Mitsubishi Heavy Industries Ltd. (MHI) has introduced DIAVAULT, a new industrial-grade edge data center brand designed to enhance on-premises data and AI utilization. The solution promises high security and low latency, addressing challenges often faced in cloud environments.
DIAVAULT is equipped with a stable power supply and advanced cooling systems, ensuring long-term reliable operation. MHI aims for DIAVAULT to be a pivotal element in accelerating digital transformation across the industrial sector, particularly as the Internet of Things (IoT) and AI become more prevalent.
The rise of IoT and AI has amplified the need for real-time data processing, which cloud-centric systems struggle to meet due to latency and data confidentiality concerns. Edge computing, which allows for faster and distributed data processing close to data sources and users, is becoming increasingly important. MHI's development of on-premises digital infrastructure with DIAVAULT seeks to address these needs.
DIAVAULT offers flexibility, catering to various customer requirements from small-scale setups to larger inference data centers of several megawatts. It provides edge solutions for diverse applications, including manufacturing sites, research facilities, defense sectors, and 5G connectivity, all of which demand low latency.
Recently, MHI inaugurated a demonstration AI data center at the Yokohama Hardtech Hub (YHH) within its Yokohama Works premises. This facility features servers with next-generation GPUs optimized for AI processing, cooled by two-phase direct chip cooling technology. This setup delivers high computing performance and an impressive Power Usage Effectiveness (PUE) rating.
The compact server room, equivalent to two 20-foot containers, offers ample space for maintaining large-depth servers and the Coolant Distribution Unit (CDU). It includes an uninterruptible power supply for continuous power delivery and employs proprietary digital integrated air conditioning control technology, which optimizes air conditioning and heat sources based on server load for year-round energy efficiency.
Security is a priority, with robust entry and exit protocols involving two-factor authentication and remote monitoring capabilities for stable unmanned operations. MHI's engineering expertise in social infrastructure, data center power, cooling, and digital technologies underpins the development of DIAVAULT, aiming for high resilience and security to support next-generation AI.
Looking ahead, MHI plans to promote the adoption and co-creation of advanced AI infrastructure through partnerships both domestically and internationally, leveraging DIAVAULT to meet the evolving demands of the digital landscape.
Comments are closed.