AMAX Unveils Ultra High-Density 128-GPU Liquid-Cooled Rack for Next-Gen NVIDIA AI Workloads

Built to accelerate hyperscale AI workloads, AMAX’s new 128-GPU liquid-cooled rack delivers dense, energy-efficient performance for next-gen data centers.

Amax Holding Co., Ltd. (TWSE:(6933.TW))

FREMONT, CA, UNITED STATES, May 19, 2025 /EINPresswire.com/ -- AMAX, a global leader in high-density GPU compute cluster solutions and liquid-cooling technology, announces the launch of its LiquidMax^® RackScale 128, expanding its portfolio of total rack-scale solutions. Built for the growing demands of AI training, inference, and HPC workloads, the solution delivers exceptional compute performance and energy efficiency for AI-driven data centers.

Modern AI workloads are pushing traditional infrastructures to their limits. The RackScale 128 combines a fully liquid-cooled design with support for up to 128 NVIDIA Blackwell GPUs, providing scalable, efficient AI performance. Its modular architecture with integrated leakage detection increases density, improves reliability, and reduces energy consumption.

“This solution represents a major step forward in rack-scale density,” said Rene Meyer, Chief Technology Officer of AMAX. “With compute scale for 128 NVIDIA Blackwell GPUs and liquid cooling built into the architecture, we’re helping customers deploy powerful AI infrastructure with greater efficiency and less complexity.”

Key Features of the AMAX LiquidMax^® RackScale 128

High-Density GPU Computing
The RackScale 128 delivers up to 128 NVIDIA Blackwell GPUs in a single 51OU rack, utilizing 5OU liquid-cooled compute trays. Each tray houses two compute nodes paired with two 8-GPU systems, providing high density and scalability for large-scale AI training and inference tasks.

Advanced Liquid Cooling Technology
Direct-to-chip liquid cooling ensures efficient heat dissipation from CPUs and GPUs, minimizing thermal limitations even under peak workloads. Options for both liquid-to-liquid Coolant Distribution Units (CDUs) and Rear Door Heat Exchangers (RDHx) enable adaptability for facilities with varying air-conditioning capabilities.

Leakage Detection and Protection
The system integrates intelligent leakage detection trays and monitoring, safeguarding high-value infrastructure and ensuring reliable data center operations.

Efficient Rack-Scale Power and Serviceability
The rack uses a centralized 48V bus bar for power distribution and supports blind-mated server and power connections, enabling easier installation, maintenance, and upgrades without manual cabling. This design optimizes airflow, supports higher thermal loads, and reduces cable complexity.

Energy Efficiency and Low TCO
Through efficient power and cooling design, the RackScale 128 reduces operational costs and Power Usage Effectiveness (PUE), delivering energy efficient performance for demanding AI and HPC workloads.

Optimized for Growing AI Demands
By integrating liquid cooling at the rack level and applying an open, modular rack design, AMAX’s RackScale 128 provides a highly scalable, energy-efficient foundation for data centers providing the infrastructure for the next generation of AI. The modular architecture also simplifies expansion and upgrade paths, allowing organizations to maintain long-term flexibility without disruptive retrofits.

As part of AMAX’s portfolio of liquid-cooled solutions, this rack represents a major step forward for deploying high-density AI infrastructure with efficiency and reliability.

About AMAX
AMAX is a leading solutions provider of advanced computing solutions for AI, HPC, and data center applications. With over 40 years of engineering excellence, AMAX specializes in designing, building, and deploying customized systems that deliver superior performance, reliability, and efficiency. For more information, visit www.amax.com.

Charla Bunton-Johnson
AMAX Engineering
+1 408-888-6162
email us here
Visit us on social media:
LinkedIn
YouTube
X

Legal Disclaimer:

EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.