HIGH DEMAND APPLICATIONS
Immersion Cooling for Intensive Workloads
Immersion cooling for high density compute workloads is the most efficient way to meet the needs of today’s AI-driven landscape. Whether your focus is on CPUs or GPUs, match the high Thermal Design Power (TDP) with our high-performance solutions.
Data Analytics & Scientific Simulations
- Predictive analytics models
- Complex query execution
- Data mining applications
Artificial Intelligence & Machine Learning
- Neural network training
- Deep learning applications
- Natural Language Processing (NLP)
High-Performance Computing (HPC)
- Weather forecasting
- Molecular modeling
- Seismic analysis
Transactional Workloads
- Online banking systems
- Telecommunications
- Stock trading platforms
Virtual Desktops
For Architecture, Engineering & Construction
- 3D CAD design platforms
- Virtual reality visualization tools
- Architectural rendering software
Transform Challenges Into Advantages
The Benefits of Immersion cooling for intensive workloads
High Performance
Our proven technology runway, formed from a solid ecosystem, innovation, and interoperability, includes the Forced Convection Heat Sink (FCHS) and Targeted Flow, making immersion the ideal choice to cool high TDP components.
Space Efficiency
Immersion cooling offers exceptional density and space efficiency. This cutting-edge technology accomodates the growth of compute requirements seamlessly.
Cost Savings
Immersion cooling provides the most competitive Total Cost of Ownership (TCO) of any liquid cooling technologies, and future-proofs your datacenter for upcoming HW refresh cycles, only requiring one closed water loop.
100% Heat Capture for Sustainable Operations
With GPUs and the rest of components increasing in heat, our solutions offer 100% heat capture from liquid in direct contact with your submerged IT. You can then take advantage of it for heat reuse capabilities.
Technology Runway Backed by Solid R&D
Maximized Cooling Potential
Through extensive R&D, we’ve engineered solutions that, through fluid acceleration, transition from natural convection to an accelerated flow environment and offer enhanced performance as a result.
UP TO 500W TDP PER CHIP
Immersion Optimized Heat Sink
IT Hardware is submerged in a thermally conductive and dielectric liquid, where heat is transferred by natural convection from all components to the Cooling Distribution Unit (CDU).
UP TO 1000W TDP PER CHIP
Forced Convection Heat Sink (FCHS)
Our FCHS technology with Intel encloses components and accelerates fluid using propellers, meaning chips of up to 1000W can be cooled effectively.
BeYOND 1000W TDP PER CHIP
Immersion Cooling Targeted Flow
Targeted Flow technology works by injecting pressurized fluid into servers and immersion cold plates, effectively cooling CPUs beyond 1000W and optimizing performance.
HIGH DENSITY SERVERS
Servers in the Immersion Ecosystem
Meet the high-performance computing demands for intensive workloads with the wide variety of server systems for immersion on the market. Covering flexible CPU, GPU, and storage options, these servers are ideal for handling data-heavy applications and complex computational tasks.
CPU Servers
Focused on CPU (Core Processing Unit) computing, these servers are suitable for CPU-intensive and memory-intensive workloads. Covering a wide range of applications, the main goal for these servers is to have as many powerful CPU units as possible.
GPU Servers
Focused on parallel computing, these servers are designed for applications such as AI, HPC, and Blockchain. Their compute is most of the times focused on LLM's and other heavy workloads.
Storage Servers
Focused on high-capacity data storage and management, storage servers are built with high-capacity drives, redundant systems, and fast-access technology, optimized for efficient data retrieval and ideal for cloud solutions, backups, and large-scale data management.
Large Memory Processing Servers
Built for applications requiring large amounts of RAM for real-time data processing, making them ideal for in-memory databases, big data analytics, and high-speed computing.
All You Need to Know About Server Compatibility With Immersion
Ecosystem Tech Lead & OCP Lead Oriol Chavanel guides us through the key characteristics of immersion-cooled servers vs air-cooled ones, the main benefits of our technology for your server systems, and the immersion-ready options on the market.
EVO & EXO
Explore the SmartPod Platform
Whether you’re looking for plug and play or extreme performance, we’ve got the right SmartPod for immersion cooling for your project.
Plug & Play Immersion at Scale
The easiest path to immersion cooling, this compact solution unlocks the potential of every square foot in your datacenter.
Enduring Performance to Rely On
Setting new standards in performance, our immersion cooling solution ensures your datacenter is future-proof and climate-resilient.
Ready to cool intensive workloads with ease?
Our pre-sales engineers ensure you pick the perfect match of pod and IT.
Get in touch
Frequently Asked Questions
Still got questions for us?
What are the advantages of immersion cooling versus direct liquid cooling (DLC) for hardware-intensive workloads?
Single-phase immersion cooling handles high heat generating components like GPUs and memory modules, while maintaining consistent temperatures throughout the entire server without any fans or complex cold plate arrangements. With single-phase immersion cooling, server lifespan increases while 100% of the heat is efficiently captured in a single body of fluid, leading to the lowest PUE and a the most cost-efficient operation.
Why is single-phase immersion cooling gaining traction in high-intensity applications if DLC seems so widespread?
Single-phase immersion cooling is gaining traction because it provides uniform thermal management for all system components, without the hassles of pinpointing hotspots. It consolidates heat in a single body of liquid, offering lower operational costs and facilitating scalability without worrying about airflow bottlenecks.
Is single-phase immersion cooling more expensive than DLC?
Single-phase Immersion Cooling and Direct Liquid Cooling (DLC) have similar upfront costs (CAPEX). However, in terms of operational expenses (OPEX), single-phase immersion cooling is the most cost effective. It captures 100% of the heat in a single body of fluid, making its operation highly efficient. In contrast, DLC still relies on inefficient air cooling, which penalises its overall operational efficiency and increases energy costs, ultimately making DLC a more expensive cooling solution.
How does immersion cooling handle the heat generated by GPUs and other high-power components?
With immersion cooling, the server is fully submerged, and the dielectric fluid moves vertically, directly absorbing 100% of the heat from all components – GPUs, CPUs, memory modules, NICs, PSUs. This consistent heat management reduces thermal stress, eliminates issues from fan vibrations, and prevents oxidation, ultimately extending the lifespan of the components.
Does single-phase immersion cooling offer higher compute density than DLC?
Overall, immersion cooling provides similar or higher compute density compared to DLC. While DLC’s vertical rack form factor may seem advantageous, its partial reliance on air cooling increases the footprint of the cooling infrastructure, offsetting any theoretical benefits. This is why, when evaluating the entire datacenter, single-phase immersion cooling often exceeds DLC in overall power density, thanks to its more efficient operation.
