Janice R. Andrews
Intel’s imminent release of its desktop CPUs, the Arrow Lake-S, is set to mark a significant milestone. These chips will be the first to be used in a desktop PC with a neural processing unit (NPU). Even though this is an excellent milestone, the NPU won’t be a major player in AI. It might not even be eligible for Microsoft’s Copilot+ PC program, in instance. Does it matter, though?
Unverified Rumors and Speculations
Intel has not yet verified the Arrow Lake-S specifications. On X (previously Twitter), a person going by the handle Jaykihn posted a list of speculated specs. These comprise the GPU, CPU, and NPU’s trillion operations per second (TOPS) numbers. The leak indicates that the NPU will have a meagre 13 TOPS of AI performance. In contrast, AMD’s Ryzen AI 300 series has 50 TOPS, while Intel’s Lunar Lake CPUs, which are exclusively found in laptops, have 45 TOPS. The Microsoft Copilot+ certification, which calls for at least 40 TOPS, is applicable to both of them. Why therefore would Intel introduce a next-generation product that has only a small portion of the AI functionality found in laptops already?
NPUs: Not Required for Desktop Computers
The unique requirements of desktop users hold the solution. Ultralight workstation notebooks use NPUs to minimize battery effect while handling AI and machine learning tasks. Desktop users do not have to worry about battery life, hence a high-TOPS NPU is not as necessary. High TOPS counts in NPUs are mostly found in laptops where battery efficiency is a key concern, as Tom’s Hardware notes. One important benefit of having an NPU is practically null and void in a desktop environment.
Discrete GPUs’ Function
Discrete GPUs operate well in desktop PCs designed for AI applications. For example, Nvidia’s RTX 4090 boasts up to 1,321 AI TOPS, a substantial improvement above any NPU. This explains why the RTX 4090 is still so expensive even after almost two years of release. These days, desktop GPUs and laptops with AI are driven by demand. Even with a more powerful NPU, Intel’s Arrow Lake-S couldn’t compete with a discrete GPU from Nvidia’s most recent generation. An NPU would not match the AI power of a less expensive GPU.
Marketing versus Usefulness
Even if the Arrow Lake-S has limited effect on real AI workloads, Intel might use it to promote the first desktop NPU for customers if the rumored specs are true. Basically, if the CPUs have to have an NPU, it’s better if it’s not very powerful because there aren’t many desktop situations where a 45 TOPS NPU would be useful.
AI’s Role in Desktop Environments
Some features, including the now-delayed Recall, are unclear regarding their futures given that Arrow Lake-S has an NPU. This capability has only been available on Copilot+ PCs, but if it becomes popular, Microsoft might soon release it on desktops, according to rumors. Although the role of NPUs in desktop processors may change as a result of this shift, the optimization of AI capabilities through separate GPUs is still the major priority.
With their small NPUs, Intel’s Arrow Lake-S processors mark a significant milestone, but they also emphasize the disparities in requirements between desktop and laptop consumers. For ultralight laptops to effectively handle AI tasks, NPUs are essential, while desktop PCs can achieve better AI performance by using discrete GPUs with greater power. The future of AI in desktop computing will probably change as long as Intel keeps coming up with new innovations, but for the time being, Arrow Lake-S’s tiny NPU is a useful advancement that won’t drastically change the nature of desktop AI tasks.
