Computational fluid dynamics can be one of the most computationally demanding parts of engineering simulation. Large models, parameter sweeps and high-fidelity aerodynamic studies can turn CFD into a waiting game measured in hours or days.
Flexcompute Flow360 takes a different approach.
Flow360 is a cloud-native, GPU-native CFD solver developed by Flexcompute for high-performance simulation across applications including aerospace aerodynamics, automotive aerodynamics, eVTOL, drones, rotors and propellers, aeroacoustics, thermal management, hydrodynamics and supersonic flows. Flexcompute says its GPU-native architecture can deliver simulation performance up to 100 times faster than traditional CPU-based approaches for some workloads.
More importantly for engineers evaluating CFD software, Flexcompute currently offers a two-week Flow360 evaluation program with unlimited RANS simulations.
Is There a Free Trial of Flexcompute Flow360?
Sort of — but Flexcompute deliberately does not call it a free trial.
The company offers a 14-day evaluation program designed for professional engineering teams that want to test Flow360’s performance and accuracy on their own CFD problems.
During the evaluation, approved users receive:
- Two weeks of Flow360 access
- Unlimited RANS simulations
- Models up to 600 million mesh nodes
- Access to the full Flow360 solver capabilities
- Cloud-based GPU computing without requiring dedicated local GPU hardware
- Browser-based access and a Python API
The major requirement is that participants bring their own Pointwise mesh.
Flexcompute specifically describes this as a time-boxed evaluation rather than a traditional free trial, and requests are reviewed for suitability. The program is intended for commercial or research engineering teams rather than students or educational users.
Flow360 Evaluation:
Request the Flexcompute Flow360 Evaluation
Wait—What Is a Pointwise Mesh?
Flexcompute’s current Flow360 evaluation asks you to bring an existing Pointwise mesh.
If you searched for Pointwise and had trouble finding it, there is a reason: Cadence acquired Pointwise in 2021, and the software is now called Cadence Fidelity Pointwise.
Fidelity Pointwise is dedicated CFD mesh-generation software used to create structured, unstructured, hybrid, overset and high-order meshes for a variety of CFD solvers.
For the Flow360 evaluation, Flexcompute says you do not need to own Pointwise during the evaluation itself—you need to arrive with a compatible mesh that was created in Pointwise.
That distinction matters.
If you already use Pointwise/Fidelity Pointwise, the requirement may be straightforward.
If you are starting from CAD geometry and do not already have a Pointwise mesh, determine how you will create the mesh before starting your Flow360 evaluation period. Otherwise, you could lose valuable evaluation time preparing the model instead of testing Flow360.
Looking for Pointwise? See our guide to the Cadence Fidelity Pointwise free trial and CFD meshing software
What Is Flow360?
Flow360 is Flexcompute’s high-performance computational fluid dynamics solver.
Unlike traditional CFD applications that were originally developed around CPU computing and later adapted to GPUs or cloud infrastructure, Flow360 was designed around GPU computing from the beginning.
That distinction matters because CFD workloads often involve millions of cells or nodes and thousands of iterative calculations. Efficiently distributing those calculations across modern GPUs can substantially reduce simulation time.
Flexcompute provides the computing infrastructure through its cloud platform, so engineers do not need to install and maintain their own GPU cluster.
Flow360 includes both a browser interface and a Python API, allowing engineers to work interactively or automate large simulation campaigns.
What Can You Simulate With Flow360?
Flow360 is particularly oriented toward fluid-flow problems where large models, repeated simulations or extensive design exploration can create computational bottlenecks.
Applications highlighted by Flexcompute include:
- Aircraft aerodynamics
- Automotive aerodynamics
- eVTOL aircraft
- Drones and UAVs
- Rotors and propellers
- Aeroacoustics
- Supersonic flows
- Hydrodynamics
- Thermal management
This makes Flow360 especially interesting for teams that need to run many CFD cases rather than simply producing an occasional single simulation.
Why Does GPU-Native CFD Matter?
Traditional CFD can create an uncomfortable engineering tradeoff.
You may know that you should evaluate another 20 design configurations, but each simulation takes long enough that there simply isn’t time.
So instead of exploring the entire design space, engineers make educated choices about which configurations deserve simulation.
Faster computation changes that equation.
If the simulation bottleneck becomes substantially smaller, engineers can potentially:
- Evaluate more design variants
- Run larger models
- Build aerodynamic databases
- Perform parameter sweeps
- Investigate more operating conditions
- Repeat simulations earlier in the design process
- Use CFD more extensively within optimization workflows
That may ultimately be more important than simply saying that one simulation finished faster.
The real value of faster CFD may be the simulations engineers can finally afford to run.
How Fast Is Flow360?
Flexcompute advertises Flow360 as delivering performance up to 100× faster than traditional CFD tools, depending on the simulation and comparison being made.
As always with simulation performance claims, engineers should test this against their own workloads.
Solver speed can be influenced by many factors, including:
- Mesh size
- Physics
- Turbulence models
- Convergence requirements
- Boundary conditions
- Hardware
- Solver settings
- Parallelization
- Model complexity
Fortunately, this is precisely where Flow360’s evaluation program becomes useful.
Rather than deciding whether a 100× performance claim applies to your workflow based on a benchmark, you can bring one of your existing CFD models and measure the difference yourself.
What Should You Test During the Flow360 Evaluation?
Don’t spend the two-week evaluation simply reproducing a colorful contour plot.
Use a CFD case you already understand.
Ideally, choose a model previously solved using another CFD package or supported by experimental data. That lets you evaluate three separate questions:
1. Accuracy
Compare important engineering outputs such as:
- Lift
- Drag
- Pressure coefficients
- Flow separation
- Forces and moments
- Wake behavior
- Temperature
- Pressure drop
A faster solver provides little value if the engineering results aren’t trustworthy.
Flexcompute maintains a public validation library comparing Flow360 results with experimental and established reference datasets, including aerospace, automotive, marine and fundamental CFD cases.
2. Time to Solution
Measure the entire workflow, not simply solver runtime.
Consider:
- Model preparation
- Upload
- Configuration
- Solver initialization
- Convergence
- Post-processing
- Results extraction
The fastest solver in the world doesn’t automatically create the fastest engineering workflow if everything surrounding it takes longer.
3. Simulation Throughput
This may be Flow360’s most important evaluation metric.
Ask:
How many useful engineering questions can our team answer in one working day?
Try running several configurations rather than only one.
That will give you a better sense of whether GPU-native CFD changes how your engineering team works.
Bring a Pointwise Mesh
There is one important qualification to the current evaluation.
Flexcompute requires participants to bring an existing Pointwise mesh.
You do not necessarily need to own Pointwise during the evaluation, but the mesh supplied to Flow360 needs to have been created in Pointwise.
That means this is not quite the same experience as downloading a traditional CFD package and beginning with raw CAD geometry.
Teams already using Pointwise may find the evaluation particularly straightforward.
For others, the mesh requirement should be considered before requesting access.
Is Technical Support Included?
No.
Flexcompute describes the current two-week evaluation as self-service and says hands-on technical support is not included.
That’s worth knowing before beginning the clock.
Prepare your mesh and your comparison case before the evaluation starts so that you spend the two weeks evaluating Flow360 rather than assembling your test problem.
Flow360 Is Cloud-Based
Flow360 runs on Flexcompute’s cloud computing infrastructure.
For many engineering organizations, that removes one of the barriers associated with GPU-accelerated simulation: purchasing and maintaining expensive computing hardware.
Engineers can access the platform through a browser, while more automated workflows can use Flow360’s Python interface.
Cloud simulation also introduces evaluation questions of its own, however.
During your trial, consider:
- Data upload and download times
- Security requirements
- IP policies
- Collaboration
- Automation
- Integration with existing CAD/CAE systems
- Cost predictability for production workloads
Solver performance is only one part of deciding whether a cloud-native CFD platform fits your organization.
Who Should Evaluate Flow360?
The Flow360 evaluation makes the most sense for professional engineering teams already performing significant CFD work.
It may be particularly interesting if:
- CFD runtime is slowing development
- You run large aerodynamic models
- Your team performs many design iterations
- Engineers regularly queue simulations
- You need aerodynamic databases
- You’re evaluating cloud CFD
- Your organization is considering GPU-accelerated simulation
- You want to automate CFD studies using Python
- Existing computing infrastructure limits the number of simulations you can run
Organizations running occasional small CFD models may see less benefit from dramatically increased solver throughput than teams running large simulation campaigns.
Who Is Flexcompute?
Flexcompute is an engineering simulation software company founded by engineers from MIT and Stanford.
The company develops GPU-native simulation technologies intended to make high-fidelity engineering simulation faster and more scalable.
Its major simulation products include:
Flow360 for computational fluid dynamics.
Tidy3D for electromagnetic and photonic simulation.
Flexcompute has also expanded into geometry automation, aerodynamic optimization, RF simulation and AI-driven engineering workflows.
We’ll cover those products separately because Flow360 and Tidy3D address very different engineering disciplines.
Flow360 vs. Traditional CFD Software
Flow360 shouldn’t necessarily be evaluated simply as a replacement for an existing CFD solver.
A more useful question may be:
What would your engineering team do differently if CFD stopped being the bottleneck?
Traditional CFD platforms such as Ansys Fluent, Siemens Simcenter STAR-CCM+, OpenFOAM and other established solvers have enormous ecosystems, broad physics capabilities and mature engineering workflows.
Flow360’s proposition is different.
It is particularly focused on combining high-fidelity CFD with GPU-scale computational throughput.
That makes it worth evaluating alongside established CFD software rather than assuming that either traditional or GPU-native architecture is automatically better for every problem.
The Evaluation Is Especially Good for Benchmarking
One thing we particularly like about Flexcompute’s approach is that the evaluation encourages engineers to use their own simulation models.
That’s how engineering software should be evaluated.
Vendor benchmark models can demonstrate capabilities, but they cannot tell you whether a solver performs well with your geometry, meshes, physics and workflows.
A useful Flow360 evaluation therefore begins with something you already know.
Run it in your existing CFD environment.
Run it in Flow360.
Then compare:
Accuracy.
Runtime.
Cost.
Setup effort.
Automation.
Overall simulation throughput.
At the end of two weeks, you should have considerably more useful information than a generic product demo can provide.
How to Request the Flow360 Evaluation
Flexcompute currently offers qualifying professional engineering teams two weeks of Flow360 access for evaluation.
The program includes unlimited RANS simulations with models up to 600 million mesh nodes and full solver capabilities.
A Pointwise mesh is required, and evaluation requests are reviewed by Flexcompute for fit. The program excludes existing customers and is not currently intended for student or educational use.
Start here:
Request the two-week Flow360 CFD evaluation
EST Evaluation Takeaway
Trial type: Two-week professional evaluation
Cost: No-cost evaluation for approved participants
Simulation limit: Unlimited RANS simulations during the evaluation
Maximum model size: 600 million mesh nodes
Solver access: Full Flow360 solver capabilities
Installation: Cloud-based
Mesh requirement: Existing Pointwise mesh
Technical support: Self-service; hands-on support not included
Student access: Not through this evaluation program
Best for: Engineering teams benchmarking high-performance CFD and GPU-native simulation
Our take
This is a strong engineering-software evaluation offer.
It isn’t a casual browser demo, and it isn’t designed for somebody who simply wants to experiment with CFD for the first time.
But for an engineering team with an existing Pointwise model and known CFD results, two weeks of unlimited RANS simulation provides enough access to conduct a meaningful solver comparison.
That is exactly the kind of evaluation access we like to see.