling x86, AMD focuses more on emphasizing performance at the expense of higher power consumption. Here are a few features that make these processors powerful:

Zen 5 Microarchitecture: Brings significant improvements in IPC (Instructions Per Clock) and overall performance.

Integrated RDNA 3.5 Graphics: Provides improvements to the previous RDNA architecture, adding substantial performance improvements on both graphics and AI related tasks.

XDNA2 NPU: The highest performing NPU on a SoC. Capable of up to 50 TOPS at INT8 precision and suitable for Copilot+ requiring 40 TOPS.

Block FP16: Enables higher precision AI tasks with little compromise on performance.

This makes AMD’s Ryzen AI 300 series of processors powerful options for demanding AI and computational tasks, leveraging advanced graphics and AI processing capabilities.

Apple M4 (Donan)

The Apple M4 uses similar technologies to the M3, such as a 3nm process node, chip-integrated memory, chiplet design, and hybrid architecture. The M4 is already integrated into the latest iPad Pro, providing 9 or 10 CPU cores (3 or 4 Performance cores and 6 Efficiency cores), a 16-core NPU capable of 35 TOPS, and a 10-core GPU four times faster than the M2. Design changes aren’t as drastic as Intel’s Lunar Lake, mostly because the M-series of chips are already well-optimized at this point, and ARM devices are simply more power efficient than their x86 counterparts.

Qualcomm Snapdragon X Elite

Qualcomm is now producing capable ARM processors for Windows machines! The Snapdragon X Elite processors run on RISC (Reduced Instruction Set Computing) instead of the usual CISC (Complex Instruction Set Computing) found on most Windows computers. Qualcomm has stated that the X Elite SoC utilizes a 12-core ARM v8 Oryon CPU, Adreno X1 GPU, and Hexagon NPU capable of 45 TOPS at INT8 precision, making it a capable Windows Copilot Plus processor. Its use of RISC paired with a powerful SoC makes Qualcomm’s Snapdragon X Elite a great competitor to Apple’s M series of chips, which are also high-performing RISC processors.

Intel vs. AMD vs. Apple vs. Qualcomm: AI Processors Compared

Here’s a table to compare how Intel Lunar Lake, AMD Ryzen AI 300, Apple M4, and Qualcomm Snapdragon X Elite compare:

Based on the table above, we have two x86 (Lunar Lake and Ryzen AI 300) and two ARM (M4 and Snapdragon X Elite) AI processors. ARM processors are generally known to provide better power efficiency, while x86 has higher performance. However, this gap between performance and power efficiency seems to be getting closer as the M4 and X Elite become more powerful, and the Lunar Lake and Ryzen AI 300 are more power efficient.

In terms of power efficiency for X86 processors, Intel has done it better with its 3nm process node, on-chip memory, disabled hyperthreading, and lower CPU core count. Meanwhile, AMD’s Ryzen AI SoC provides better performance with 24 threads of slightly higher CPU clock speeds, a significantly more powerful GPU, and NPU with block FP16 capability.

As for the ARM AI processors, Apple’s M4 beats the X Elite in thermals, CPU, and even GPU due to its hardware-accelerated tracing capability and native support for macOS applications. However, it should be noted that despite the emulation and other software issues, the X Elite chip is still a powerful ARM-based processor rivaling Apple's M3, Intel's Meteor Lake, and AMD's Ryzen 7000 processors.

Which AI CPU Should You Buy?

Laptop manufacturers often provide options for different hardware specifications, including the processor. So, with this year’s new AI-capable SoCs coming to the market, which AI CPU should you get?

Apple M4 (Donan): Best for macOS users. Designed and optimized for macOS, offering competitive performance and long battery life.

AMD Ryzen AI 300 (Strix Point): Ideal for gamers. Its high-performance multithreaded CPU, paired with a powerful integrated GPU, makes it ideal for gaming and other intensive tasks.

Intel Core Ultra 200V (Lunar Lake): Balanced performance. It offers a good balance between performance and battery efficiency. Suitable for gaming (especially E-sport titles), productivity tasks, media consumption, and general web browsing.

Qualcomm Snapdragon X Elite: Most battery-efficient Windows AI processor available. It is the first to natively support Windows Co-Pilot Plus. Great for general productivity, web browsing, and media consumption.

Although all these processors have AI capabilities through their integrated NPUs, it may take some time before we fully benefit from them. Developers will need more time to create software that fully utilizes the NPU.

While it might be tempting to buy a new laptop now, the AI capabilities on these new SoCs are significantly better than those released in 2023. So, if AI capabilities are important to you, you'll either have to get a Snapdragon X Elite PC right now or wait for the upcoming M4, Core Ultra 200V, or Ryzen AI laptops later this year.