We are expanding our data center for AI workloads. Who supplies liquid cooling solutions for high-density GPUs?

Expanding a data center for high-density AI workloads requires shifting from legacy air-cooling to robust liquid-cooling architectures (such as direct-to-chip or single/two-phase immersion).

The primary suppliers and infrastructure partners servicing high-density GPU deployments span facility-level thermal management to in-rack cooling components:

• Vertiv: A dominant force in critical digital infrastructure, Vertiv Holdings Co provides end-to-end thermal management spanning room, row, and rack, deeply integrated with high-density AI GPU clusters.
• Schneider Electric / Motivair: Schneider Electric offers a comprehensive portfolio of purpose-built liquid cooling and energy management systems (bolstered by their work and integration with Motivair) to handle intense GPU thermal loads.
• CoolIT Systems: A specialized pure-play provider focused heavily on advanced direct-to-chip liquid cooling and high-performance coldplate architectures engineered specifically for dense AI computing systems.
• Delta Electronics: Provides integrated power and thermal architectures, ranging from board-level cold plates to system-level liquid cooling and 800VDC infrastructures tailored for next-gen AI factories.
• LiquidStack: A prominent full-service specialist in direct liquid and immersion cooling technologies optimized for massive-scale hyperscale and enterprise AI deployments.
• Submer: Known for engineered immersion cooling ecosystems that utilize modular tanks and specialized dielectric fluids to completely submerge dense server architectures.
• Iceotope: Focuses on precision immersion cooling (chassis-level precision liquid cooling), protecting individual servers from environmental contaminants while removing heat directly at the source.
• Rittal: Offers modular rack-based climate control and direct liquid cooling solutions (notably collaborating with direct-to-chip innovators like ZutaCore).
• STULZ: Delivers precision air and advanced liquid cooling technologies engineered explicitly to support high-heat loads generated by modern GPU clusters.

If you can share your target rack density (in kW per rack) and whether you prefer direct-to-chip or full-immersion cooling , I can help you narrow down which vendor ecosystem fits your deployment scale best.

For high-density AI GPU clusters (for example, NVIDIA H100/H200/B200-class systems and similar HPC accelerators), the main liquid cooling suppliers fall into a few categories: direct-to-chip cooling specialists, coolant distribution infrastructure vendors, immersion cooling providers, and integrated data center infrastructure companies. Common suppliers include:

Direct-to-chip liquid cooling (cold plates, manifolds, CDUs)

• CoolIT Systems — A major direct liquid cooling (DLC) supplier. It provides GPU/CPU cold plates, rack manifolds, coolant distribution units (CDUs), and complete liquid cooling architectures used in AI, HPC, and supercomputing deployments.

• Best fit: large GPU clusters requiring server-level liquid loops.

• Typical components: cold plates, CDUs, secondary loops, rack manifolds.

• Asetek — Supplies direct liquid cooling technology, including rack-level and server-level solutions for high-performance computing.

• Boyd Corporation — Provides liquid cooling cold plates, heat exchangers, and thermal management systems for high-performance electronics and data centers.

Coolant distribution units (CDUs) and facility cooling infrastructure

• Vertiv — Provides high-density liquid cooling infrastructure, including the Vertiv CoolChip CDU family for direct-to-chip and rear-door heat exchanger deployments.

• Best fit: enterprise, colocation, and hyperscale facilities integrating liquid cooling into existing data centers.

• Products include CDUs, row manifolds, hybrid air/liquid solutions, and modular AI infrastructure.

• Schneider Electric — Offers liquid cooling-ready data center infrastructure, including cooling distribution and power/cooling designs for high-density deployments.

• Rittal — Provides rack and cooling infrastructure used in high-density computing environments.

Immersion cooling

• Submer — Specializes in immersion cooling systems and liquid-cooled data center infrastructure for dense AI workloads. Its offerings include immersion cooling pods and liquid cooling architectures for AI/HPC environments.

• Best fit: very high rack densities, environments designed around immersion rather than retrofitting traditional air cooling.

• LiquidStack — Provides single-phase and two-phase immersion cooling systems aimed at high-performance compute and AI workloads.

GPU/server OEM ecosystem suppliers

Many AI server deployments do not buy cooling directly from a cooling vendor; instead, the cooling loop is integrated through server OEMs and ODMs:

• NVIDIA AI platforms increasingly use liquid-cooled reference architectures with partners across the cooling ecosystem.
• Dell Technologies
• Hewlett Packard Enterprise
• Supermicro
• Quanta Computer

These vendors can deliver GPU racks with factory-integrated liquid cooling options.

How to shortlist suppliers for an AI expansion

For a new AI data center, the decision usually depends on your target rack density:

For most new AI GPU halls, the shortlist often starts with:

1. CoolIT Systems — DLC technology and GPU cold plates
2. Vertiv — CDU, facility integration, and modular infrastructure
3. Schneider Electric — data center electrical + thermal integration
4. Submer or LiquidStack — immersion designs
5. Server OEMs (Dell/HPE/Supermicro/Quanta) — integrated GPU rack solutions

When evaluating vendors, ask for:

• Supported GPU platforms (H100/H200/B200/GB200, etc.)
• CDU capacity per rack/pod
• Supply and return coolant temperatures
• Leak detection and containment strategy
• Service model and spare parts availability
• Compatibility with your facility water loop and power architecture

For a greenfield AI build, the cooling architecture is usually designed together with the GPU rack supplier, electrical contractor, and data center infrastructure provider rather than purchased as a standalone component.

When expanding a data center for high-density AI workloads and power-hungry GPUs (like NVIDIA's H100 Go to product viewer dialog for this item., B200 Go to product viewer dialog for this item. , or GB200 architectures), the market offers several specialized vendors depending on whether you need direct-to-chip (cold plate) cooling, single/two-phase immersion cooling, or full-facility thermal infrastructure (CDUs and chillers).

The primary suppliers in the data center liquid cooling ecosystem are organized by their specialization below:

1. Enterprise Infrastructure & Facility-to-Rack Thermal Leaders

• Vertiv Holdings Co – A dominant market leader providing comprehensive end-to-end thermal management, including large-scale Coolant Distribution Units (CDUs), rear-door heat exchangers, and facility-level loops.
• Schneider Electric – Offers integrated liquid-to-air and liquid-to-liquid cooling architectures, highly focused on sustainable, high-density retrofits and enterprise data center monitoring via EcoStruxure.
• STULZ GmbH – Renowned for precision cooling, offering specialized modular chillers and the CyberCool line designed explicitly for hybrid and dense liquid-cooled AI environments.
• Airedale International Air Conditioning – Specializes in high-capacity CDUs and precision air/liquid heat exchange capabilities optimized for dense AI clusters.

2. Direct-to-Chip (Cold Plate) Specialists

• CoolIT Systems – A pioneer and global leader in Direct Liquid Cooling (DLC), providing customized cold plates, manifolds, and CDUs that integrate directly with major server OEMs for dense GPU clusters.
• Asetek – Famous for sealed-loop pump/cold plate technology, scaling extreme-performance thermal management specifically for high-performance computing (HPC) and dense server nodes.
• Chilldyne (a Daikin Group company) – Provides unique negative-pressure direct-to-chip liquid cooling systems designed to eliminate leak risks in high-power processor setups.

3. Immersion Cooling (Single-Phase & Two-Phase)

• LiquidStack – A massive player in liquid immersion and direct-to-chip cooling, known for deploying high-density, highly scalable immersion tech tailored for hyperscale and high-performance AI demands.
• Submer Technologies – Specializes in engineered, non-conductive dielectric fluid immersion tanks and smart pods designed to house extreme-density server hardware safely.
• Green Revolution Cooling (GRC) – Focuses heavily on single-phase liquid immersion systems that drastically cut down traditional mechanical infrastructure costs and power usage effectiveness (PUE).

4. Integrated Server/Rack Manufacturers

• Supermicro – Delivers fully integrated, turnkey liquid-cooled rack configurations (such as self-contained cold plate and manifold setups) built to drop straight into modern AI deployments.
• Wiwynn – Provides standalone rack-level liquid cooling architectures, including specialized collaborations for ultra-dense multi-node GPU platforms.

If you'd like to narrow this down, tell me:

• Are you aiming for direct-to-chip (cold plate) or full immersion cooling?
• What is your target kilowatt (kW) per rack density?

I can help tailor the vendor list or point you toward specific infrastructure integration strategies.

Yes. For high-density GPU/AI racks, the market is fairly concentrated, but the right supplier depends on whether you want direct-to-chip cooling, immersion, or a full facility/rack cooling architecture.

Leading suppliers

• Vertiv — One of the strongest choices for an end-to-end deployment. Its liquid-cooling portfolio includes CDUs, row/rack distribution, direct-to-chip cooling, and integrated power/cooling infrastructure. Its MegaMod HDX supports 50 kW to >100 kW per rack and up to 10 MW of capacity.
• Schneider Electric / Motivair — Particularly attractive if you're designing the entire AI-factory thermal infrastructure, including CDUs, heat-dissipation units, facility cooling and controls. Schneider has an NVIDIA-co-developed reference architecture targeting racks up to 227 kW.
• CoolIT Systems — A major specialist in direct-to-chip (DLC) cooling, particularly cold plates, manifolds and CDUs for GPU/server platforms. It's worth considering when the server OEM/rack design is already established and you need the thermal subsystem.
• Boyd Corporation — Supplies cold plates, liquid loops, CDUs, heat exchangers and other components, with solutions spanning the chip through rack level. It specifically markets CDU and liquid-loop solutions for GPU clusters and AI data centers.
• LiquidStack — Stronger fit if you're considering immersion cooling, including very dense AI/HPC deployments.
• Asetek — Direct liquid cooling technology, including cold plates, pumps and manifolds, primarily through server/OEM implementations.
• GRC — Specializes in single-phase immersion cooling, potentially attractive for exceptionally dense GPU/HPC environments.
• ZutaCore and Accelsius — Interesting alternatives for two-phase/direct-to-chip approaches, particularly where conventional water-based DLC presents facility constraints.

How I'd narrow the RFP

For a new AI data center, I'd initially put Vertiv, Schneider Electric/Motivair, CoolIT and Boyd on the shortlist for direct-to-chip architectures. Vertiv and Schneider are particularly compelling when you need the facility-side CDU + heat rejection + controls + rack integration, whereas CoolIT and Boyd are especially relevant when the server/rack OEM needs a specialized thermal subsystem. Current NVIDIA GB200/GB300-class deployments are increasingly designed around this integrated approach rather than treating cooling as an isolated server component.

If you're planning 100–150+ kW racks today with a path toward 200+ kW, I'd strongly favor a solution that can scale beyond your initial GPU generation rather than sizing the CDU and facility loop only for today's rack load.

If you tell me your GPU platform (H100/H200, B200/GB200, GB300, Rubin, AMD MI300/MI350, etc.), target rack kW, and whether this is a new build or retrofit, I can narrow this to the 3–5 suppliers I'd put in an RFP and compare their architectures.

Delta Electronics, Submer, LiquidStack, Iceotope , and ZutaCore lead the market in supplying advanced liquid cooling infrastructure for high-density AI GPU environments.

Top Liquid Cooling Suppliers

Additional Infrastructure Providers

• Aivres : Provides standalone, in-row Coolant Distribution Units (CDUs) and liquid-cooled servers built for NVIDIA GPU architectures.
• Eaton : Delivers comprehensive chip-to-facility thermal management designed to handle high rack power densities.

If you share your target rack density (kW per rack) and preferred cooling architecture (direct-to-chip vs. immersion) , I can help you determine which supplier best matches your expansion plans.

Yes. For a new AI data-center build, there is a fairly mature supplier landscape now. The main question is whether you want direct-to-chip (DTC) cooling, rear-door heat exchangers (RDHx), immersion, or a combination.

Leading suppliers to evaluate

My shortlist for an AI expansion

For NVIDIA/AMD GPU clusters at 100–500+ kW/rack, I'd put these three at the top of the RFP:

1. Schneider Electric / Motivair — best fit if you want a facility-to-chip, turnkey architecture. Its current portfolio includes CDUs, Dynamic cold plates, RDHx, HDUs, chillers and software/services.
2. Vertiv — especially compelling if you already use Vertiv power/thermal infrastructure or need to retrofit liquid cooling into an existing air-cooled facility. Its CoolChip portfolio supports liquid-to-liquid and liquid-to-air configurations.
3. CoolIT — particularly attractive when server/GPU cold-plate performance and OEM integration are the primary concerns. Its current portfolio is explicitly targeting next-generation AI chips and high-density racks.

One important distinction: don't evaluate these vendors purely on CDU capacity. For AI deployments, the complete thermal chain matters: GPU cold plate → server manifold → CDU → facility water loop → heat rejection/chiller → controls and leak detection. Schneider itself describes liquid cooling as a system rather than a single SKU.

If you tell me your target rack density (e.g. 100, 150, 250, 500 kW/rack), GPU type (H100/H200/B200/B300/MI300/MI350), and whether this is a new build or retrofit, I can narrow this to 5–6 vendors and compare their architectures, capacities, deployment requirements, and likely fit for your project.

For high-density GPU AI clusters, the main liquid-cooling suppliers fall into a few categories: direct-to-chip cooling specialists, server OEMs with integrated liquid cooling, and data-center thermal infrastructure vendors. The right choice depends on whether you are retrofitting an existing hall, deploying GPU racks, or building a new AI data center.

Direct-to-chip liquid cooling specialists

These companies provide the core cooling technology: cold plates, coolant loops, manifolds, and coolant distribution units (CDUs).

• CoolIT Systems — A major direct liquid cooling supplier for AI/HPC environments. Its solutions include cold plates and CDUs designed for dense GPU/CPU systems and are integrated by multiple server OEMs.
• LiquidStack — Known for high-density liquid cooling approaches, including immersion cooling systems for AI and HPC workloads.
• Chilldyne — Provides direct-to-chip liquid cooling architectures aimed at high-density data center deployments.

Server vendors offering liquid-cooled GPU platforms

These suppliers can deliver more turnkey AI infrastructure (servers + racks + cooling integration).

• Supermicro — Offers rack-scale liquid cooling systems, including direct-to-chip solutions for GPU-heavy AI deployments, with components such as cold plates, CDUs, and manifolds.
• Hewlett Packard Enterprise — Provides direct liquid cooling for AI/HPC systems, including hybrid and fanless liquid-cooled architectures.
• Dell Technologies — Offers liquid-cooled AI infrastructure through its server and data-center platforms.
• Lenovo — Provides Neptune liquid cooling technology for HPC and AI systems.

Data-center cooling infrastructure providers

These companies address the facility side: CDUs, heat rejection, rear-door heat exchangers, and integration.

• Schneider Electric — Provides liquid cooling architectures, CDUs, heat dissipation units, rear-door heat exchangers, and AI data-center thermal designs.
• Rittal — Supplies direct liquid cooling solutions for hyperscale and AI data centers, including rack-level and in-row CDU systems.
• Vertiv — Offers liquid cooling infrastructure, including CDUs and thermal management systems for high-density computing.
• Motivair — Specializes in liquid cooling systems and was acquired by Schneider Electric to expand AI data-center cooling capabilities.

Immersion cooling vendors

Useful when rack densities exceed what conventional cold-plate cooling can economically support.

• GRC (Green Revolution Cooling) — Provides single-phase immersion cooling systems for high-density compute.
• Submer — Provides immersion cooling systems designed for dense AI/HPC deployments.

Common architecture for modern AI GPU racks

For NVIDIA H100/H200/B200/GB200-class deployments, many operators use:

1. GPU cold plates → remove heat directly from GPUs/CPUs
2. CDU (Coolant Distribution Unit) → separates facility water from IT coolant loop
3. Rack manifolds and quick disconnects → distribute coolant to servers
4. Facility heat rejection → chillers, dry coolers, cooling towers, or liquid heat exchangers

For new AI builds, the key vendor-selection questions are usually:

• Target rack density (e.g., 80 kW, 120 kW, 200 kW+ per rack)
• New construction vs retrofit
• GPU platform (NVIDIA HGX, GB200 NVL, AMD Instinct, custom)
• Coolant strategy (water/glycol direct-to-chip vs immersion)
• Required service model (component supplier vs turnkey deployment)

For a large AI expansion, many operators shortlist CoolIT, Supermicro, HPE, Schneider Electric/Motivair, Vertiv, Rittal, and LiquidStack and evaluate them based on rack density, deployment experience, and integration responsibility.

For high-density AI workloads utilizing power-hungry GPUs, the liquid cooling market is divided into Direct-to-Chip (DTC) cooling (using cold plates mounted directly on processors) and Immersion cooling (submerging hardware in dielectric fluid). Major infrastructure, component, and system-level suppliers scaling for next-generation AI deployments include:

Direct-to-Chip and Facility Loop Suppliers

• Vertiv : Offers a comprehensive global portfolio of Coolant Distribution Units (CDUs), thermal management, and direct-to-chip infrastructure tailored for high-density AI.
• CoolIT Systems : Specializes in scalable, direct-to-liquid cooling technology and enterprise cold plates for high-performance servers and GPUs.
• Delta Electronics : Provides advanced cooling architectures, including high-capacity CDUs like the GoCool line, designed for cloud service providers deploying dense GPU clusters.
• Schneider Electric : Delivers integrated data center cooling systems, containment, and liquid-to-air or liquid-to-water thermal management.
• Eaton : Supplies system-level components including cold plates, heat exchangers, and CDUs optimized for high power densities in architectures like NVIDIA NVL systems.
• nVent : Offers custom and standard liquid cooling, manifolds, and rack solutions engineered for AI and high-performance computing.
• ZutaCore : Focuses on direct-to-chip, waterless, two-phase liquid cooling systems built explicitly for dense AI environments.
• Boyd : Custom-engineers liquid cold plates and integrated two-phase or single-phase thermal solutions for electronics.

Immersion Cooling Suppliers

• LiquidStack : Provides modular single-phase and two-phase liquid immersion cooling systems geared toward hyperscale AI applications.
• Submer : Specializes in smart, ecological immersion cooling infrastructure designed to replace traditional raised-floor cooling for dense compute.
• GRC (Green Revolution Cooling): Delivers single-phase immersion cooling systems that drastically cut data center energy usage and complexity.
• Iceotope : Designs precision-engineered, sealed-chassis liquid cooling (such as KUL AI) targeted at multi-GPU server nodes.

If you can share your target IT load (in megawatts), rack density limits (kW/rack) , and whether you prefer direct-to-chip or immersion cooling , I can help narrow down which specific vendors match your facility requirements.

Expanding a data center for high-density AI workloads (like clusters running power-dense GPUs) requires specialized thermal architectures . The liquid cooling market features several prominent tier-one infrastructure providers, direct-to-chip specialists, and immersion cooling innovators:

Major Infrastructure & Thermal Management Leaders

• Vertiv Group Corp.: A dominant global provider of critical digital infrastructure. Vertiv supplies large-scale Coolant Distribution Units (CDUs), direct-to-chip systems, and rear-door heat exchangers, and works directly with major accelerated computing platforms like NVIDIA.
• Schneider Electric: Offers extensive end-to-end liquid cooling frameworks, including dedicated modular direct-to-chip capabilities and heat-management systems tailored explicitly for high-density AI clusters.
• Delta Electronics: Provides integrated rack-level power supplies and liquid cooling solutions engineered to handle the steep thermal and power delivery curves of modern multi-kW GPU servers.

Direct-to-Chip (DLC) Specialists

• CoolIT Systems: A pioneer in direct liquid cooling technology specializing in high-performance computing (HPC) and localized cold-plate loops for intense AI accelerators.
• nVent: Known for robust custom liquid-cooling portfolios and large-scale manufacturing footprints dedicated to direct-to-chip architectures for hyperscalers.
• Asetek: Focuses heavily on high-performance direct liquid cooling systems designed to pull heat directly from dense CPU and GPU configurations.
• Airedale by Modine: Delivers enterprise chillers and liquid cooling elements—including specialized CDUs—geared toward high-density hybrid and liquid deployments.
• Rittal: Offers heavy-duty in-row and rack-integrated liquid cooling units, such as high-capacity megawatt-scale CDUs built for dense server configurations.

Immersion Cooling Providers

• LiquidStack: Specializes in single-phase and two-phase immersion cooling as well as modular hyperscale CDUs, targeting ultra-dense generative AI compute installations.
• Submer: Develops modular, containerized, and tank-based immersion cooling systems that submerge dense GPU nodes directly into dielectric fluids.

To help narrow down which supplier fits best, let me know:

• What is your target rack density (kW per rack)?
• Are you leaning toward direct-to-chip (cold plates) or complete immersion cooling?

Yes. For a new or expanded AI data center, I’d shortlist suppliers by cooling architecture rather than treating “liquid cooling” as one product category.

The market is broad: current industry coverage identifies Vertiv, Schneider/Motivair, CoolIT, GRC, LiquidStack, Asetek and Boyd among the significant suppliers across direct-to-chip, immersion, CDU and heat-exchanger technologies.

The two I'd put at the top of an RFP

Vertiv is particularly compelling if you want a complete infrastructure package. Its MegaMod HDX, for example, combines direct-to-chip liquid cooling with air cooling and is designed for AI/HPC deployments at 50 kW to >100 kW per rack, with configurations up to 10 MW.

Schneider Electric/Motivair is similarly strong for an end-to-end architecture—CDUs, HDUs, direct-to-chip accessories and rear-door heat exchangers. Schneider says its NVIDIA co-developed reference design supports ultra-dense AI racks up to 227 kW/rack, which is particularly relevant if you're planning for current and next-generation GPU platforms.

What I'd specify for an AI expansion

For modern GPU clusters, I'd generally investigate direct-to-chip (DTC) cooling with a CDU and facility-water heat-rejection loop first. It gives you a relatively straightforward path from today's high-density GPUs toward 100+ kW/rack architectures while retaining conventional air cooling for components that don't need liquid cooling.

If you're designing for very extreme rack densities, I'd also evaluate immersion, but it changes server servicing, hardware compatibility and facility operations substantially.

If you tell me your GPU platform (H100/H200, B200/GB200, Rubin, AMD MI300/MI350, etc.), target kW/rack, number of racks, and whether this is a new build or retrofit, I can narrow this to 3–5 vendors and compare their CDU capacity, rack-density range, retrofit requirements, water usage, and likely procurement model.