dicehub.de

What It Is

dicehub is a cloud-based CFD simulation platform positioned as a “template-driven CFD platform.” It focuses on helping engineers quickly create high-fidelity fluid dynamics simulations through preconfigured templates, and is already used by 1,000+ engineers across 25+ countries. The key use cases shown on its site include pedestrian wind comfort, automotive aerodynamics, UAV aerodynamics, and cloud HPC parallel computing based on OpenFOAM.

Core Capabilities

The platform centers on template reuse and cloud scalability. Users can start from ready-made templates and adjust geometry, boundary conditions, and solver settings instead of rebuilding each case from scratch. The page explicitly mentions STL geometry import, OpenFOAM setup, comfort maps, and other outputs. For collaboration, it offers Google Docs-like real-time collaboration, allowing multiple engineers to edit the same simulation case, discuss changes, and manage access permissions. On the compute side, dicehub can scale from a single machine to 4, 16, or 64 nodes, automatically handling case decomposition, solver launch, and result aggregation. It also provides observability features such as run logs, CPU/memory metrics, and residual alerts.

Pricing and Value

dicehub uses pay-as-you-go pricing and only charges while simulations are running. The page gives examples: 18 cores at €2.83/hour, 36 cores at €5.66/hour, 96 cores at €7.41/hour, and 192 cores at €13.56/hour, with prices listed as net amounts excluding tax. For occasional or project-based CFD workloads, this model can avoid the need to purchase and maintain long-term HPC infrastructure. However, the page does not clarify whether there is a free tier, team seat pricing, storage fees, data transfer fees, or enterprise pricing, so actual costs still need to be evaluated based on project scale.

Pros and Cons

Its strengths are lowering the barrier to using OpenFOAM and HPC, while its templated workflows help improve result consistency and team review. Real-time collaboration is useful for multi-engineer iteration, and automatic parallelization, logs, and monitoring reduce the infrastructure operations burden. The limitations are that the captured content does not disclose whether it is open source, whether self-hosting is supported, or whether there are clear API/SDK options, data security details, compliance certifications, or simulation accuracy validation materials. Its “managed fleet” description mentions CUDA, GPUs, and batch processing, but the specific GPU capabilities and supported scope are not clear.

Who It’s For and Access from China

It is better suited to building wind environment teams, automotive/aerospace aerodynamics teams, and small to midsize engineering teams or consultancies that need to run OpenFOAM in the cloud. It is less suitable for organizations with strict requirements around offline deployment, domestic infrastructure environments, or data residency. Access from mainland China, payment methods, and local support are not explained on the site, so china_access can only be considered unknown. If network access or payment is restricted, alternatives include SimScale, Ansys Fluent, COMSOL, or building an OpenFOAM/HPC workflow on a domestic cloud platform in China.