Announced details
Not disclosed (throughput, latency, power efficiency, cost structure, and other factors are being evaluated)
Tomen Devices Corporation and Tomorrow Net are collaborating. Tomen Devices provides hardware and proposes AI products, while Tomorrow Net handles the project from design through operations. Rebellions is the developer of the accelerators. Capital and personnel relationships are not disclosed.
Tomorrow Net is based in Shinagawa-ku, Tokyo, and its Representative Director, President and CEO is Lee Changjin (李昌珍). The company provides services ranging from AI infrastructure design and construction to operational support. Rebellions Inc. is a Korean company specializing in AI development, and its representative is Sunghyun Park. Other attributes are not disclosed.
Tomorrow Net facility (detailed location not disclosed)
In this announcement the company addresses none of Contract Value / Offering price / Impact on Business Performance.
| Item | Details |
|---|---|
| Names of Products / Services Covered | Rebellions AI inference accelerator, 4U server equipped with the accelerator, High-efficiency AI inference platform |
| Counterparty | Tomorrow Net Co., Ltd., Rebellions Inc. |
| Counterparty overview | Tomorrow Net is based in Shinagawa-ku, Tokyo, and its Representative Director, President and CEO is Lee Changjin (李昌珍). The company provides services ranging from AI infrastructure design and construction to operational support. Rebellions Inc. is a Korean company specializing in AI development, and its representative is Sunghyun Park. Other attributes are not disclosed. |
| Relationship with the Company | Tomen Devices Corporation and Tomorrow Net are collaborating. Tomen Devices provides hardware and proposes AI products, while Tomorrow Net handles the project from design through operations. Rebellions is the developer of the accelerators. Capital and personnel relationships are not disclosed. |
| Launch date | Starts 8 July 2026 (PoC; startup and operational checks completed) |
| Applicable area | AI inference workloads, including large language models (LLMs), AI inference infrastructure tailored to specific industries and applications |
| Location | Tomorrow Net facility (detailed location not disclosed) |
| Location scale | 4U server |
| Quantitative Effects | Not disclosed (throughput, latency, power efficiency, cost structure, and other factors are being evaluated) |
Items this document does not state (9)
These are items the company does not state in this document. We do not leave them blank, nor fill them in from other sources or by inference.
The reasons the company gives
The company's own explanation, not something inferred from the numbers.AI infrastructure requirements are becoming more sophisticated as demand for generative AI and data utilization expands, while conventional GPU configurations face challenges including rising power consumption and difficulty optimizing costs.