Direct-to-Chip Cooling
CHOOSING THE PERFECT COOLING SOLUTION
Direct-to-chip cooling represents a cutting-edge approach to thermal management in data centers, where cooling mechanisms are applied directly to the heat-generating components of processors and other hardware.
By bringing cooling elements closer to the heat source, direct-to-chip cooling solutions such as liquid cooling or microfluidic systems enhance heat dissipation efficiency and enable more precise temperature control. This innovative cooling method not only improves the overall performance and reliability of servers but also allows data centers to operate at higher power densities.
Direct-to-chip cooling technology plays a critical role in optimizing energy efficiency, reducing cooling costs, and maximizing the computing capacity within a given space. As data centers continue to evolve, direct-to-chip cooling emerges as a key solution in the pursuit of sustainable, high-performance computing infrastructures.
Benefits of Direct-to-Chip Cooling in Data Centres
Liquid cooling through Coolant Distribution Units (CDUs) has revolutionized cooling performance in data centers, offering unparalleled efficiency and enhanced thermal management capabilities.
By utilizing liquid as a cooling medium, CDUs facilitate the transfer of heat more effectively than traditional air cooling methods, allowing for greater heat dissipation and improved temperature regulation.
This advanced cooling technology not only optimizes the thermal management of servers and equipment but also enables data centers to operate at higher power densities without compromising performance or reliability.
To fully realize the benefits of direct-to-chip cooling, advanced infrastructure is required. Coolant Distribution Units (CDUs) are the workhorses of this ecosystem. They safely and precisely isolate the facility water loop from the IT equipment cooling loop, ensuring that the massive heat loads generated by clusters of NVIDIA GB300s or AMD Helios chips are efficiently rejected without compromising the safety of the server racks.
We recommend establishing direct-to-chip cooling as a core data centre strategy. The advantages of deploying this technology extend far beyond simply managing temperatures:
- Enabling technologies such as CDUs and liquid cooling support the operation of incredibly high-density racks.
- The strategy significantly improves overall data centre energy efficiency.
- Liquid loops enhance hardware reliability by maintaining consistent operating temperatures.
- This cooling method is critical for enabling intensive AI and HPC workloads.
COOLING REDEFINED WITH DIRECT-TO-CHIP INNOVATION
Direct-to-chip cooling offers a host of benefits that significantly enhance the thermal management and efficiency of data centers. By delivering cooling directly to heat-generating components at the chip level, this innovative approach maximizes heat dissipation efficiency and enables precise temperature control, resulting in improved performance and reliability of servers and hardware.
Direct-to-chip cooling systems also facilitate higher power densities, allowing data centers to operate at increased computational capacities within the same physical footprint. Additionally, this method reduces cooling energy consumption, lowers operational costs, and enhances the overall sustainability of data center operations.
By harnessing the advantages of direct-to-chip cooling, data centers can achieve superior thermal management, optimize energy efficiency, and bolster the performance and longevity of their infrastructure.
Direct-to-chip innovation relies on custom-engineered cold plates that physically attach to the hottest components of your server—the CPU and GPU dies. For systems like the NVIDIA GB200 NVL72, where dozens of GPUs operate in a tightly integrated architecture, micro-channel cold plates capture heat instantaneously at the source. This targeted approach dramatically reduces the need for power-hungry server fans and oversized computer room air conditioning (CRAC) units.
Coolant Distribution Units (CDUs) for Direct-to-Chip Cooling
The precision control and efficiency of liquid cooling systems through CDUs have significantly contributed to reducing energy consumption, minimizing carbon footprint, and maximizing the overall cooling performance of modern data centers, making them a cornerstone of sustainable and high-performance computing infrastructures.
To establish the proper product category, it is vital to understand the systems that drive the fluid through the cold plates. The CDU is the heart of any direct-to-chip liquid cooling deployment, managing flow rates, pressure, and coolant temperatures to ensure optimal thermal transfer from high-wattage silicon
Future of Coolant distribution units & Direct to chip 1MW rack incoming!
RackChiller CDU800 Coolant Distribution Unit
FEATURES
- Redundant high-performance, leak-free pump system
Integrated variable speed drives - Coolant connections through top or bottom panel
- Integrated 10-inch touch panel display.
- Remote control features through Ethernet, SNMP v3, Modbus
- On-board integrated leak detection
- Unrivaled power density – fits into standard data center footprint
- Serviceable during operation – no need for shut down during system maintenance.
- Redundant system layout minimizes risk for single points of failure.
- Integrates with nVent Guardian Management Gateway and sensors portfolio.
SPECIFICATIONS
General Data
- 800+kW of cooling capacity @ 6K (850 LPM Primary)
- Pipe Connection: 3-inch ID hygienic tri-clamp
Liquid Temp Range: 20 - 70 C (68 - 158 F)
Primary Rating
- Coolant: treated water with up to 20% PG
- Maximum Allowable Flow Rate: 1200 LPM (317 GPM) •
- Maximum Head Loss (at 850 LPM, Water): 1.3 Bar (19 psi)
- Maximum System Pressure: 10.3 Bar (150 psi)
- System Volume: 50 L (13 Gal)
Primary Filter Size: 250 micron
Secondary Performance
- Coolant: treated water with up to 30% PG
- Maximum Flow (single pump): up to 1100 LPM (290 GPM) at 2.6 bar (38 psi)
- Maximum Flow (dual pumps): up to 1100 LPM (290 GPM) at 3.4 bar (49 psi)
- Maximum Allowable Static Pressure: 3.5 Bar (50psi)
- Maximum System Pressure: 8.6 Bar (125 psi)
- Pressure Relief Valve Activation Pressure: 9.0 Bar (130 psi)
- System Volume: 100 L (26 Gal)
- Secondary Filter Size: 50 micron
Videos
