- Sakura Internet, bolstered by government subsidies, is expanding its cloud services for generative AI, with a reported fivefold increase in GPU procurement, including NVIDIA’s B200.
- The company aims to deploy approximately 10,000 GPUs, featuring NVIDIA’s latest offering, to achieve a computational power of 18.9 EFLOPS by March 2028.
- Previous government subsidies in June 2023 and the current subsidy of JPY 50.1 billion highlight Sakura’s strategic positioning in Japan’s generative AI landscape.
- Sakura’s revised investment plan, totaling around JPY 100 billion, seeks to acquire an additional 8,000 GPUs between April 2024 and December 2027.
- NVIDIA’s commitment to support Japan’s generative AI aspirations, as evidenced by CEO Jensen Huang’s discussions with Prime Minister Fumio Kishida, underscores the collaborative efforts to propel generative AI development.
Main AI News:
Sakura Internet, a leading Japanese digital infrastructure service provider, fueled by government backing, is amplifying its cloud capabilities tailored for generative AI. As per reports from MoneyDJ, Sakura Internet is poised to quintuple its GPU procurement from initial projections, with acquisitions encompassing NVIDIA’s cutting-edge offering, the “B200,” unveiled in March.
In an announcement dated April 19th, Sakura revealed its successful securing of Japanese government subsidies to fortify its cloud service “Koukaryoku” for generative AI. The company charts a trajectory to escalate the GPU count in “Koukaryoku” fivefold from the initially outlined figure, with aspirations to integrate approximately 10,000 GPUs, inclusive of NVIDIA’s latest “NVIDIA HGX B200 system” introduced recently. The ultimate objective? To erect a vast cloud infrastructure boasting a computational prowess of 18.9 EFLOPS by March 2028’s denouement.
Notably, this marks Sakura’s second bout of government subsidies, following a similar instance in June 2023.
Elaborating on previous endeavors, last June, Sakura channeled JPY 13 billion into procuring around 2,000 NVIDIA GPUs (packing a computational might of 2.0 EFLOPS) slated for acquisition between July 2023 and March 2025. Owing to an exponential surge in demand, the acquisition of these 2,000 GPUs is now slated for completion before June 2024 draws to a close, ahead of schedule.
Under this new investment blueprint, which commands an approximate JPY 100 billion allocation (inclusive of server component expenses besides GPUs and maintenance costs), Sakura aims to secure an additional 8,000 GPUs (bolstered with a computational capability of 16.9 EFLOPS) between April 2024 and December 2027.
Consequently, the overall GPU procurement tally of nearly 10,000 units will outstrip the original plan by a factor of five, from approximately 2,000 units initially envisaged. According to insights from Japanese media outlets, Sakura will furnish server computational prowess, leveraging these GPUs, to enterprises vested in generative AI research.
On April 19, the Ministry of Economy, Trade and Industry of Japan divulged plans to extend a maximum subsidy of JPY 72.5 billion to five Japanese entities, aiming to forge requisite supercomputing infrastructure for domestic generative AI advancement. Of this, Sakura stands to gain the lion’s share with a maximum subsidy pegged at JPY 50.1 billion.
Noteworthy is NVIDIA CEO Jensen Huang’s visit to Japan in December of the previous year, where he conferred with Japanese Prime Minister Fumio Kishida. During the rendezvous, Prime Minister Kishida pressed for NVIDIA’s robust GPU supply to underpin Japan’s generative AI endeavors. NVIDIA commits to synergizing efforts with Japanese counterparts, including Sakura, SoftBank, NEC, NTT, and others, to fast-track generative AI evolution.
Conclusion:
Sakura Internet’s substantial GPU procurement surge, backed by government support and strategic collaborations with industry giants like NVIDIA, signifies a significant advancement in Japan’s generative AI capabilities. This surge not only enhances Sakura’s competitive edge but also underlines Japan’s commitment to fostering innovation in the AI sector, potentially reshaping the market dynamics and driving technological innovation across various industries.
