Modular
Modular offers the MAX platform, which includes tools for deploying AI inference pipelines, and hosts the annual ModCon developer conference.
Platform Offerings
Modular's flagship platform, MAX, includes a suite of tools and libraries designed for deploying low-latency, high-throughput AI inference pipelines. The components of MAX are MAX Engine, MAX Serving, and the Mojo programming language. Each of these tools is tailored to ensure seamless AI model execution and deployment, irrespective of the underlying hardware or container infrastructure.
MAX Engine
MAX Engine is a model inference runtime and API library provided by Modular. It enables the execution of AI models on any hardware without the need for rewriting or conversions. This capability is crucial for developers who require flexibility in deploying AI models across diverse environments.
MAX Serving
MAX Serving is a model serving library that integrates with existing serving systems like NVIDIA Triton. It is designed to be easily deployable within container infrastructures such as Kubernetes. This integration allows for the smooth deployment and management of AI models in scalable, production-ready environments.
Mojo Programming Language
Mojo is a programming language developed by Modular, combining the usability of Python with the performance of C. This design enhances the programmability and extensibility of AI models and hardware, making it easier for developers to build and optimize AI applications.
Developer Support and Community Engagement
Modular is committed to supporting developers through extensive documentation and tutorials available on their website. They also host an annual developer conference called ModCon, which features industry experts and insights on AI development and deployment. The company actively engages with the developer community via blogs, detailed documentation, and participation in discussions on platforms like GitHub and Discord.