DeGirium hopes developers find DeLight platform for edge AI app development changes game
A California-based AI startup developing software and hardware infrastructure for edge computing, DeGirum, announced DeLight, a cloud platform designed for edge AI application development. DeLight provides easy-to-use IT solutions for organizations deploying edge computing technology. The company says the platform can help developers reduce edge AI application development time from months to days.
DeLight’s cloud platform, along with the DeGirum ORCA family of hardware accelerators, is a user-friendly AI solution that makes developing applications and testing hardware more straightforward, the company says. The cloud platform features hardware-agnostic application programming interfaces and pre-compiled ML models. Developers and engineers can additionally access multiple-edge AI hardware in the cloud.
“Before DeLight, developers had to spend months prototyping their application SW on HW from different vendors to compare the HW options and make the right choice for their deployment,” said Winston Lee, chief executive officer of DeGirum. “Now, the DeLight platform enables developers to focus exclusively on their application logic, without having to deal with the frustration of setting up HW and modifying code multiple times.”
DeLight makes it possible for product engineers to prototype AI applications in the cloud before setting up the hardware at edge locations. After developing an application on the cloud platform, organizations can use the ORCA AI hardware accelerator to deploy their application closer to the edge. The cloud platform provides access to PySDK documentation, token generation, DeGirum Labs and PySDK examples.
DeGirum PySDK is a Python package that lets users develop edge AI applications with just a few lines of code. PySDK’s easy-to-use APIs and widely used utilities make it a breeze to manage images, camera feeds, audio streams and video files. Developers can use multiple models simultaneously to get valuable information from data.
DeGirum’s ORCA AI accelerator hardware can provide high performance and sophisticated edge functionalities. Its efficient compute architecture with support for pruned models powers the hardware suite. ORCA’s ability to process prune models allows for multiplied compute and bandwidth resources, giving more processing power and accuracy to real-time applications at the edge.
Article Topics
AI/ML | application development | DeGirium | DevOps | edge AI | model
Comments