#光伏# #人工智能# Not a few weeks ago, artificial intelligence chip giant NVIDIA CEO Jensen Huang pointed out that “the future development of AI is closely linked with photovoltaic and energy storage. We can not only think about the arithmetic power, if we only consider the computers, we need to burn down 14 of the earth's energy” in a public speech, whose expression triggered heated debate among Photovoltaic industry. I have no way to verify the source of the original text, but quietly think about "why artificial intelligence will be coupled with energy", that is still worthwhile.
Electricity formation and applications
With the social and economic development, leads to scientific experiments progress. in 1831, the British physicist Michael Faraday discovered the theory of electromagnetic, meanwhile get the method of generating alternating current, and for the first time to create the first generator. Therefore this lays the foundation for realization for the electrical era of the second industrial revolution. American famous inventor and physicist Thomas Edison invented the phonograph, movie camera, tungsten filament lamps and other electrical industrial products came into being, promoting the human society to enter into the secondary industrial revolution electrical era, in other words relying on primary non-renewable fossil energy development changes to electricity. From that human production and life have been closely related to the use of electricity but only for a short period of 100 years.
Fossil energy share over 60%
It has been less than 300 years since the 1860s (the first industrial revolution) that human society has attempted to improve the efficiency of energy utilization as a means of advancing society's scientific and technological development. During this period, because of humans have taken fossil energy from the earth to obtain energy power, result in emitted greenhouse gases carbon dioxide that far exceeded the cumulative value of all human activity on the planet before the Industrial Revolution.
As of 2022Y statistics show that global electricity from fossil energy, sources of coal accounted for 35.63%, natural gas accounted for 22.48%, oil accounted for 3.15%, the three together accounted for 61.26%, is the main source of electricity. Next is hydropower accounted for 14.96%, followed by nuclear power 9.18%, wind power 7.32%, solar energy 4.57%. 2022Y wind power and solar energy annual power generation and consumption accounted for 11.89%, the two annual installed capacity also exceeded coal power for the first time. And it is expected to start from 2023Y, the global additional demand of electricity all met by the wind power and solar energy. This is also a milestone year, the driving force of human social activity is shifting from fossil energy to renewable energy.
AI is major electricity consumer
Energy supply or power consumption is always a power source for the development of human society. Back to NVIDIA CEO Jensen Huang expressed the content, we found that the core of its expression should be understood as "artificial intelligence needs more power consumption, and its source needs more forms of energy support, including photovoltaic and energy storage". Its photovoltaic intermittent power generation characteristics, combined with the form of energy storage to overcome the problem of intermittent power output, the views put forward are also quite targeted.
So why does AI need to consume more electrical energy? AI algorithms require a lot of computing resources to run and train, especially in areas such as deep learning and neural networks. These computational resources usually come from large data centers or cloud computing platforms, which require large amounts of power to keep them running. Here it is also good to understand why Microsoft will think of sinking servers to the bottom of the sea, Huawei thought of building data centers in high-altitude caves, the motivation behind these business practices is to think of the use of natural water or wind to reduce the temperature of the data center, in order to reduce power consumption and costs. Another example is the current use of liquid cooling technology for energy storage products, look at the same thing.
With the increasing popularization and application of AI technology, more and more smart devices and systems also require electrical energy supply. For example, smart homes, self-driving cars, smart factories and other fields require a large amount of electrical energy support to drive their operation. The coupling relationship between artificial intelligence and electrical energy is inevitable. In order to better develop AI technology, it is necessary to consider how to effectively utilize and manage electric energy resources to ensure the reliable operation and sustainable development of AI systems. The development of human society to change the way of electrical energy and energy sources is itself a social progress, isn't it?
The massive use of Artificial Intelligence or AI technology to promote the transformation of human society's production and lifestyle will inevitably derive a greater demand for electricity, and at the same time continue to change and optimize the scheduling and consumption in the power system. Only by increasing the power supply from renewable energy sources, reducing fossil energy dependence and consumption, and lowering greenhouse gas carbon dioxide emissions, can we achieve a harmonious coexistence of human beings and the earth.