Microsoft Announces Win 10 API for Windows 10 Developers

Microsoft points to the new trends that are coming in the coming years, that of artificial intelligence and machine learning. During the day of the date Microsoft made the announcement of a new API called WinML that aims to help developers take advantage of these two facets in the Windows 10 operating system.

Microsoft announces WinML to develop AI on Windows

WinML is a new set of APIs that will allow developers to take advantage of the full capabilities of any Windows 10 device to use pre-trained machine learning models, allowing AI tasks to be downloaded from the cloud.

The first reason is performance. Despite the enormous power of cloud computing available, we still live in a world where moving data to the cloud can be prohibitive in terms of cost and speed. The latency of any network connection is slower than accessing local memory, and working with massive data sets can be difficult without expensive, dedicated, high-bandwidth interconnects. Performing computing tasks locally can significantly improve performance thanks to lower latency and deliver real-time results. Operational costs can be saved by reducing network bandwidth as well as reducing computation time in the cloud.

Microsoft's new suite of AI APIs offer several key benefits that should help developers integrate them into their products. Arguably the most important thing is that the API does all the heavy lifting for the developer, so you wouldn't have to worry about what kind of hardware is available on any machine your app is going to run on. The WinML engine will leverage hardware dynamically and create code to get the best performance available from whatever hardware the device is running on.

The engine is built in Direct 3D, and if the system has a DX12 compatible GPU, it will use the DX12 calculation shaders dynamically. If you have a massive GPU with a lot of VRAM, the workload will be offloaded to the GPU. If a DX12 GPU is not available, or performance is an issue due to integrated graphics, the engine can also use the computing power of the CPU.

Perhaps most interestingly, the WinML engine will work even on SnapDragon 835-based computers, or even IoT devices. If it has a lot of performance available, it will take advantage of it, but if the system is a low power device, it will continue to work.

Anandtech font