SoftBank is making a significant push into the field of generative artificial intelligence. According to Nikkei, the company plans an ambitious $960 million investment aimed at enhancing its computing capabilities. By 2025, the Japanese telecom giant plans to develop a world-class Japanese-language generative AI, signaling its intent to compete globally in this rapidly advancing sector.
The company will purchase advanced graphics processing units (GPUs) from Nvidia, not only to power its AI development but also to provide computational resources to other companies. This move follows an initial $160 million investment last year in computing infrastructure, underscoring SoftBank’s commitment to becoming a leader in AI technology.
SoftBank’s ongoing project includes the development of a large language model (LLM) with 390 billion parameters, scheduled for completion in fiscal 2024. Subsequently, the company plans to escalate its ambitions by developing a model with 1 trillion parameters, aligning with global benchmarks for AI performance.
The strategic investment is timely as the Japanese generative AI market is expected to expand significantly, from $760 million in 2023 to approximately $13 billion by 2030. This growth trajectory reflects a broader global trend where AI technologies are being integrated across diverse sectors.
SoftBank’s initiative also reflects a growing emphasis on data sovereignty in Japan, mirroring global concerns about relying too heavily on foreign technology which could pose risks to national security and economic stability.
Moreover, the surge in SoftBank’s share price to its highest since its 2018 listing, following the announcement, reflects investor confidence in the profitability and strategic importance of generative AI technologies. This investment not only positions SoftBank as a pivotal player in Japan’s tech landscape but also as a significant contributor to the global AI innovation ecosystem.
