The "computing power arms race" of autonomous driving was launched, and domestic AI chips counterattacked the global hegemon
Some time ago, a few years ago, the rapid development of unmanned driving, in this process, AI has played a great role in promoting. Artificial intelligence plays a key role in the field of autonomous driving. Not long ago, Tesla released the FSD12Beta version, which demonstrated that only one manual operation can achieve fully artificial intelligence terminal-to-terminal driving control, which further strengthens the prospect that artificial intelligence computing power will affect unmanned driving.
1. The battle between unmanned driving and artificial intelligence computing power.
With the development of autonomous driving technology, the demand for AI computing power has also increased dramatically. Behind this need, there are two main lines: GPGPU and ASIC. The modeling process of driverless cars involves massive matrix operations, and the GPU architecture is good at solving such problems, so GPGPU can be better applied to various artificial intelligence application environments and has higher universality.
2. The rise and problems faced by domestic artificial intelligence chips.
At present, China's investment in GPGPU is still relatively small, but Haiguang Information, Biqian Technology, Mu Xi and others have begun to conduct in-depth research in this area. Haiguang Deep Computing No. 1 is currently the only artificial intelligence chip with high-precision computing power in China, its computing power is not worse than NVIDIA's NVIDIA's processor, and it can also be combined with NVIDIA's CUDA operating system, which is its biggest advantage.
In contrast to GPGPU, an ASIC is an application-specific integrated circuit. For example, Cambrian, Huawei, and some other domestic companies are mainly ASICs. Although ASIC chips must be designed to take into account the adaptability of various algorithms, and its ecological environment is relatively fragmented, making it difficult for developers to master, its performance can still adapt to most application environments.
Third, overtaking in corners, domestic artificial intelligence technology is facing huge opportunities and challenges.
In the battle of computing power in this round of unmanned driving, it is entirely possible for domestic AI chips to catch up with those large foreign companies. For example, Haiguang Semiconductor's "Deep Computing No. 1" has achieved hundreds of thousands of commercial applications, and its service objects cover intelligent computing centers, industrial users, artificial intelligence and network companies of the "new infrastructure" project. It can be seen that domestic artificial intelligence chips have been greatly improved in terms of performance and actual use. However, how to realize its industrialization has always been a huge problem in the development of AI chips in China. Although the R&D strength of domestic AI chips is not much worse than that in the world, we still need to speed up the pace of ecological construction and commercial development of artificial intelligence.
In the "computing power war" of unmanned vehicles, how will Chinese artificial intelligent chips stand out and finally complete the "anti-killing" and dominate the whole world?This is a question that we have all been waiting for for a long time. This requires us to continuously improve the performance of our processors to keep up with the increasing computing power;At the same time, it is necessary to increase the research and development of the artificial intelligence industry and promote the commercial development of the artificial intelligence industry, so as to promote China's artificial intelligence chips to the international market.
But in any case, the future of domestic artificial intelligence chips is bright. Let us look forward to the fact that in the future unmanned "computing power war", domestic artificial intelligence chips can stand out and complete the real counterattack and catch-up!This is not only a battle of technical strength, but also a battle of wisdom. The era of unmanned driving has come, and you are fully prepared for domestic artificial intelligence chips?