We are providing an ecosystem with tools and resources that help you
build, train and run your AI models. Store your data, analyze your data, learn from your data and build knowledge.
Data is strategic resource. We foster collaboration towards competition bringing AI experts together and sharing the data to create innovation. We are aware of data having different value and criticality and support various levels of sharing with an option to define usage policies and keep data sovereignty.
The key requirement for successful AI applications is the availability of large comprehensive datasets. Such datasets are usually not available in a single company but have to be built up in a collaborative fashion by many stakeholders.
For different types of problems and depending on your data object’s features you may need to do some transformations before starting to train your model.
We are offering tools to do:
- Data Quality Assessment (Handling missing, inconsistent or duplicate data)
- Standardizing data
- Feature Aggregation
- Feature Sampling
- Feature Encoding
- Rescaling data
AI-LAB platform supports Maplotlib, Pandas, Numpy, R and Julia.
BUILD AND TRAIN
Using your data, you can build your solution or train your model. We are continuously working on making the training easier and include in AI-LAB the most popular training algorithms. We are also implementing the visual programming with Elyra.
AI-LAB supports following well-known platforms:
With AI LAB you can deploy your solution to your target platform. We are working on special deployment toolchains for different edge devices implementing ONNX standard, that enables AI developers to use models with a variety of frameworks, tools, runtimes and compilers.
NVIDIA DGX™ Systems combined with DELL C4140
NVIDIA DGX 2 is purpose-built for complex AI challenges, integrating 16 Tensor Core GPUs for 2 petaFLOPS of AI performance.
NVIDIA A100 is the universal system for all AI workloads. It integrates 8 Tensor Core GPUs, delivering the very first 5 petaFLOPS AI system.
2 x DELL C4140 is designed for the most demanding cognitive and technical computing, with its unique interleaved GPU design and no-compromise acceleration technology.