What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Del Complex proposed a floating AI data center called the BlueSea Frontier Compute Cluster (BSFCC), designed to carry more than 10,000 Nvidia H100 GPUs and operate in international waters. The company described it as solar-powered and water-cooled, with an offshore location intended to reduce exposure to national AI rules. The available reporting supports treating BSFCC as a proposal—not a built, funded or operating data center.
What did Del Complex propose?
Del Complex described the BSFCC as a floating, barge-based facility for training and running AI models. Its stated design called for more than 10,000 H100 GPUs, water cooling and solar power. TechRadar also reported a $500 million estimate for the GPUs alone; that is a reported estimate, not an audited purchase price or the total cost of the project. TechRadar’s report describes the announced configuration and the company’s broader ambitions.
Del Complex linked the offshore location to two aims: access to seawater and distance from land-based constraints, and reduced exposure to national AI regulation. The second aim is a claim about the project’s purpose, not proof that a vessel offshore would be outside national law.
Is the project operational?
There is no evidence in the available reporting that the BSFCC has been built or launched, that Del Complex bought the GPUs, secured financing or a vessel, or started providing compute. Tom’s Hardware questioned whether Del Complex had the conventional operating capacity implied by its offering and treated the concept as speculative. That supports describing it as an announced proposal, not calling the company fake or claiming no activity could exist beyond public reporting. Tom’s Hardware’s coverage discusses the company’s public footprint and regulatory pitch.
Recommended Free Tools
#1 Best Overall
- PLEASE NOTE: Exporting an NVIDIA RTX Pro 6000 GPU outside the US requires strict adherence to the U.S. Export Administration Regulations (EAR) and issuance of an export license from the Bureau of Industry and Security (BIS). Compliance and Know Your Customer (KYC) screening may be required as a condition of order acceptance. [NVIDIA Blackwell Streaming Multiprocessor] The new SM features increased processing throughput, and new neural shaders that integrate neural networks inside of programmable shaders | DLSS 4: Multi Frame Generation ensures ultra-smooth frame pacing for lifelike simulations.
- [Double-Flow-Through Design] The RTX PRO 6000 Blackwell features a double-flow-through cooling design, optimizing efficiency and airflow to sustain peak performance under 600W power loads. | [5th Gen Tensor Cores] Deliver up to 3X the performance of the previous generation and support for FP4 precision for faster AI model processing times with reduced memory usage, enabling local fine-tuning of LLMs and generative AI | [4th Gen Ray Tracing Cores] Double the ray-triangle intersection rate of the previous generation to create photoreal, physically accurate scenes and immersive 3D designs with RTX Mega Geometry, which enables up to 100X more ray-traced triangles.
- [PCIe Gen 5] Support for PCIe Gen 5 provides double the bandwidth of PCIe Gen 4, improving data-transfer speeds from CPU memory and unlocking faster performance for data-intensive tasks like AI, data science, and 3D modeling. | [GDDR7 Memory] With 96 GB of GPU memory and 1.8 TB ps bandwidth, it can tackle massive 3D and AI projects, fine-tune AI models locally, explore large-scale VR environments, and drive larger multi-app workflows.
- [DisplayPort 2.1] Achieve unparalleled visual clarity and performance, driving high resolution displays at up to 8K at 240 Hz and 16K at 60 Hz. Increased bandwidth enables seamless multi-monitor setups while HDR and higher color depth support ensures superior color accuracy for precision work, such as video editing, 3D design, and live broadcasting.
- [Universal MIG] Divide a single RTX PRO 6000 Blackwell into multiple isolated instances, each with dedicated resources, allowing for concurrent execution of multiple workloads, optimized GPU utilization, and secure isolation of different applications or users. [WARRANTY] 3 YR Manufacturer's Warranty. Bulk OEM Packaging. Retail Packaging is NOT included.
To assess whether the proposal had become a real infrastructure project, readers would need evidence such as identified financing, hardware procurement, a platform contract, permits or maritime registrations, detailed engineering, connectivity arrangements and customer commitments. None is established in the reporting cited here.
Could 10,000 H100s work as a computing cluster?
At that scale, AI training is technically possible in conventional data-center environments. Meta’s MegaScale paper examines the engineering of training large language models on clusters exceeding 10,000 GPUs; it is evidence that the computing problem is tractable, not evidence that Del Complex built its system. Read the MegaScale paper.
The GPUs would be only one part of the facility. A working cluster would also need servers, CPUs and memory, storage, high-speed networking, power distribution, cooling, fire suppression, backup systems, security, connectivity, maintenance and staff. Putting those systems on a floating platform adds marine construction and operating requirements.
Power: 7 MW is only a GPU-load illustration
As an illustrative calculation, 10,000 GPUs at an assumed 700 watts each would draw 7 megawatts for the GPU boards alone. That is not a BSFCC specification. It excludes the rest of the servers, networking, storage, electrical conversion losses, pumps, cooling, lighting and redundancy, so a facility would need materially more continuous electrical capacity.
Rank #2
- NVIDIA Ampere Architecture-based CUDA Cores - Double-speed processing for single-precision floating point (FP32) operations and improved power efficiency provide significant performance improvements for graphics and simulation workflows, such as complex 3D computer-aided design (CAD) and computer-aided engineering (CAE), on the desktop.
- Second-Generation RT Cores - With up to 2X the throughput over the previous generation and the ability to concurrently run ray tracing with either shading or denoising capabilities, second-generation RT Cores deliver massive speedups for workloads like photorealistic rendering of movie content, architectural design evaluations, and virtual prototyping of product designs. This technology also speeds up the rendering of ray-traced motion blur for faster results with greater visual accuracy.
- Third-Generation Tensor Cores - New Tensor Float 32 (TF32) precision provides up to 5X the training throughput over the previous generation to accelerate AI and data science model training without requiring any code changes. Hardware support for structural sparsity doubles the throughput for inferencing. Tensor Cores also bring AI to graphics with capabilities like DLSS, AI denoising, and enhanced editing for select applications.
- Third-Generation NVIDIA NVLink - Increased GPU-to-GPU interconnect bandwidth provides a single scalable memory to accelerate graphics and compute workloads and tackle larger datasets.
- 48 Gigabytes (GB) of GPU Memory - Ultra-fast GDDR6 memory, scalable up to 96 GB with NVLink, gives data scientists, engineers, and creative professionals the large memory necessary to work with massive datasets and workloads like data science and simulation.
The solar-power claim cannot be assessed from the information cited here: no public energy budget, panel-area calculation, battery specification or generation model is provided. A credible design would have to address nighttime generation, storms, energy storage, peak demand, backup power and black-start capability—the ability to restart after a complete loss of power.
Cooling: seawater is a heat sink, not a complete system
Seawater could help reject heat, but the facility would still need pumps, heat exchangers, filtration, controls, redundancy and regular maintenance. A closed freshwater loop with seawater on the other side of heat exchangers is one possible approach; direct seawater circulation creates more serious corrosion, fouling and environmental concerns. Saltwater, marine growth, blocked intakes, leaks, storms and remote repairs all complicate operations. Calling ocean cooling free would ignore those costs and risks.
Connectivity: training and inference have different needs
A remote facility would need high-capacity fiber connections, ideally with redundant routes, and a shore landing station. Satellite links could support management or some workloads, but they are not an obvious replacement for the fast, low-latency networking used between GPUs in tightly coupled training. Training data and model checkpoints can also be substantial; outages, cable cuts or difficult data movement could undermine operations.
Inference and batch workloads may tolerate more distance than tightly coordinated training, but customer requirements still matter. Remote siting can create difficulties for latency, data residency, privacy and regulated workloads.
Rank #3
- NVIDIA Volta GV100 Architecture — 4,608 CUDA Cores, 640 1st-Gen Tensor Cores delivering 14 TFLOPS FP32 and 112 TFLOPS deep learning performance for AI training, inference, HPC, and scientific computing workloads
- 32GB HBM2 ECC Memory — 900 GB/s Bandwidth — High-bandwidth memory on a 4096-bit bus with ECC error correction provides the memory capacity and throughput required for the largest AI models, simulations, and datasets
- PCIe 3.0 x16 Interface — 250W TDP — Standard PCIe Gen3 connectivity with passive cooling designed for enterprise rack server deployment in HPE ProLiant, Dell PowerEdge, and Supermicro platforms with adequate chassis airflow
- NVLink — Scale to 96GB Unified Memory — Connect two V100 GPUs via NVLink at 300 GB/s bi-directional bandwidth to scale GPU memory from 32GB to 96GB for larger AI training and HPC workloads
- Multi-Precision Computing — Supports FP64 (7 TFLOPS), FP32 (14 TFLOPS), FP16 (112 TFLOPS) and INT8 precision modes for flexible deployment across training, inference, and scientific simulation workloads
Does “international waters” mean the facility is outside the law?
No. “International waters” is not a synonym for a law-free zone. A vessel is generally connected to the jurisdiction of its flag state, while the company, owners, operators, suppliers, workers and customers may remain subject to laws based on their own locations and activities. Port calls, resupply and maintenance can bring additional port-state, coastal-state, customs, environmental, labor, insurance and maritime-security requirements into play.
Moving equipment offshore does not automatically remove export-control or sanctions obligations. The precise legal outcome would depend on ownership, shipment, licensing, the vessel’s flag, the people and companies involved, and the transactions supporting the operation. Del Complex’s regulatory-avoidance rationale should therefore be read as an ambition, not a settled legal conclusion.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Could Del Complex obtain and operate the H100s?
Procurement is a central uncertainty because H100s are advanced U.S.-origin accelerators subject to export-control policy. Hardware movement is only one part of the picture: suppliers, financing, software and support, personnel, beneficial ownership, customers and onward transfers may all matter.
A public comment submitted on February 7, 2024, urged the U.S. government not to provide H100 GPUs to Del Complex and raised concerns about the proposed offshore barges. The filing shows that the issue was raised publicly; it does not establish that Del Complex obtained or was denied GPUs, or that the government reached a decision on the project. Read the public comment.
Rank #4
- Standard Memory: 40 GB
- Host Interface: PCI Express 4.0
- Cooler Type: Passive Cooler
- Product Type: Graphics Card
Would the “AI nation” become a sovereign country?
Del Complex’s messaging extended beyond infrastructure to self-governance and the idea of a sovereign or quasi-sovereign AI entity. TechRadar reported that the company invoked the Montevideo Convention and the United Nations Convention on the Law of the Sea in discussing recognition. Citing treaties does not itself create a country or establish international recognition. The sovereignty pitch is best understood as a political and legal claim attached to the concept, not a status the floating platform has demonstrated.
What would make the proposal financially credible?
The reported $500 million GPU estimate is not a project budget. A complete financial case would have to account for servers and networking, platform construction or conversion, energy generation and storage, cooling, subsea cables, crew, security, insurance, resupply, repairs, replacement hardware, financing and eventual decommissioning. The cited reporting does not establish that Del Complex had the funding or customer commitments to cover these costs.
Before treating the headline GPU count as an investable or financeable plan, look for documented capital, purchase agreements, engineering studies, power and connectivity contracts, expected utilization, customer revenue, maintenance assumptions and insurance arrangements. Without those, the number describes an aspiration rather than a demonstrated business case.
What is the practical takeaway?
The BSFCC proposal combines a technically conceivable large AI cluster with unresolved power, cooling, connectivity, maritime, legal and financing questions. Its most consequential claims—solar operation, regulatory insulation and AI sovereignty—are not established by evidence of a working facility. It is a case study in the challenges of offshore compute, not evidence that an offshore AI nation or 10,000-H100 data center exists.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
