VESSL AI Targets 200,000 GPUs by 2030 in Neocloud Push
VESSL AI announced a roadmap on October 15, 2026 to secure 8 AI data center bases by next year and 200,000 GPUs by 2030.
VESSL AI unveiled a mid-to-long-term roadmap on October 15, 2026 to become a global top-tier neocloud company, according to datanet.co.kr. The plan includes securing 8 AI data center (AIDC) bases across South Korea by next year and 200,000 GPUs by 2030.
The company targets 10,000 GPUs this year and 50,000 GPUs by 2027. It expects 250 billion won in revenue this year and aims for 2 trillion won in revenue by 2028, while pursuing 400 billion won in global GPU financing.
VESSL AI also plans to open an Nvidia DGX B200 NVL72 cluster in the first half of next year and introduced an inference service that supplies computing on a token-by-token basis.
Earlier growth and partnerships
In the first half of this year, VESSL AI recorded 130 billion won in revenue. Founded in 2020, the company initially grew on a platform supporting AI model development, training, deployment and operations, and it began its neocloud business in earnest last year. It has been providing global inference infrastructure for Upstage's LLM Solar Pro 4 and participates in the Nvidia Inception program. CEO Ahn Jae-man and CTO Jeon Ji-hwan led the roadmap announcement, which centers on Fluid Computing as a core technology.
Quick answers
What are VESSL AI's GPU targets?
VESSL AI targets 10,000 GPUs this year, 50,000 GPUs by 2027 and 200,000 GPUs by 2030, along with 8 AI data center bases across South Korea by next year.
What revenue does VESSL AI expect?
The company expects 250 billion won in revenue this year and targets 2 trillion won in revenue by 2028.
What Nvidia cluster is VESSL AI planning?
VESSL AI plans to open an Nvidia DGX B200 NVL72 cluster in the first half of next year.