AHEAD, a data center infrastructure integrator, has announced a new integration facility focused on rack-scale, direct-to-chip, liquid-cooled infrastructure. The 68,000-square-foot facility, located in Libertyville, Illinois, is a response to the growing demand for AI and high-performance computing (HPC) infrastructure. It is scheduled to begin operations in 2025 and is expected to create approximately 130 tech manufacturing jobs.
Meeting the demands of AI and HPC workloads
The new facility adds to AHEAD’s existing Libertyville campus, which already includes multiple integration and distribution facilities. The company also operates integration facilities in the U.K. and EU. The facility is designed to address the increasing power and cooling demands of AI and HPC workloads, which are exceeding the capabilities of traditional data center infrastructure.
Reclaim Health and WEX partner to integrate AI-powered healthcare financial management
The facility will specialize in both air and direct-to-chip, liquid-cooled rack integration projects. Key benefits of liquid cooling, according to AHEAD, include:
- Superior heat management: Direct-to-chip liquid cooling is significantly more efficient at dissipating heat than air cooling, crucial for GPU-intensive workloads.
- Reduced energy consumption: Liquid cooling reduces power usage, leading to cost savings and supporting sustainability goals.
- Space optimization & high rack density: Liquid-cooled racks allow for higher compute density, maximizing floor space utilization.
- Lower water & carbon footprint: Direct-to-chip cooling can achieve significantly lower power usage effectiveness (PUE) values compared to air-cooled data centers.
The facility will include a 60-rack innovation lab where organizations can benchmark performance gains across different AI and HPC architectures. The lab will feature various accelerated compute, storage, and network architectures, and will enable real-world deployment scenario testing and side-by-side comparisons of thermal efficiency and power consumption.
AHEAD cites a Gartner report from October 2024 projecting that by 2028, 25% of new servers will be equipped with dedicated workload accelerators for generative AI, highlighting the need for efficient cooling technologies.
Featured image credit: AHEAD
