A major milestone in programming language development has been reached as Mojo, a high-performance language created by Chris Lattner, now allows Python users to call Mojo code. This integration, announced by Modular.AI’s product manager Brad Larson, marks a significant step toward broader adoption of Mojo, particularly in AI and accelerator-based computing.
Mojo, a high-performance programming language developed by Chris Lattner, has taken a significant step forward with the ability to interoperate with Python. Created as part of his work at Modular.AI, Mojo is designed as a superset of Python, offering enhanced capabilities for high-performance computing on modern accelerators. Today, Modular.AI’s product, manager Brad Larson announced that Python users can now call Mojo code, a development that is expected to broaden the language’s reach and utility in AI and computational tasks. The integration is part of an ongoing effort to make Mojo compatible with existing Python codebases, enabling developers to migrate performance-critical sections of their code to Mojo for improved efficiency. This update is available through several installation methods, including pip and Conda, and includes new examples demonstrating how Mojo code, even when utilizing GPUs, can be accessed from Python. The move represents a strategic effort to leverage Mojo’s capabilities within the Python ecosystem, potentially unlocking new applications and enhancing computational performance.
According to the announcement, one of the primary goals has been to gradually introduce MAX and Mojo into the vast Python codebases currently in use. The team believes that enabling selective migration of performance bottlenecks in Python code to fast Mojo, especially when running on accelerators, will unlock entirely new applications. The developers are especially excited about how this will expand the reach of the Mojo code that many of you have been writing. This is a significant achievement that has taken months of deep technical development, and it’s just the first step in the rollout of this new language feature. The post also mentions the importance of understanding the list of current known limitations to avoid potential frustrations and prevent the filing of duplicate issues for areas that are already being addressed. The conclusion of the post emphasizes the team’s interest in what users will build with this new functionality as well as their desire to hear feedback on how this could be made even better. The Mojo licensing makes it free on any device, for any research, hobby or learning project, as well as on x86 or ARM CPUs or NVIDIA GPU.
Read more of this story at Slashdot.