VESSL AI, an artificial intelligence cloud company, has teamed up with Korea Investment & Securities to secure a funding base for expanding its graphics processing unit infrastructure. The plan is to combine a securities firm's fundraising capabilities with a GPU infrastructure business that requires heavy upfront investment, and use that to grow its neocloud operations.
VESSL AI said on the 16th that it had signed a memorandum of understanding on GPU financing with Korea Investment & Securities. Under the agreement, Korea Investment & Securities will cooperate on expanding investment and raising funds for GPU infrastructure, while VESSL AI will support the projects based on its experience operating a GPU cloud platform. The specific investment size and financial structure will be determined through consultations on each project.
Neocloud refers to cloud services that specialize in providing the GPU computing resources needed to develop and run AI. Meeting rising demand for AI computing requires large sums to secure GPUs and build out infrastructure, making stable funding a key element of any expansion. VESSL AI plans to use the partnership to build capacity for GPU infrastructure investment and widen its supply of cloud services.
The company has been expanding its GPU infrastructure business in rapid succession. A prime example is a GPU cluster based on Nvidia's next-generation AI chip, Vera Rubin, which it plans to open in the first half of next year. For that project, VESSL AI is also pursuing 400 billion won (about $290 million) in GPU financing together with Nvidia and global investment banks in the United States and Singapore. If it proceeds as planned, it would be the first case of GPU financing carried out in South Korea. VESSL AI is targeting 200,000 GPUs by 2030.
"Competitiveness in GPU infrastructure comes from having both a funding base and operational capability," VESSL AI Chief Executive Ahn Jae-man said. "Through this partnership with Korea Investment & Securities, we will respond more quickly to demand from global AI customers."