NVIDIA to invest 4 trillion won in Meta Platforms, expand collaboration on AI assistant ‘Muse’ [Global AI Briefing]

[Financial News] NVIDIA Corporation will invest $3 billion (about 4 trillion won) in Meta Platforms, expanding its collaboration on the development and operation of AI models, including the AI assistant ‘Muse.’
The strategy goes beyond supplying AI semiconductors to help strengthen customers’ competitiveness in AI services as well. The move is drawing attention as the boundary between companies that develop AI models and semiconductor companies that power them grows increasingly blurred.


NVIDIA Corporation officially announced on the 27th (local time) that it would invest $3 billion in Meta Platforms’ paid-in capital increase, worth a total of $97.5 billion (about 130 trillion won), to fund Meta’s AI investments.
Alongside discussions on the investment, NVIDIA Corporation and Meta Platforms agreed to expand their collaboration to improve the training and inference performance of Meta’s AI assistant Muse.
Muse is Meta Platforms’ AI assistant service. At users’ direction, it uses various tools and services to carry out tasks such as shopping, sending emails, and making reservations.
For Muse to gain wider adoption, it will be crucial to provide useful features while increasing response speeds and reducing operating costs. NVIDIA Corporation’s high-performance semiconductors and networking technology could help with Muse’s training and service performance, which is seen as the reason the two companies are deepening their collaboration.
In effect, the partnership will allow Meta Platforms to secure the infrastructure needed to develop AI models and operate its services, while NVIDIA Corporation expands the AI ecosystem that uses its technology.
Industry observers say the collaboration between NVIDIA Corporation and Meta Platforms demonstrates a trend in which competition in the AI market is shifting toward operational efficiency and cost.
As competition in the AI market rapidly shifts toward agents, experts expect that for AI assistants to be put to practical use in daily life and at work, competitiveness will depend not only on high-performing models but also on computing infrastructure capable of supporting them and efficient service strategies.
[email protected] Lee Gu-sun Reporter