The original intention of laying out the new track
Now, with the further deepening of the understanding of AI in all walks of life, a more pragmatic perception has emerged in the market-Rome was not built in a day.
The most important thing in the era of artificial intelligence is the collection and acquisition of various data, because artificial intelligence needs a large amount of data to “feed” so that they can become smarter.
But quantifying data availability as a digitization rate is also a major bottleneck for AI.
For AI to become more ubiquitous, its task-processing capabilities must become more efficient, which means we should focus on developing AI-enhanced hardware—this does not mean going back to using discrete AI or neural networks. The number of ways of the chip.
On the contrary, in the future, artificial intelligence solutions will tend to be more balanced and integrated – as we can see, in almost all types of applications, there is artificial intelligence involved, such as computer-aided design and customer service, as well as retail and communication, etc. wait.
The investment logic of card bit AI
Storage manufacturers started the AI investment boom around 2018, which is also a carnival year for artificial intelligence. Micron held the micron insight 2018 conference, inviting Amazon, BMW, Google, Qualcomm, Microsoft, nvidia and Visteon, as well as MIT professors, writers and explorers to gather together to discuss emerging technologies such as AI, machine learning and deep learning. technology.
At this conference, Micron announced the establishment of a $100 million investment fund to invest in startups in the AI field, 20% of which will be invested in companies led by women and disadvantaged groups.
In 2019, Micron Technology directly acquired Fwdnxt, an artificial intelligence hardware and software startup, which caused quite a stir in the industry at the time.
Micron believes that Fwdnxt’s technology, when combined with Micron’s memory chips, will equip Micron with the ability to explore deep learning AI solutions needed for data analytics, especially for the Internet of Things and edge computing.
SK hynix AI investment battle
SK Hynix has also joined this AI investment war, even with a fierce offensive. At the beginning of 2019, SK hynix invested in Horizon; in September 2020, it announced its investment in Gauss Labs, aiming to lead innovation in the semiconductor manufacturing industry through industrial artificial intelligence solutions.
In January 2022, SK Hynix, together with SK Telecom and SK Square, announced the establishment of a joint development association. First, it will spend 80 billion won to establish an AI semiconductor company SAPEON in the United States. SK Hynix holds 25% of the shares. For SK hynix, the business in the fields of NAND flash memory and AI will further expand.
New growth areas targeted by the Samsung Group
Samsung Group announced in August 2018 that it will invest more than US$22 billion in AI, auto parts and other fields in the next three years, most of which will be borne by Samsung Electronics; in August 2021, Samsung announced again that it will invest US$2.4 million in the next three years. billion won (about 205.5 billion U.S. dollars) to fields such as biopharmaceuticals, artificial intelligence, semiconductors, and robotics.
From US$22 billion in 2018 to US$205.5 billion in 2021, a tenfold increase. Although AI is not the only area where Samsung Electronics has increased investment, it is clearly a new growth area that has been locked.
Kioxia joins the research and development of artificial intelligence technology
Recently, Kioxia has developed an image classification system based on Memory-Centric AI. Memory-Centric AI is an artificial intelligence technology that utilizes large-capacity memory. The system uses neural networks to classify images. In the future, Kioxia will use Memory -Centric AI expands from image classification to other fields, and promotes research and development of artificial intelligence technology using large-capacity storage.
Kioxia pointed out that traditional AI technology uses a large amount of data to train the model, while memory-centric AI is based on its accumulated knowledge, and performs tasks by searching and referring to this knowledge, so the more new knowledge and memory it accumulates The more, the more mature the artificial intelligence will be.
end:
At present, AI technology has been implemented in many fields such as finance, medical care, and security, and the application scenarios are becoming more and more abundant. It is realizing all-round commercialization and triggering profound changes in various industries. This will accelerate the digitization of enterprises and improve the industry. It has played a positive role in improving the chain structure and improving the efficiency of information utilization.
At the same time, AI has also fully entered the era of machine learning, and the future development of AI will be the combination of key technologies and industries.
Part of the content comes from: Semiconductor Industry Observation: Memory Chip Giants Flock to New Tracks; Financing Business Plan: Artificial Intelligence Track Card Positions, let’s see how the capital market operates? ; Samsung’s AI development; Micron Micron Technology: Micron Technology talks about the primary development path of artificial intelligence