Gartner Reports That The Market For AI Chips Is Expected To Grow At More Than 10% Cagr In The Next Five Years And Remain Robust At Least Through 2028.
Since 2023, the global revenue from AI semiconductors is expected to reach $71 billion, a YoY growth of 33%, as outlined by Gartner’s latest forecast.
According to the research, the global revenues from AI chip sales are anticipated to reach $91 by the year 2025. The report released today states that the revenue will be 5 billion and will grow in the double-digit to at least 2028.
By the end of the year 2026, 100 percent of PC purchases in enterprises will be AI PCs which are computers which contain an NPU that allows for on-computer Ai processing. These PCs operate for longer and quieter periods are cooler and always in the background, and have AI operations through Gartner’s report. The firm estimates that AI PC shipments will account for 22% of total PC shipments by 2024.
For this year alone, more than 40% of the revenue for AI chips is projected to come from sales of PCs with integrated AI capabilities. In the computer electronics industry alone, the revenue from AI chips is estimated to reach $33 by the end of this year. This is expected to nearly double to $4 billion in the next five years and will make up 47% of total AI semiconductors revenue, Gartner says.
“Today, GenAI is rapidly driving the adoption of high-performance AI chips in data centers. In 2024, the market for AI accelerators applied in servers is expected to be worth $21 billion, and this is expected to rise to $33 billion by 2028,” said Alan Priestley, vice president analyst at Gartner.
This year, the revenue from AI chips in automotive electronics is also anticipated to be $7. 1 billion, and $1. Consumer electronics accounted for 8 billion dollars.
IDC research reveals that sixty-six percent of enterprises globally declared that they would be planning to invest in GenAI in the next eighteen months. Of the companies that revealed their intention to raise IT budgets on GenAI in the year 2024, 46% of the overall budget will be on infrastructure.
The problem: the one crucial element that is required to implement this AI infrastructure is currently scarce: hardware. Although GPUs have become indispensable in their role to train the biggest large language models (LLMs) that power GenAI, the market continues to need high-speed memory chips for AI applications. Currently, the market is still rather limited for both of them.
This is because the training and inference of LLMs on GPUs is computationally intensive and expensive to perform. Smaller and more targeted models, more focused on industries or businesses are usually more effective in providing better results to businesses, and they can work with standard x86 processors with NPUs.
”New ai workloads are being put onto high performance GPUs and the major cloud providers (AWS, Google, Meta, Microsoft) are all dipping their toes into the pool to develop their own silicon for new AI” said Priestley.
Although it is expensive to develop chips, the use of custom-designed chips to implement AI-based services can significantly enhance efficiency, lower the costs of providing AI services to the end-users and decrease the cost that users have to incur to access new AI applications as explained by Priestley.
‘This is because as the market moves from the development stage to the deployment stage we anticipate this trend to remain constant,’ added Priestley.
In a previous interview with Reuters last month, Intel CEO Pat Gelsinger shared that he envisions Intel’s future as part of an AI-for-everything strategy, with NPUs forming the backbone of its new line of Intel Core Ultra processors. The chipmaker will be shipping 40 million AI PC processors in 2024 and a whopping 100 million the following year.
Some of the factors contributing to the growth in the usage of AI on edge devices include that the current lifetime expectancy of mobile phones is relatively low; that people and corporations switch their phones more often.
“This change enabled the device spending to get to $688 billion in the year 2024 from $664 billion that was spent in the year 2023 which was the low spending level and this will be a growth of 3. 6% as stated in the report. This change is thus fueled by the integration of GenAI features in premium as well as basic phones.
0 Comments