According to the Zhitong Finance and Economics APP, with the rapid development of generative artificial intelligence technology, the demand for computing power is experiencing explosive growth, leading to a massive increase in electricity consumption. Goldman Sachs' latest view believes that the electricity required for one query by ChatGPT is almost 10 times that of Google search. This difference in electricity consumption will bring about significant changes in the way electricity is consumed and the costs in the United States, Europe, and even globally.
Recently, the nuclear power plant invested in by Bill Gates with 1 billion USD broke ground in Wyoming, USA. On the 16th local time, he mentioned that he will continue to invest: &#34;I have invested over a billion, and I will invest tens of billions more.&#34;
The reason Bill Gates continues to invest in this nuclear power plant is largely due to the enormous electricity pressure brought about by the development of artificial intelligence (AI). &#34;The data centers we are going to build will increase the electricity load by as much as 10%.&#34; He stated that the rise of electric cars and various household heating devices such as heat pumps has increased the electricity demand in the United States, and the emergence of data centers is &#34;adding insult to injury&#34;. Therefore, large tech companies are researching how to contribute more to the power supply to enable these data centers to meet the explosive growth in AI demand.
Furthermore, the surge in demand for next-generation technologies has raised concerns about potential power shortages, causing the deceleration of coal retirement progress in the U.S. Some U.S. companies have postponed the retirement deadlines for coal-fired power plants. Alliant Energy recently postponed its plan to convert its Wisconsin coal-fired power plant to natural gas from 2025 to 2028. FirstEnergy has abandoned its goal of phasing out coal by 2030 due to &#34;resource adequacy issues&#34;.
According to IEA data, using the chatbot ChatGPT under OpenAI for one query consumes 2.9 watt-hours of electricity, while Google search only requires 0.3 watt-hours, about 1/10 of the former. By 2026, electricity consumption related to data centers, cryptocurrencies, and AI could increase to 620 to 1050 terawatt-hours. In 2023, Germany's total electricity consumption was 465 terawatt-hours, and Japan's total power demand was 870 terawatt-hours.
For the domestic market, Huatai Securities pointed out that while data center construction is advancing overseas, there is also high growth potential domestically under the drive of AI implementation. The overseas growth in electricity demand provides global growth opportunities for excellent domestic power generation and grid equipment companies, and the upward cycle of the global electricity-related industry chain is still strengthening.
It is worth mentioning that with the arrival of high temperatures in summer, China's electricity consumption is entering a peak period. According to the National Energy Administration's forecast, during this year's summer, the national electricity load will grow rapidly, with the peak load increasing by over 100 million kilowatts year-on-year, exceeding 7% compared to the peak load of 1.339 billion kilowatts set in July last year. Under the catalysis of high electricity demand due to high temperatures and electricity shortages in some regions, investment opportunities in the power sector are worth paying close attention to.
Guojin Securities research report indicates four structural investment themes in the power equipment sector:
<ol>
<li>
Overseas Expansion of Power Equipment: The export prosperity of China's power equipment is on the rise, with a significant mismatch between overseas supply and demand. Subsequent catalysts include overseas grid investment planning + customs exports exceeding expectations + individual stocks exceeding expectations.
</li>
<li>
Power Reform: Policies will drive the industry into the second growth curve.
</li>
<li>
Ultra-High Voltage Lines &amp; Main Grid: The strong demand for the second batch of large-scale wind and solar bases, coupled with multiple line gaps, will catalyze subsequent disclosures of new lines + increased penetration rates + individual stocks exceeding expectations.
</li>
<li>
Distribution Network Transformation: Top-level policies + year-on-year high growth in some provincial tenders, with subsequent catalysts including policy changes + tenders exceeding expectations + individual stock breakthroughs.
</li>
</ol>
Huaxi Securities predicts that the demand for new energy IT will experience a rapid outbreak, entering a golden development stage. This year, an unprecedented large-scale equipment update is brewing in the power industry, signaling a new opportunity period for the sector.
Related Concept Stocks:
China Resources Power (00836): Recently, China Resources Power announced that in May 2024, the sales volume of its subsidiary power plants reached 1602.47 million megawatt-hours, an increase of 6.7% year-on-year. Among them, the sales volume of its wind power plants reached 360.58 million megawatt-hours, a decrease of 2.7% year-on-year; the sales volume of its photovoltaic power stations reached 65.49 million megawatt-hours, an increase of 276.8% year-on-year. In the first five months of 2024, the cumulative sales volume of its subsidiary power plants reached 8138.68 million megawatt-hours, an increase of 5.7% year-on-year.
China Datang Corporation Renewable Power (01798): In mid-June, CICC released a research report stating that for China Datang Corporation Renewable Power to achieve breakthrough profits and have significant valuation upside, adjustments due to power reform will take time to reflect. The target price was raised by 10.5% from HKD 2 to HKD 2.21. Due to limited upside potential, the rating was downgraded from &#34;Buy&#34; to &#34;Neutral&#34;.
China Nuclear Power (01816): At the end of April, CICC released a research report stating that China Nuclear Power has risen by 29% since the beginning of the year, with the dividend yield of H shares narrowing to 4.4%. The certainty of earnings seems insufficient to explain its valuation premium relative to other Hong Kong-listed power peers. Based on the potential for China Nuclear Power to double its installed capacity in the next decade, the target price was raised to HKD 3.2, with a reiterated &#34;Buy&#34; rating

06886

601688

OpenAI

HTSC believes that the rapid development of generative artificial intelligence technology will bring a massive demand for electricity, with ChatGPT consuming almost 10 times the electricity of a single Google search. In response, Bill Gates is investing $1 billion to build a nuclear power plant to meet the power needs of data centers. Additionally, as the demand for next-generation technologies surges, the decarbonization progress in the U.S. power industry is slowing down. It is estimated that by 2026, the electricity consumption related to data centers, cryptocurrencies, and AI could increase to 620 to 1050 terawatt-hours

Zhitong

HTSC

Hong Kong Stock Concept Tracking | ChatGPT's electricity consumption for one query is 10 times that of Google search. The development of AI computing power has spurred a massive demand for electricity (with concept stocks attached)