We are moving towards the AGI of the engineering design

Our software can be used by heat exchanger manufacturers, electronics cooling manufacturers, electric vehicle industry, and data center operators.

Example Image

Analysis of Heat Exchanger design Automation Software

Compressible flow feature

The geometry, meshing, and optimization engine of the current version is capable of performing compressible flow simulation without any changes

No external CAD software

our software does not require pre-defined design and external CAD software and provides a global free form deformation optimization framework.

More effective design

Significantly more effective design can be adopted as the current version of the software is based on the geometries comprised of Bezier curves instead of straight-line components

Adaptive Design

Our software can design a cold plate with increased fin density in the presence of high heat flux and, lower fin density in lower heat flux region of the battery.

AI generated heat sink design
for NVIDIA A100 TENSOR CORE GPU

AI Heat Sink Design
Our DRL based shape optimization for heat exchanger design technology generates optimized designs that can provide up to 30% improvement in heat transfer and 60% reduction in pressure drop. Therefore it has applicability in various sectors of the market including consumer electronics, high performance computing systems at datacenters, HVAC systems for both industrial setups and residentials, automotive, aviation and aerospace, etc.
...
/ Magnative AI

Design Automation of Heat exchangers using Deep reinforcement learning

No need to external CAD software
...

our software does not require pre-defined design and external CAD software and provides a global free form deformation optimization framework.

Considering manufacturing limits
...

Our software is capable of considering manufacturing limits on the size of the fins and separating sheets of heat exchangers.

Training a surrogate model
...

our solver engine gathers data during high fidelity simulation and trains a surrogate model to accelerate the solver for optimization purpose. Interviewees of the market assessment study rated the advantages of our technology at 8.6 out of 10.