Blockmate Ventures는 NVIDIA B200 GPU 투자를 통해 AI 컴퓨팅 시장에 진출한다.

Blockmate Ventures Chairman Domenic Carosa joined Steve Darling of Proactive to announce that the company has entered into a binding agreement with Scannit Inc. to acquire its first NVIDIA B200 AI compute node. The acquisition marks Blockmate's entry into the AI neocloud and GPU computing market.

Under the agreement, dated October 27, 2026, Blockmate will acquire a high-performance computing node featuring eight NVIDIA B200 GPUs for approximately US$635,000. The turnkey purchase price includes shipping, data centre setup, installation, network connectivity and onboarding.

The acquisition represents a shift in Blockmate's business model, as the company moves from a venture builder toward an owner of AI infrastructure. Blockmate intends to generate revenue by owning specialised computing hardware and making its capacity available to AI developers and businesses through a specialist operating partner. Blockmate is expected to receive the majority of the node's gross revenue. The initial management term will run for 36 months from the date the equipment becomes operational.

Delivery of the initial node is currently expected in the first quarter of 2027, subject to vendor timing and NVIDIA allocation. Installation, testing and commissioning are expected to take approximately one week after delivery, after which the equipment will be available for commercial workloads, subject to operational readiness.

The business targets growing demand for specialised computing power used to train and operate AI models. Rather than investing in their own expensive infrastructure, AI developers and businesses can rent GPU capacity on demand or through longer-term agreements.

Blockmate views the first deployment as a test of its operating model. The company aims to collect real-world data on customer demand, utilisation rates, rental pricing, uptime, hosting and power costs, and the effectiveness of different sales channels.

The initial acquisition gives Blockmate an opportunity to establish a presence in the AI infrastructure market without independently managing all aspects of data centre operations. Its ability to generate attractive returns will depend on commercial utilisation, operating expenses and the economics of renting out its computing capacity.