MXMACA, introduced by MetaX, is a heterogeneous computing platform that uses a general-purpose parallel computing architecture to solve complex computing problems. It includes self-developed Instruction Set Architecture (ISA) and a parallel computing engine inside the GPU. Integrating General-Purpose computing and machine learning frameworks, it provides a highly flexible and high-performance open software platform for scientists, researchers, and industry experts in various application fields.

Flexible programming language

This heterogeneous computing platform provides an easy-to-use C-like programming language for users to write programs for the MXMACA architecture. This programming language is simple, flexible, easy to read and write, and expressive. It is also compatible with the mainstream C/C++ heterogeneous computing language, which brings easy-to-use software adaptation and efficient operator development for users.

Support for multiple open source technologies

The MXMACA heterogeneous computing platform supports multiple open source technologies, which include AI neural network frameworks (TensorFlow and PyTorch etc.), libraries (such as Blas and DNN) and Linux Kernel support. Through continuous optimization, the MXMACA computing platform can achieve higher performance and scalability, which allows users to better empower the society with AI. What's more, the MXMACA computing platform provides various system and application management tools to allow users to conduct development, verification, environment deployment and quality monitoring more efficiently and flexibly.

  • China Business
  • International Business
  • Media