Image source: Photo.com
Recently, Israel's Ministry of Finance, Ministry of Economy and Ministry of Taxation issued a joint statement announcing the approval of Intel Corporation of the United States to invest $25 billion in a new chip factory in the country. In the U.S. stock market, Nvidia shares rose 4.5%.
The statement said it was the largest investment by a foreign company in the country's history, and the factory would be built in the southern town of Ghat and would employ thousands of workers. Intel has pledged to complete the investment and open the plant by 2028 and operate it until at least 2035. Under Israel's laws to encourage investment, Intel will receive a grant of $3.2 billion for 12.8 percent of the investment.
——AI chips are the mainstream mode of intelligent computing
Accelerated computing based on AI chips is the mainstream mode of intelligent computing. AI chips are co-designed with AI algorithms to meet the ultra-high demand for computing power in AI computing. At present, mainstream AI accelerated computing mainly uses CPU systems equipped with heterogeneous acceleration chips such as GPUs, FPGAs, and ASICs.
In recent years, domestic AI acceleration chip manufacturers have continued to make efforts and made rapid progress in this field, and related products have been released one after another, covering the needs of AI inference and AI training, including chips based on general-purpose GPU architecture, chips based on ASIC architecture, and brain-like architecture chips, showing a diversified development trend in general.
——Analysis of the global artificial intelligence chip market size
Artificial intelligence chips are the brains of artificial intelligence, with the growth of the number of global artificial intelligence terminal devices and the gradual increase in the demand for edge computing, the global demand for artificial intelligence chips is growing rapidly, and the market scale is expanding. According to data released by Tractica, the global artificial intelligence chip market reached $11 billion in 2019, and the global artificial intelligence chip market is expected to increase to $17.5 billion in 2020, and the global artificial intelligence chip market is expected to exceed $72 billion in 2025.
-- AI chip competition pattern
From the perspective of application field classification, NVIDIA is the largest in the global cloud training chip market, and NVIDIA's GPU+CUDA computing platform is the most mature AI training solution. In addition, there are two solutions: the third-party heterogeneous computing platform OpenCL+AMD GPU and the self-developed acceleration chip of cloud computing service providers.
In terms of the global cloud inference chip competition pattern, hundreds of schools of thought compete for cloud inference chips, each with its own merits. Compared with the training chip, the inference chip considers more comprehensive factors: computing power per unit power consumption, latency, cost, etc. The initial inference is also accelerated by GPUs, but due to the particularity of the application scenario, the optimization according to the specific neural network algorithm will bring higher efficiency, and the performance of FPGA/ASIC may be more prominent. In addition to traditional chip manufacturers such as Nvidia, Google, Xilinx, and Altera (Intel), startups such as Wave computing and Groq have also joined the competition. Among Chinese companies, Cambrian and Bitmain are also actively deploying cloud chip business.
IDC pointed out that due to the weak demand in the chip market this year, the supply chain inventory consumption process is still continuing, although there were sporadic short orders and urgent orders in the second half of this year, it is still difficult to reverse the 20% year-on-year decline in the first half of this year. As a result, the global semiconductor sales market is still expected to decline by 12% in 2023. With the gradual recovery of demand for smartphones and the strong demand for AI chips, IDC expects the semiconductor market to return to its growth trend in 2024, with an annual growth rate of more than 20%.
Galen Zeng, senior research manager of semiconductor research in IDC Asia Pacific, said that the tight control of supply and production by memory chip manufacturers has led to chip prices having risen since the beginning of November this year. The demand for artificial intelligence will drive the recovery of the overall semiconductor sales market in 2024. The semiconductor supply chain, including design, manufacturing, packaging, and testing, will bid farewell to the downturn in 2023.
Prospective Economist APP Information Group
For more research and analysis of this industry, please refer to the "Chinese Intelligent Chip (AI Chip) Industry Development Prospect Forecast and Investment Strategic Planning Analysis Report" of the Prospective Industry Research Institute.
At the same time, the Prospective Industry Research Institute also provides solutions such as industrial big data, industrial research reports, industrial planning, park planning, industrial investment, industrial mapping, smart investment promotion system, industry status certification, IPO consulting/fundraising feasibility study, IPO working paper consulting, etc. To quote the content of this article in any public information disclosure such as prospectus and annual report, formal authorization from Qianzhan Industry Research Institute is required.
More in-depth industry analysis is available in the [Prospective Economist APP], and you can also communicate and interact with 500+ economists/senior industry researchers.