Source: Content compiled from eenewseurope

The World Semiconductor Trade Statistics (WSTS) said that although the global chip market is growing, the European chip market is expected to shrink in 2024.

WSTS raised its forecast for the global chip market in 2024, helped by growth in the Americas and Asia-Pacific, but at the expense of slower growth in 2025.

However, while it was able to raise its 2024 growth forecasts for the Americas and Asia-Pacific from five months ago to 38.9% and 17.5%, respectively, it forecast a 6.7% decline in annual growth for Europe.

WSTS raised its forecast for the global chip market in 2024 to $626.87 billion in its fall forecast, up 19.0% from the 2023 total. WSTS had expected a 16.0% increase in its spring forecast.

WSTS now forecasts that the global chip market will reach $697.18 billion in 2025, an annual growth rate of 11.2%, down from the previous forecast of 12.3%. This still makes WSTS one of the more optimistic forecasters among forecasters. Other forecasts range from 6% to 14% for 2025.

The upward revision to the short-term global chip market data reflects improved performance in Q2 2024 and Q3 2024, especially in the computing sector, driven by demand for AI-enabled chips.

As a result, growth in 2024 will be driven by memory (expected to grow 81.0%) and logic (expected to grow 16.9%). The growth in memory is mainly due to rising average selling prices.

Europe has almost no computer manufacturing industry, and its development is mainly driven by the industrial and automotive sectors, both of which have been hit by low demand and excess inventory.

As a result, categories such as discrete devices, optoelectronics, sensors, and analog semiconductors are expected to decline in 2024.

The forecast for the following year reflects a more traditional pattern, with torrid growth in memory being tempered but discretes, optoelectronics, sensors and analog returning to single-digit percentage growth.

Gartner predicts that global semiconductor revenue will grow 14% in 2025

According to the latest forecast from Gartner, global semiconductor revenue is expected to grow 14% in 2025 to $717 billion. The market size is expected to grow 19% in 2024 to $630 billion.

After a decline in 2023, semiconductor revenue is rebounding and is expected to achieve double-digit growth in 2024 and 2025 (see Table 1). "Growth is driven by a continued surge in demand for semiconductors related to artificial intelligence and a recovery in electronics production, while demand in the automotive and industrial sectors continues to be weak," said Rajeev Rajput, senior principal analyst at Gartner.

Table 1

In the short term, the memory market and graphics processing units (GPUs) will drive global semiconductor revenue growth.

The global memory market revenue is expected to grow 20.5% to $196.3 billion in 2025. The continued supply shortage in 2024 will drive NAND prices up 60% in 2024, but prices will fall 3% in 2025. Due to the reduced supply and weak pricing environment in 2025, NAND flash memory revenue is expected to reach $75.5 billion in 2025, an increase of 12% from 2024.

DRAM supply and demand will rebound due to improved supply shortages, unprecedented high bandwidth memory (HBM) production and demand rise, and double data rate 5 (DDR5) price increases. Overall, total DRAM revenue is expected to grow to $115.6 billion in 2025 from $90.1 billion in 2024.

GPUs have dominated the training and development of AI models since 2023. Its revenue is expected to reach $51 billion by 2025, up 27%. "However, the market is now moving to the return on investment (ROI) phase, and inference revenue needs to grow to a multiple of training investment," said George Brocklehurst, vice president analyst at Gartner.

This includes a sharp increase in demand for HBM, a high-performance AI server memory solution. "Vendors are investing heavily in HBM production and packaging to meet next-generation GPU/AI accelerator memory needs," said Brocklehurst.

HBM revenue is expected to grow by more than 284% in 2024 and 70% in 2025, reaching $12.3 billion and $21 billion, respectively. Gartner analysts predict that by 2026, more than 40% of HBM chips will be used for AI inference workloads, compared with less than 30% today. This is mainly due to increased inference deployments and limited repurposing of training GPUs.

View more at EASELINK