Artigo

NVIDIA Vera Rubin NVL72: technical and executive guide to liquid cooling and the AI Factory

EnQ Digital·04 de setembro de 2026

An NVIDIA Vera Rubin NVL72 is not just an ultra-high-density GPU rack. It is the core of an AI Factory: a platform that converts large volumes of data and electrical power into model training, inference, AI reasoning, scientific simulation and business outcomes at scale. This guide walks through the entire decision chain — from the business case to thermal and hydraulic sizing, CDU selection, commissioning and operation.

The chain is always the same: demand defines scale; scale defines power and thermal load; thermal load defines flow, ΔT and ΔP; and ΔP defines the pump, the CDU and the hydraulic plant. Skipping a step transfers uncertainty to the construction site, where correction costs more and compromises schedule.

About this material: the figures here are public platform and pre-design reference values. NVIDIA still classifies several Rubin specifications as preliminary. The Reference Design and the Site Planning Guide for the configuration actually purchased must take precedence over any preliminary calculation in this text.

NVIDIA Vera Rubin NVL72 rack displayed next to an 800 VDC power rack on exhibit
Vera Rubin NVL72 on display next to the 800 VDC power rack. Credit: NVIDIA / editorial record reproduced in the EnQ Digital e-book.

Executive summary

Each part answers a specific decision. Executives start with the first three lines; engineering moves through the technical disciplines; operations focuses on commissioning, telemetry and indicators; procurement uses the matrices and the RFI to structure quoting, technical equalization and acceptance.

PartContentMain decision
Business caseWhen an AI Factory is justifiedBuild, rent GPU in the cloud, or hybrid?
ApplicationsSectors, workloads and objectivesWhat workload will occupy the cluster?
Real casesBlackwell GB200/GB300 in productionWhat economic model has already been validated?
ArchitectureRubin GPU, Vera CPU, HBM4, NVLink 6What does the rack deliver as a system?
CoolingDLC, FWS, CDU, TCS and 45 °CHow does heat leave the chip and reach the atmosphere?
SizingkW, LPM, m³/h, ΔT, ΔP and Pump HeadWhat flow and pressure does the project require?
CDU selectionCriteria, DSX marketplace and manufacturersWhich CDU meets the actual operating point?
ElectricalCabinet TDP, kVA and consumptionHow much power and energy should be reserved?
OperationBMS, commissioning, risks and RFIHow do you prove and sustain performance?

Why build a supercomputer of this class

The central goal is to drastically reduce the time needed to train, fine-tune and run models and simulations that would be impractical on conventional infrastructure — while preserving low latency between GPUs, large aggregate memory and high efficiency per unit of work. The value is not just in the FLOPS.

  • Time-to-solution: reducing weeks or months of training and simulation to days or hours.
  • Larger models: training, fine-tuning and serving LLMs, multimodal models and Mixture-of-Experts with hundreds of billions or trillions of parameters.
  • Reasoning at scale: sustaining inference with long context, agents and reasoning chains that consume far more compute than a traditional chatbot.
  • Sovereignty and privacy: keeping data, models and intellectual property within infrastructure controlled by the organization or the country.
  • AI + HPC convergence: using the same factory for AI, massive data analysis, numerical simulation and scientific workloads.
  • Per-token economics: in large, continuous workloads, performance per watt, utilization and communication speed determine TCO.
  • Monetization: cloud providers and data centers can turn the infrastructure into GPU-as-a-Service, training-as-a-service and inference-as-a-service.

When it makes sense — and when it doesn't

An NVL72 cluster is justified when there is sustained demand for parallel computing, large datasets, strong GPU-GPU communication, a need for low latency, data sovereignty, or an inference volume sufficient to amortize the infrastructure. For small, experimental projects or those with sporadic utilization, public cloud or smaller clusters tend to offer a better economic fit.

IndicatorSignal that an AI Factory may make sense
TrainingLarge proprietary models, recurring training or continuous fine-tuning.
InferenceBillions of tokens/day, low latency and agents executing complex workflows.
DataPetabytes or strategic datasets that must not leave the controlled domain.
HPCHighly parallel simulations: CFD, FEA, climate, molecular dynamics or seismic.
BusinessCapacity used continuously or resold as a GPU/AI service.
SovereigntyData residency, security, compliance or intellectual property requirements.

Which areas the platform is useful for

The same infrastructure serves generative AI and HPC. The gain appears mainly when the problem can be parallelized and requires large memory bandwidth and communication between accelerators.

AreaTypical workloadsObjective
AI Labs / LLMFoundation models, MoE, multimodal, reasoning, fine-tuning and RL.Create proprietary models and reduce training time.
Cloud / Data CenterGPUaaS, bare metal GPU, inference endpoints, dedicated clusters.Monetize capacity and serve AI/HPC customers.
Banking and financeFraud, AML, risk, pricing, tabular models, agents and document analysis.Process large datasets and run private AI with low latency.
Healthcare and life sciencesDrug discovery, molecular dynamics, genomics, protein design and medical imaging.Shorten research cycles and test more hypotheses.
Industry and engineeringCFD, FEA, digital twins, generative engineering, robotics and industrial vision.Reduce physical prototyping and speed up optimization.
Energy, oil and gasSeismic imaging, reservoir simulation, grid optimization and predictive maintenance.Increase resolution and reduce operational decision time.
Climate and environmentWeather forecasting, climate modeling and data assimilation.Run faster, higher-resolution forecasts.
TelecomRF planning, network optimization, AI for NOC and traffic models.Automate operations and predict congestion and failures.
Government / Sovereign AINational LLMs, research, cyber defense and public services.Preserve data sovereignty and strategic AI capacity.
CybersecurityLarge-scale detection, malware analysis and SOC with agents.Correlate large volumes of telemetry and accelerate response.
Media and contentVideo, image, audio generation, translation and neural rendering.Produce multimodal content at large scale.
Scientific researchPhysics, chemistry, materials, astronomy, bioinformatics and hybrid methods.Explore solution spaces previously unfeasible due to computational cost.

Practical example: an institution can use the same cluster during the day for inference and corporate agents and, in planned windows, allocate thousands of GPUs for training, simulation or scientific processing via Kubernetes/Slurm. The value comes from shared utilization and the ability to orchestrate distinct workloads.

Real cases with Blackwell B200 and GB200 NVL72

Blackwell systems that predate Rubin already demonstrate how an AI Factory is used in production. These cases help clarify the type of workload and the economic model that Vera Rubin tends to expand.

Naming: B200 and B300 refer to the Blackwell and Blackwell Ultra GPU generations. GB200 and GB300 combine Grace CPU + Blackwell GPU in the Superchip and, in the NVL72, form the rack-scale system with 72 GPUs.

CaseSystemWhat it demonstrates
CoreWeaveGB200 NVL72In April 2025, NVIDIA announced thousands of active Grace Blackwell GPUs; Cohere, IBM and Mistral AI began training and serving reasoning and agentic AI models.
Oracle / OCIGB200 NVL72 liquid-cooledFirst wave of racks in operation, with thousands of Blackwell GPUs for OCI and DGX Cloud, combining Quantum InfiniBand and Spectrum-X Ethernet.
NeoSpaceGB200 NVL72 via OCILatin American case: NeoData in a multi-cloud environment for one of the largest private banks in Latin America, with more than 60 million customers.
DeepLGB200Training of next-generation Mixture-of-Experts models, seeking efficiency in both training and inference.
Microsoft / OpenAIGB300 NVL72Large-scale Azure cluster: throughput, latency, memory and network scaling as a single system.
LambdaGB300 NVL72AI Factory of more than 100 MW in Kansas City, planned with more than 10,000 GPUs — energy and data center turned into sellable capacity.
Together AI / 5CB200, GB200 and GB300AI Factory in Maryland with B200 and a new facility in Memphis with GB200/GB300 for AI-native applications.
Global AI128 GB300 NVL72 racksMore than 9,000 GPUs in New York: the rack stops being an isolated unit and becomes a repeatable block of an AI Factory.

What these cases teach: the goal is not simply owning GPUs. Value appears when compute, NVLink, scale-out networking, storage, software, power and cooling are treated as a single factory capable of delivering tokens, training, simulations and services with operational predictability.

Why Vera Rubin is the next step

In 2026, NVIDIA reported that the Vera Rubin NVL72 entered production ramp, with racks operating at partners such as CoreWeave, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure and Nebius. Microsoft announced Rubin for future Fairwater sites, while CoreWeave began integrating Rubin systems for training, inference and agentic AI workloads.

Product view of the Vera Rubin NVL72 rack highlighting the density of the stacked modules
Product view of the Vera Rubin NVL72 rack, highlighting module density. Credit: NVIDIA.

Executive message: Vera Rubin should be understood as strategic infrastructure for the AI Factory and HPC — high CAPEX and high density, but able to concentrate in a few racks a capacity that conventional servers cannot match.

Systemic architecture of the AI Factory

Vera Rubin is a rack- and POD-scale AI infrastructure architecture designed so that computing, memory, communication, networking, storage, security, power and cooling function as a coordinated system. In current models, performance does not depend only on FLOPS: it is necessary to move weights, activations and KV cache, synchronize accelerators, maintain low latency and deliver more tokens per watt.

The strategic shift is clear: from an isolated server to a rack-scale execution domain; from peak FLOPS to useful tokens per watt and per megawatt; from pure training to training, post-training and agentic inference; from peripheral interconnect to memory and communication as part of the accelerator.

The platform brings together Rubin GPU, Vera CPU, NVLink 6 Switch, ConnectX-9 SuperNIC, BlueField-4 DPU and new networking families and specialized racks — compute, CPU, storage and Ethernet. The NVL72 is the compute core of this ecosystem.

The Rubin GPU in detail

The Rubin GPU is oriented toward reasoning, generation, data retrieval and tool-use workloads. The combination of HBM4, Transformer Engine and NVLink 6 aims to keep accelerators busy even in workloads with intensive communication and multiple inference steps.

ItemRubin GPUVera Rubin NVL72
Transistors336 billion
SMs / Tensor Cores224 / 89672 GPUs
NVFP4 inference50 PFLOPS3,600 PFLOPS
Memory288 GB HBM420.7 TB HBM4
HBM4 bandwidth22 TB/s1,580 TB/s aggregate
InterconnectNVLink 6: 3.6 TB/s260 TB/s per rack

Correct interpretation: the values above are specifications and peaks published for the platform. Application performance depends on framework, kernel, precision, model, batch, context and parallelism strategy.

In inference, especially during the decode phase, moving weights and model state can be just as decisive as running multiplications. That is why Vera Rubin combines memory capacity, large bandwidth and a low-latency communication fabric.

  • Up to 288 GB of HBM4 per Rubin GPU and up to 22 TB/s of memory bandwidth per GPU.
  • More room for weights, activations, KV cache and concurrency, with less need to offload state to slower tiers.
  • Up to 3.6 TB/s of scale-up per GPU for the NVLink domain, with all-to-all communication among the GPUs in the NVL72.
  • High-bandwidth NVLink-C2C for coherent CPU-GPU communication.
  • Reduced synchronization overhead and less time lost in transfers for MoE and distributed workloads.

Why this matters: Mixture-of-Experts models distribute tokens across experts. The more time the rack spends transferring weights and synchronizing GPUs, the smaller the fraction of energy converted into useful work. HBM4 and NVLink 6 directly attack this bottleneck.

Training, inference and agentic AI

An agent does not just answer a question: it plans, calls tools, consults sources, verifies results, and can generate several token sequences before completing a task. This increases the inference volume and makes per-step latency critical.

TrainingInference and agentic AI
Large-scale MoE and foundation models.Interactive services with low latency.
Tensor, pipeline and expert parallelism.Long-context and large KV caches.
High communication for synchronization and token routing.High concurrency and continuous batching.
HBM4 to keep parameters and states close to compute.Optimization for tokens/s, tokens/W and cost per million tokens.

In specific scenarios, NVIDIA positions Vera Rubin to deliver significant gains in tokens per megawatt and cost per million tokens compared to previous generations. Treat these figures as platform reference values, conditioned on model, software, precision and configuration.

Networking, storage and security

The platform does not end at the GPU. To keep 72 accelerators busy, east-west traffic, storage access and workload protection need to be treated as parts of the computer.

  • Networking: ConnectX-9 SuperNIC for ultra-high-bandwidth scale-out; Quantum-X800 InfiniBand and Spectrum-X Ethernet for multi-rack clusters; RDMA and low-latency paths; spine-leaf architectures and high-density optics sized alongside the compute.
  • Contextual storage: the BlueField-4/STX architecture extends the role of storage beyond a file repository, bringing context and KV cache layers closer to the execution domain — combined, in real projects, with NVMe, parallel file systems, object storage and tiering policies.
  • Security: BlueField-4 for network processing, offload and policies; Confidential Computing and protection of data in use, in transit and at rest as supported by the platform; tenant isolation in cloud and GPUaaS; secure boot, signed firmware, observability and rack-scale RAS.

Vera Rubin versus Blackwell

The comparison should be made at the system level. Rubin is not just a GPU update: it changes how the data center delivers computing for AI, with new demands on memory, networking, power and cooling.

CriterionBlackwell / GB200Vera Rubin / NVL72
FocusTraining and inference at scaleAgentic AI, long-context and efficiency
MemoryHBM3e, depending on modelHBM4, up to 288 GB/GPU
InterconnectNVLink 5NVLink 6
Reference rackGB200 NVL72Vera Rubin NVL72
CoolingDirect liquid coolingDirect liquid cooling, integrated rack
EconomicsHigh throughputFocus on tokens per watt and per MW

Technical view of the rack

The NVIDIA Vera Rubin NVL72 is a third-generation MGX rack-scale system. The rack integrates 72 Rubin GPUs and 36 Vera CPUs, interconnected via NVLink 6, with ConnectX-9 and BlueField-4 networking. The goal is to make the rack operate as one large single accelerator within the AI Factory.

ParameterReference
GPUs72 × NVIDIA Rubin
CPUs36 × NVIDIA Vera
Total HBM420.7 TB
HBM4 bandwidthup to 1,580 TB/s
CPU memory54 TB LPDDR5X
Total DGX fast memory75 TB
NVLink6th generation
NVLink switch system9 × L1 NVLink Switch Units
Scale-outConnectX-9 / BlueField-4
Facilities Cabinet TDPup to 330 kW — DSX reference

Physical structure: 18 compute trays and 9 NVLink 6 switch trays. Each compute tray integrates 2 Vera Rubin Superchips, totaling 4 Rubin GPUs and 2 Vera CPUs. The tray design is modular, cable-free, hose-free and fanless, with a redesigned internal manifold. The third-generation MGX rack uses UQD08 rack manifolds and liquid-cooled busbars for high currents.

Specification of each compute tray

The compute tray is the service and calculation unit repeated 18 times in the rack. Each tray concentrates compute, networking, cooling, management and power delivery.

ItemPer compute tray
Vera Rubin Superchips2
Rubin GPUs4
Vera CPUs2
Total HBM41,152 GB (4 × 288 GB)
HBM4 bandwidth88 TB/s aggregate
CPU cores176 Olympus cores
CPU memoryup to 3 TB LPDDR5X
NVFP4 inference200 PFLOPS per tray
NVLink 614.4 TB/s aggregate
NetworkingConnectX-9 + BlueField-4
CoolingDirect Liquid Cooling, fanless tray

Attention to power per blade: the public documentation details tray capacity and architecture, but the definitive electrical Max Power/TDP per compute tray must be obtained from the final Site Planning/Reference Design. Do not derive the electrical circuit simply by dividing 330 kW by 18.

Detail of modules, network cables and indicator lights in a validation rack
Real detail of modules, cables and indicators in a validation environment. Credit: editorial record reproduced in the EnQ Digital e-book.

Facilities executive summary

The Vera Rubin NVL72 represents a change in scale in the design of physical infrastructure for AI. For facilities, the NVIDIA DSX reference evolves to a Cabinet TDP of up to 330 kW per rack and to a high-temperature liquid cooling strategy.

Base parameter for pre-designReference value
DSX Cabinet TDP330 kW per rack
TCS design flow≥ 1.5 LPM/kW
Reference flow per rack≥ 495 LPM (29.7 m³/h)
CDU typeLiquid-to-liquid
CDU redundancyN+1
Cooling design point45 °C
Infographic of the Vera Rubin NVL72 liquid cooling solution, from the dry cooler to the rack, with reference parameters and CDU manufacturers
Overview of the liquid cooling architecture: dry cooler/chiller, FWS, CDU, TCS supply and return. Consolidated technical illustration — values must be confirmed in the final NVIDIA/OEM project.

DLC architecture: FWS, CDU, TCS and rack

The solution uses two hydraulic circuits separated by a heat exchanger. Understanding this separation is what avoids the project's most costly mistakes.

DRY COOLER / CHILLER heat rejection to the atmosphere FWS — primary circuit (~41 °C supply) CDU LIQUID-TO-LIQUID heat exchanger, pumps, control and N+1 TCS supply — secondary circuit (~45 °C) VERA RUBIN NVL72 — 330 kW cold plates, UQD08 manifolds, busbar cooling minimum flow ≈ 495 LPM (1.5 LPM/kW) TCS return typical 55–65 °C HEAT PATH: CHIP → COLD PLATE → TCS → CDU → FWS → ATMOSPHERE each link requires its own capacity, flow, pressure and monitoring

What FWS, CDU and TCS are

FWS — Facility Water System. This is the data center's primary circuit. It carries heat between the CDUs and the heat rejection plant — typically chillers, dry coolers, pumps and heat exchangers in the Central Utility Building. FWS must not be confused with the technical fluid that circulates inside the racks. It typically uses industrial water or chilled water, with temperatures in the range of 20–32 °C depending on the project, and requires chemical treatment and filtration.

CDU — Coolant Distribution Unit. This is the thermal and hydraulic interface between the FWS and the TCS. In a liquid-to-liquid CDU, a heat exchanger keeps the two circuits separated. The CDU controls temperature, flow, pressure and differential pressure, and normally includes redundant pumps, filtration, sensors, controls and BMS integration. It is available in capacities from 380 kW to more than 3 MW depending on the manufacturer.

TCS — Technology Cooling System. This is the secondary, technical-quality circuit that leaves the CDU and directly feeds the manifolds and cold plates of the IT equipment. For the Vera Rubin project, the TCS must be sized based on the rack's thermal and hydraulic operating point, not just the CDU's nominal kW capacity.

Rule of thumb: FWS = facility water | CDU = separation and control | TCS = technical fluid reaching the rack | DLC = heat removal directly at the chip.

Diagram of the liquid cooling architecture separating the primary FWS circuit and the secondary TCS circuit through the CDU
Two separate loops: Facility Water System (primary) and Technology Cooling System (secondary), with the CDU as the interface. Technical illustration.

45 °C cooling: why this changes the data center

NVIDIA describes MGX racks as designed to operate with a 45 °C warm-water inlet. In a reference scenario, facility water on the order of 41 °C feeds the CDU, which supplies approximately 45 °C to the rack. The goal is to widen the heat rejection window without continuously relying on mechanical compression.

  • Greater possibility of dry cooling and free cooling in various climates.
  • Lower compressor consumption and potential PUE reduction.
  • More of the site's electrical budget can be directed to compute.
  • The liquid-to-liquid architecture allows IT fluid quality to be separated from facility water.
  • 45 °C is a design point: final temperatures, approach and limits must come from the contracted design.

Do not confuse: a 45 °C TCS supply does not mean the Facility Water necessarily enters the CDU at 45 °C. The CDU requires a thermal difference (approach), and the FWS/TCS design must be verified with the manufacturer.

Thermal and hydraulic sizing

DSX states that the Cabinet TDP scales up to 330 kW for the Vera Rubin NVL72. For cooling pre-design, this is the most consistent reference value for sizing heat rejection, CDU and distribution, until the final Site Planning indicates specific MaxP/MaxQ values. DSX further defines a TCS design flow of at least 1.5 LPM/kW.

Per-rack calculation: 330 kW × 1.5 LPM/kW = 495 LPM = 29.7 m³/h per Vera Rubin NVL72.

RacksThermal loadMinimum TCSFlow
1330 kW495 LPM29.7 m³/h
2660 kW990 LPM59.4 m³/h
41.32 MW1,980 LPM118.8 m³/h
82.64 MW3,960 LPM237.6 m³/h
165.28 MW7,920 LPM475.2 m³/h

ΔT: turning kW and flow into temperatures

The thermal balance of a water circuit can be approximated by Q = ṁ × Cp × ΔT. In water, 495 LPM is equivalent to approximately 8.25 kg/s. For 330 kW, the theoretical ΔT is on the order of 9.6 °C.

Illustrative example: if the TCS enters at 45 °C and removes 330 kW at about 495 LPM with properties close to those of water, the theoretical return would be close to 54–55 °C. This value is merely indicative; the coolant and final temperatures must follow the OEM and the CDU.

  • A higher ΔT can reduce the flow required for the same load, but must respect the rack's limits.
  • Higher glycol concentration alters density, viscosity and thermal capacity.
  • A CDU's published capacity depends on the approach temperature and the operating point.
  • Use the manufacturer's curves, not just the nominal catalog power.

How to interpret ΔP and the Pump Head

ΔP reads "Delta P" and means pressure difference. In a hydraulic circuit, it represents the pressure loss between two points. If the rack receives 3.0 bar and returns 2.2 bar, the rack's ΔP is 0.8 bar.

Total Pump Head: ΔP CDU + ΔP piping + ΔP main manifold + ΔP rack drop + ΔP valves + ΔP hose/flexible + ΔP Rubin rack + design margin.

The CDU pump must supply sufficient flow at the required Pump Head. A CDU rated at many megawatts may not meet the project if its pump curve does not deliver the necessary flow at the circuit's actual ΔP.

  • Obtain the rack's ΔP from the OEM at nominal and maximum flow.
  • Calculate the head losses of the supply and return headers.
  • Add valves, filters, fittings, flexibles, manifolds and accessories.
  • Apply a consistent margin for tolerances, aging, fouling and expansion.
  • Check NPSH, cavitation and parallel pump operation where applicable.
Diagram of the CDU Pump Head adding the pressure losses of CDU, piping, manifold, valves, hoses and rack
Elements that make up the total pressure loss of the secondary loop and the design margin. Technical illustration.

Manifolds, cold plates and the rack interface

In the third-generation MGX, NVIDIA describes new internal tray manifolds, UQD08 rack manifolds, and liquid-cooled busbars capable of supporting very high currents. The compute tray was redesigned into a cable-free, hose-free and fanless assembly.

  • The supply manifold distributes coolant across the trays; the return manifold collects the heated fluid.
  • Universal Quick Disconnects (UQD08) facilitate connection and maintenance.
  • Cold plates transfer heat directly from the GPU, CPU and other components to the fluid.
  • The rack's external interface — diameter, connection standard, working pressure and position — must come from the OEM's mechanical drawing.

Mandatory project information: do not assume that the internal UQD08 connector is the same as the rack's external connector. Request the Rack Supply/Return Interface Drawing with the number of connections, diameter, type, height, orientation, pressure and insulation requirements.

How to select the CDU

Selection should not be based solely on "CDU MW." It is necessary to simultaneously validate thermal capacity, flow, available pressure, approach, fluid, materials, redundancy, power supply and controls.

CriterionWhat to validate
Thermal capacity≥ rack load at the defined operating point and ATD
Secondary flow≥ 1.5 LPM/kW as a DSX reference
Pump HeadSufficient for the calculated total ΔP
TypeLiquid-to-liquid per DSX reference
RedundancyN+1 in CDU groups
PumpsRedundancy, failover and validated curve
Fluid and materialsChemical compatibility and wetted materials
FiltrationRating compatible with the OEM
ControlsConstant flow or constant ΔP depending on strategy
IntegrationBMS, DSX Exchange and Mission Control
MaintenanceIsolation, bypass and swap without undue impact

Example: for 1 rack, that is 330 kW and ≥ 495 LPM. A nominal 380 kW / 600 LPM CDU may seem sufficient, but the decision is only valid if it delivers those values at the project's actual ΔP and approach.

DSX Marketplace: MP Ready vs Sample Ready

The NVIDIA DSX Infrastructure Marketplace publishes CDUs submitted to the CDU Self-Qualification Suite and indicates supply chain status. MP Ready is not, by itself, synonymous with qualification: it is an indicator of readiness for mass production, while the Validation Type column shows the technical tests performed.

StatusPractical meaning
MP ReadyMass Production Ready: product at a scale-supply stage.
Sample ReadyProduct available at a sample/qualification stage, without the same mass production status yet.
Validation TypeLists the tests performed: hydraulic constant flow/DP, thermal capacity, pump failover, flow accuracy, group control and others.

Published tests include Hydraulic Test (Constant Flow and Constant DP), Pumping Capacity, Thermal Test (Nominal Capacity and Low Load), Flow Sensor Accuracy, Cold Start Test, Pump Failover, Group Control and Wetted Materials Compatibility.

CDU manufacturers in the DSX ecosystem

The public list is dynamic. The values below reflect the query performed on 09/04/2026 and should be reconfirmed before purchase.

ManufacturerModelCapacity @ 4 °C ATDFlow @ 35 psiStatus
AVCCDU1000-LTL-RW1.2 MW1,600 LPMSample Ready
BoydROL23001.1 MW2,600 LPMMP Ready
Carrier65LL1.2 MW2,500 LPMSample Ready
CoolITCHx15001.5 MW1,950 LPM
DeltaRDF106CDT51921.0 MW1,500 LPMMP Ready
DeltaCDU30002.0 MW3,200 LPM
Johnson ControlsSACDU-10501.0 MW
LG ElectronicsLGE600 kW850 LPM
LiquidStackL2L CDU800800 kW1,200 LPM
LiquidStackD1PM202.5 MW3,750 LPM
LITEONLC-LL-WCDU-6011(S)380 kW600 LPMSample Ready
MEPPIME-CDU 12001.25 MW
MotivairMCDU501.7 MW1,136 LPMMP Ready
MotivairMCDU551.3 MW1,616 LPMSample Ready
NautilusEcoCore FCD3.6 MW3,300 LPM

Procurement strategy in Brazil

For deployment in Brazil, it is advisable to separate three criteria: the manufacturer's local presence, availability of the exact model, and the model's validation status within the NVIDIA ecosystem. A company may have a Brazilian operation without keeping the specific Rubin project CDU in local stock.

Manufacturer / ecosystemPublic evidence in BrazilNote
Schneider Electric / MotivairBrazilian portfolio of liquid cooling and CDUsPublishes Motivair CDUs and local services; validate the specific DSX model.
Delta ElectronicsGoCool LTL on the Brazilian siteGoCool-1000/1200/1500/3000; validate equivalence and the exact model's status in DSX.
Johnson Controls / Silent-AireBrazilian data center and liquid cooling pageGood service presence; validate SACDU, model and lead time.
CarrierRegional/global presence and CDU 65LLConfirm local availability and supply chain status.
LGLocal operationConfirm CDU model, support and BOM for Rubin.

Purchasing recommendation: issue an RFI/RFQ to at least three suppliers and require pump curves, thermal maps by ATD, wetted materials, coolant quality, redundancy, BMS integration, FAT/SAT, spare parts and a support commitment in Brazil.

Electrical power, kVA and consumption

DSX publishes 330 kW as the Cabinet TDP. For an approximation of electrical infrastructure, kVA = kW / power factor. Final sizing, however, must use the electrical site planning and the rack's power supply limits, not just the thermal TDP.

Power factorCalculationEquivalence
PF 0.95330 kW / 0.95347.4 kVA
PF 0.98330 kW / 0.98336.7 kVA
PF 0.99330 kW / 0.99333.3 kVA

NVIDIA does not publish a universal average consumption figure for Rubin. Training, inference, communication and MaxP/MaxQ profiles produce different loads. Use utilization scenarios only for energy budgeting, never to reduce installed physical capacity.

ScenarioAverage powerEnergy over 730 h
50% of envelope165 kW120.5 MWh/month
70%231 kW168.6 MWh/month
80%264 kW192.7 MWh/month
90%297 kW216.8 MWh/month
100%330 kW240.9 MWh/month

Integrated requirements for the data center

DomainPoints to validate
SpaceU occupied, weight, access, movement path, floor and maintenance area.
PowerkW/kVA per rack, A/B architecture, UPS, PDU, protection, grounding, harmonics and expansion.
ThermalCDU, flow, ΔP, supply/return, temperature, fluid quality and leak detection.
NetworkInfiniBand/Ethernet, fiber, transceivers, latency, cabling and spine-leaf.
StorageNVMe, parallel file system, object storage, throughput and metadata.
OperationsNOC, DCIM/BMS, telemetry, spare parts, training, RMA and emergency procedures.

Receiving checklist: confirm model and revision; obtain dimensional and weight drawing; request Max-P/TDP, nominal, minimum and maximum flow and ΔP; validate supply/return and CDU; perform FAT/SAT, hydraulic test, load test and network validation.

BMS, Mission Control and observability

Cooling for an AI Factory needs to be observable in real time. The DSX documentation publishes a catalog of points for integration between the BMS and the NVIDIA/DSX Exchange ecosystem.

PointOperational use
CDULiquidSupplyTemperatureTCS supply temperature
CDULiquidReturnTemperatureTCS return temperature
CDULiquidDifferentialPressureΔP of the secondary circuit
CDULiquidFlowFlow in LPM
CDULiquidSystemPressureSystem and return pressure
Leak detection / alarmsLeak events and abnormalities
CDU availability/statusAvailability and operational state
  • Alarms for out-of-range temperature, flow, ΔP and pressure.
  • Historical trending to detect fouling, saturated filters and pump degradation.
  • Interlock and isolation logic per cause-and-effect matrix.
  • Integration of facility events with Mission Control/DSX for coordinated power and cooling operation.

Technical RFI: 12 mandatory pieces of information

Before the executive design, these twelve pieces of information must come from the manufacturer or the representative — not from an estimate.

#InformationWhy it is indispensable
1Exact quantity of Vera Rubin NVL72Defines total scale, power, cooling, flow, CDU, redundancy and expansion.
2Confirmed Max-P/TDP per rackBasis for power and thermal load, within the contracted operational profile.
3TCS design flow per rackNominal flow to remove the load and size pumps, headers and CDU.
4TCS minimum/maximum flowOperational limits against underflow, excess flow and operation outside the envelope.
5TCS supply/return temperaturesDetermines ΔT, CDU approach, thermal efficiency and rejection capacity.
6Rack ΔP at nominal and maximum flowNeeded to calculate Pump Head and select pumps and valves.
7Working pressure and maximum allowable pressureDefines circuit safety limits, hoses, valves and fittings.
8Coolant / water-quality specificationSpecifies fluid, chemistry, pH, conductivity, filtration and compatible materials.
9Rack Supply/Return connection type and dimensionsDefines diameter, type, position, quantity and standard of hydraulic interfaces.
10CDU approved/validated modelsAligns selection with the DSX Marketplace and OEM configuration.
11BMS/Mission Control telemetry requirementsDefines sensors, protocols, alarms and IT/OT integration points.
12Final Site Planning Guide / Reference DesignAuthoritative document for validating the executive design before deployment.

Minimum package expected in the response: a filled-in table with nominal, minimum and maximum values and their units; hydraulic curves for the rack and the CDU at the relevant operating points; mechanical drawings of the supply/return interfaces; chemical specification of the coolant and wetted materials; a list of alarms, sensors, protocols and BMS/DSX integration points; and final Site Planning, FAT/SAT documents and acceptance criteria.

Pre-sizing example with 8 racks

This example is for initial sizing only. The executive design must finalize ΔP, pump curves, hydraulic topology, diversity, redundancy, expansion and the electrical envelope.

ParameterPre-design value
Vera Rubin racks8
Rubin GPUs576
Vera CPUs288
Aggregate Cabinet TDP2.64 MW
Minimum TCS design flow3,960 LPM
Equivalent flow237.6 m³/h
Cooling architectureLiquid-to-liquid CDU group
CDU redundancyN+1
TCS supply design point45 °C class
BMSSupply/return temp, flow, ΔP, pressure, alarms

CDU selection: do not simply use 2.64 MW divided by nominal capacity. Also verify whether the CDU group delivers ≥ 3,960 LPM at the total Pump Head and thermal approach of the project, while maintaining N+1. For larger clusters, distribution should be designed as a high-capacity hydraulic plant, with main headers, per-rack and per-branch isolation, balancing, instrumentation and a concurrent maintenance strategy.

Commissioning and testing

A high-density DLC system must be commissioned as mission-critical infrastructure. The CDU FAT and the system SAT must validate not only thermal power, but also hydraulic and control behavior.

  • Flushing and cleaning of piping before connection to the rack.
  • Chemical analysis of the coolant and verification of compatible materials.
  • Hydrostatic test according to allowable pressure and manufacturer procedure.
  • Flow verification per branch and rack balancing.
  • ΔP test at nominal and maximum load.
  • Pump failover test and A/B power loss.
  • Isolation and bypass valve test.
  • Leak detection and cause/effect in the BMS.
  • Validation of temperature, pressure and flow sensors.
  • FWS loss test, recovery and restart sequence.
  • N+1 operation test with one CDU unavailable.
  • Baseline recording for operation and predictive maintenance.

Acceptance criteria: delivery must include as-built curves, control parameters, setpoints, alarm matrix, water-quality baseline, list of critical spare parts and a leak response procedure.

Main project risks

RiskImpact
CDU chosen only by MWMay lack flow or Pump Head at the actual operating point.
Unknown rack ΔPPrevents correctly sizing the pumps.
Assumed rack connectorCan cause mechanical incompatibility and rework.
Undefined water qualityRisk of corrosion, fouling and cold plate failure.
No real N+1CDU maintenance or failure can affect multiple racks.
No per-rack isolationIncreases the blast radius of maintenance and leaks.
No adequate BMS/telemetryLoss of early detection of degradation.
Using average consumption for sizingCreates risk of insufficiency at MaxP.
Ignoring ATD/approachThe CDU's nominal capacity may not exist at the project's temperatures.
Not validating local availabilityRisk of insufficient lead time, spare parts and support in Brazil.

Quick glossary

TermDefinition
DLCDirect Liquid Cooling.
FWSFacility Water System — the facility's primary circuit.
TCSTechnology Cooling System — the secondary circuit serving IT equipment.
CDUCoolant Distribution Unit — the FWS/TCS interface with heat exchanger, pumps and controls.
ΔPDelta P — pressure difference or loss between two points.
ΔTDelta T — temperature difference between supply and return.
Pump HeadPressure or manometric head that the pump must supply to overcome circuit losses.
Cold plateHeat exchanger in thermal contact with the chip or component.
ManifoldCollector and distributor of supply and return.
UQD08Universal Quick Disconnect used in the third-generation MGX architecture.
ATDApproach Temperature Difference of the heat exchanger or CDU.
MP ReadyMass Production Ready — supply chain status on the NVIDIA DSX Marketplace.
Sample ReadySample and qualification stage on the Marketplace.
BMSBuilding Management System.
DSXNVIDIA architecture and ecosystem for the AI Factory, including facilities, hardware and IT/OT.

Documents that should be requested

  • NVIDIA Vera Rubin NVL72 Reference Design — NVOnline #1151654.
  • NVIDIA DSX — Vera Rubin Facilities Infrastructure Reference Design — NVOnline #1145739.
  • NVIDIA DSX Facilities Infrastructure Design Guide — NVOnline #1152370.
  • Final Site Planning / Power & Cooling Specification for the acquired configuration.
  • Rack Supply/Return Interface Drawing and coolant/water-quality specification.

Cut-off date: technical content and verified supplier list as of 09/04/2026. The NVIDIA DSX Marketplace is dynamic; revalidate status, capacity, flow and availability at the time of contracting.

Technical references

Conclusion

The main innovation of Vera Rubin is systemic. The value is not in an isolated GPU, but in the combination of compute, HBM4, NVLink 6, scale-out networking, storage, security, power and Direct Liquid Cooling operating as a single computing factory.

For companies and data centers, the opportunity lies in transforming energy and physical infrastructure into useful AI capacity: model training, agentic inference, GPUaaS, scientific research, simulation and sovereign services. The challenge is also systemic — electrical, hydraulic, connectivity, software, observability, operations and support all need to evolve at the same pace as the accelerator.

The right question: instead of asking only how many GPUs fit in the rack, the project must answer how many useful tokens, completed agents, simulations and business workloads the data center can produce per megawatt — with availability, security and operational predictability.

  • Consolidate the actual demand for training, inference and HPC.
  • Confirm the Max-P/TDP and the TCS design flow of the contracted configuration.
  • Calculate the total Pump Head before choosing the CDU.
  • Issue an RFI to at least three suppliers with the minimum data package.
  • Define commissioning criteria before contracting.
  • Plan telemetry, operations and maintenance from the outset.

EnQ Digital: intelligent infrastructure for companies that cannot stop — strategy, engineering, deployment, operation and evolution driven by risk and performance.

Chegou ao final da matéria? Baixe o material completo em PDF.

E-book: Infraestrutura de Refrigeração Líquida para NVIDIA Vera Rubin NVL72