When the next Arrow Lake-S desktop processors—the first to have a neural processing unit (NPU) in a desktop PC—are released, Intel will have accomplished a major milestone. Although this is a significant advancement for desktop computers, these processors’ NPUs aren’t going to be very strong for AI applications. Indeed, it might not fulfill the prerequisites for Microsoft’s Copilot+ PC application. It’s possible that this restriction is not as serious as it first appears.
Intel has not yet confirmed the official specifications for Arrow Lake-S, but a leak from Jaykihn on X (formerly Twitter) provided a list of rumored specs. These include the trillion operations per second (TOPS) counts for the GPU, CPU, and NPU. According to the leak, the NPU in Arrow Lake-S will deliver only 13 TOPS of AI performance. This is significantly lower compared to Intel’s Lunar Lake CPUs for laptops, which offer 45 TOPS, and AMD’s Ryzen AI 300 series, which provides 50 TOPS. Both of these laptop processors meet the criteria for Microsoft’s Copilot+ certification, which requires at least 40 TOPS. This discrepancy raises the question: why would Intel introduce a next-generation desktop processor with an NPU that significantly underperforms compared to its laptop counterparts?
The reason lies in the differing needs of desktop and laptop users. NPUs are designed to handle AI and machine learning tasks while minimizing battery consumption, a crucial factor for ultralight workstation laptops. In desktop PCs, battery life is not a concern, so the primary benefit of an NPU—efficient energy use—is less relevant. For desktop users who require robust AI performance, a discrete GPU is a more effective solution. For instance, Nvidia’s RTX 4090 GPU delivers up to 1,321 AI TOPS, far surpassing any NPU currently available in consumer hardware. Even more affordable GPUs outperform integrated NPUs in terms of AI processing power.
This context suggests that the inclusion of an NPU in Arrow Lake-S might be more about marketing than functionality. Intel can claim the distinction of having the first consumer desktop NPU, even if its practical impact on AI workloads is minimal. Given the limited scenarios where a high-performance NPU would be advantageous in a desktop setting, a modest NPU might be a more appropriate choice. This aligns with the broader trend where desktop PCs handling intensive AI tasks typically rely on powerful discrete GPUs.
Additionally, it’s important to understand the landscape of AI hardware in desktop computing. While NPUs are becoming increasingly common in mobile and laptop devices to enhance battery life and enable efficient processing of AI tasks, desktops have traditionally relied on the brute force of discrete GPUs for heavy computational workloads. This trend is unlikely to change in the near future because GPUs offer vastly superior AI processing capabilities compared to NPUs. The Nvidia RTX 4090, for example, features up to 1,321 AI TOPS, which explains its enduring popularity and high price point nearly two years after its launch. The high demand for AI capabilities in both laptops and desktop GPUs underscores the importance of having powerful and dedicated hardware for these tasks.
Even if Intel were to equip the Arrow Lake-S with a more capable NPU, it would still be outperformed by discrete GPUs from Nvidia’s latest generation. This is because desktop PCs that are intended to handle serious AI workloads are better suited with a powerful discrete GPU, which can handle a much broader range of tasks more efficiently than an integrated NPU. Therefore, the decision to include a modest NPU in Arrow Lake-S could be seen as a strategic move to market the product as innovative while acknowledging the reality of desktop computing needs.
This decision also raises questions about future features and integrations, such as Microsoft’s Recall feature, which is currently exclusive to Copilot+ PCs. If Recall proves to be successful, there might be pressure to extend its availability to desktop PCs. This could necessitate a re-evaluation of the role and capabilities of NPUs in desktop processors to ensure that they can support such advanced features in the future.
In addition to adding passkey support, Google also shared that it is partnering with the media nonprofit Internews to provide cybersecurity support for its network of journalists and human rights advocates. The arrangement will cover ten countries, including Brazil, Mexico, and Poland. The partnership aims to provide essential cybersecurity support to those who are often targeted by cyberattacks due to their work. The collaboration will involve providing training, resources, and tools to help journalists and human rights advocates protect their digital identities and secure their communications. By doing so, Google and Internews aim to create a safer environment for those who play a critical role in holding power to account and advocating for human rights.
Even if NPUs have less AI performance than discrete GPUs, Intel’s Arrow Lake-S CPUs nonetheless mark a significant advancement in desktop computing. This calculated action demonstrates how the hardware environment for artificial intelligence is changing and how user needs drive the creation and incorporation of new technologies. A big step in the direction of building a more secure digital environment is Google’s decision to work with Internews and improve APP with passkey support. Google is contributing to ensuring that the internet continues to be a secure and dependable medium for all users by lowering the barrier to high-level security and offering tailored support to those who require it most.
If you like the article please follow on THE UBJ.