Interested in Licencing Components for Your Scientific Computing Needs?

When you've got to deliver intelligent software to your customers, quickly, on budget, and bug-free, consider TERAMARK Technologies. TERAMARK Technologies markets and distributes software components to developers engaged with scientific computing, machine learning, artificial intelligence/multi-agent systems and optimization tasks. Our software components let you focus on creating business applications and customized technology, keeping you from reinventing the wheel.

Numeric Computation

MAC++ Numeric Library

Solve many problems in science and engineering with our MAC++ Numeric Library. MAC++ is a C++ class library for scientific computing. Next to matrix, vector and tensor containers, our library uses the specific characteristics of each container to optimize its performance. Our algorithms are customized to each container, providing an improved perfomance compared to more general procedures. For specific tasks, such as statistical processing, we provide algorithms to optimize e.g. for parallel computing.

Contact our R&D department for more information

Machine Learning

MAC++ Machine Learning Library

Underpin your machine learning process with our data classifiers. Our classifiers are an efficient basis to decribe observations relying on extracted features. Description scheme follows the decision theoretic approach.

Contact our R&D department for more information


Stochastic Optimal Control

MAC++ Control

Our MAC++ controller applies for closed-loop feedback processes and resolves optimal control problems with stochastic variables. MAC++ controller is unique: It learns to adapt to the underlying dynamics of stochastic processes.

Contact our R&D department for more information

Stochastic Programming


A method that allows to for decision-making under uncertainty. Use our MAC++ GEP for multi-stage batch optimization applications in manufacturing or transportation software.

Contact our R&D department for more information

Autonomic Computing

Multi-Agent Systems

We offer agent-level methods concerned with individual agents including procedures for agent reasoning and learning. Our interaction-level technologies deal with agent communication, interaction protocols and resource allocation mechanisms. Our multi-agents coordinate their actions without the need for human intervention. Apply our multi-agents to automate business processes – from autonomous commodity trading, intelligent sourcing to manufacturing scheduling or transportation optimization.

Contact our R&D department for more information

Data Compression

MAC++ Information Compression

Machine learning techniques sometimes require that data or information is being compressed. Our MAC++ information compression reduces the amount of data to processed without losing information. MAC++ lossless compression speeds up the learning process and reduces the need for computing power.

Contact our R&D department for more information

Customer Benefits

Save development time

There is no need to reinvent the wheel, spending

years of research and millions of Euro to figure

out which learning technologies may suit your

needs best.

Use components that work

Relying on proven numerical methods, learning

machines or controllers cuts your technological

risk sigificantly.

Keep what you have

Our clients use our technology components to

ramp up their existing applications and bring them

to the next level. Our components may be well

integrated to what you already have.