The Chinese Taipei International Computer Show will be grandly opened on June 2.

As the exhibition approaches, all kinds of phenomenal AI PCs are also poised to launch.

Just last week,$LENOVO GROUP (00992.HK)$At the performance meeting, they first proposed their “five elements” definition of AI PC, which sparked lively discussions within and outside the industry.

Following that, Damo also gave their unique opinion and definition of AI PCs. Damo believes that a real AI PC should have the following two core characteristics:

1) Equipped with a dedicated neural processing unit (NPU)

2) High processing speed (over 40 TOPS)

Furthermore, Damo also boldly predicted that the popularity of AI PCs would bring unprecedented benefits to the entire industry.

1. AI PC drives industrial growth

Damo predicts that as processors and memory are upgraded, the semiconductor value of each AI PC will increase by 20%-30%, and the average PC sales price will also increase by 7%.

The main difference between AI PCs and traditional PCs is the addition of a dedicated neural processing unit (NPU). Traditional PCs use CPU+GPU architectures, while AI PCs use CPU+GPU+NPU architectures.

In order for the NPU to be fully effective, the AI PC is also equipped with a larger amount of memory.$Microsoft (MSFT.US)$It was pointed out that the basic AI model requires 16GB of memory, the standard AI model requires 32GB, and the advanced AI model requires 64GB or more.

Based on this, Damo predicts:

• The value of semiconductors in each AI PC is expected to increase by 20-30% due to processor and DRAM upgrades, which will generate an additional $30 billion in revenue for global semiconductor companies;

• More advanced components such as NPU, more powerful CPUs, and more memory will drive up the average selling price (ASP) of a computer. Compared to general market consensus, the average selling price (ASP) of a computer will increase by about 7%.

2. AI PC architectures tend to be ARM

Damo pointed out that in the AI PC era, as processors shift from x86 to ARM architectures, it is estimated that by 2027, ARM-architecture PC CPU sales will reach 50 million units, bringing in 10 billion US dollars in revenue.

x86 and arm are the two mainstream processor instruction sets.

• x86 architecture command leader, good at expressing multiple tasks with a single complex instruction;

• Arm architecture has short instructions and is good at dividing complex tasks into multiple instructions for expression;

This difference in expression makes the ARM architecture characterized by high performance and low power consumption. Compared to x86, the ARM architecture is more suitable for high-power AI computing.

Currently,$Qualcomm (QCOM.US)$,$NVIDIA (NVDA.US)$,$Advanced Micro Devices (AMD.US)$and MediaTek have successively launched AI PC processor products based on the ARM architecture.

Damo pointed out that with the growth of the Windows on Arm (WoA) ecosystem, it is expected that by 2027, sales of WoA PC chips will reach 50 million units, bringing in $10 billion in PC CPU revenue.

3. Competition for AI PC processors

Damo said that the Snapdragon X Elite processor recently launched by Qualcomm in AI PCs is worth paying attention to.

Since Qualcomm acquired Nuvia, processor performance has improved dramatically. The Snapdragon X Elite, which has already been released, also uses the latest architecture. According to the Bank of America report, the Snapdragon X Elite performance is in a leading position among AI PC processor manufacturers.

In the first batch of AI PCs recently launched by Microsoft, Qualcomm Snapdragon X Elite chips were used to support the use of Copilot. These computers include Surface Laptop and Surface Pro.

Furthermore, according to data compiled by Damo, Lenovo, Acer, Asus,$Dell Technologies (DELL.US)$,$HP Inc (HPQ.US)$Other major computer manufacturers, such as Samsung, have also released their next generation artificial intelligence computers. These computers are also equipped with X Elite/X Plus chips (see image below).

In the past, most of Microsoft's processors used Intel CPUs. According to market data, in Q4 2023, Intel's share of the PC CPU market was 78%. This share may change with Qualcomm's strong entry into the market.

4. AI PC shipment forecast

2024 is regarded as the first year of AI PC shipments, and Damo has also predicted AI PC shipments.

Daima prediction:

• In 2024, AI PCs will account for 2% of total computer shipments;

• In 2028, AI PCs will account for 65% of total computer shipments;

This means that by 2028, shipments of artificial intelligence computers will reach 179 million units, 85% of which are laptops.

Thanks to the shipment of AI PCs, Damo points out that component suppliers, computer manufacturers and supply chain participants will benefit:

• Component vendors: Companies such as Nvidia, Qualcomm, Ultrafine Semiconductors, and Intel are expected to reap huge profits as key suppliers of NPUs and CPUs for artificial intelligence computers.

• Computer manufacturers: The report notes that Lenovo, Dell, Asus, and HP will be the main beneficiaries as they are well-positioned in the computer market.

• Supply chain participants: As demand for artificial intelligence computer components continues to grow, companies such as SK Hynix (memory), MediaTek (chipset), Delta Electric, and Advantech are expected to benefit from this. Leading chip foundry$Taiwan Semiconductor (TSM.US)$It is also expected to benefit from increased CPU orders.

Editor/Somer