To appreciate the challenges of IP and SDI integration is to understand that bringing the two together is fraught with questions, many of which are yet to be answered.

Timing remains one of the most important aspects of television. That is even more so as broadcasters and media production facilities move to 4K and 8K where constraints tighten and data capacity rises exponentially. In addition, live television demands that any delays be kept small as even two frames of delay can result in visible errors.

When facilities begin moving to IP studio distribution, equipment vendors often tell the engineers their software can process high-resolution video, but they sometimes forget to mention the accompanying lag and delay. Although today’s video software can achieve most of the traditional broadcast workflow processes, it may become less reliable when processing real-time high resolution video. That can especially be the case with ultra-high resolution, high frame rates or high dynamic range imagery.

Every software program works by executing one instruction after another in the servers’ processor. Sometimes servers have multiple processor cores, and if the software has been well written, then multiple instructions can be executed simultaneously. However, resource bottlenecks may still occur as the data hungry video moves between processor, memory, disk drive and GPU.

The FPGA solution

In the hardware side, thanks to the parallelism of FPGA (Field Programable Gate Array) designs, millions of instructions can be processed in a single clock cycle, making an FPGA thousands of times faster and more energy efficient than a server. Consisting of a single silicon core, and a highly-tuned memory, an FPGA transfers data at high speeds with lower power consumption and it can guarantee timing. 

In the past, innovators and system integrators have been forced to stay away from designing their own real-time FPGA solutions. The cost of high-speed circuit design was prohibitive because of the coding knowledge and tools required.

Today, broadcast engineers have become accustomed to fast, efficient, software deployments. They now expect the same flexibility in all areas of the workflow chain. This means vendors have to be able to identify signal and workflow problems, find solutions and write code to solve those issues with accuracy and speed.

Providing this flexibility comes in the form of FPGA solutions. These cards offer off-the-shelf availability, yet provide a high-speed FPGA hardware core, tuned to accommodate high-resolution video in a small size and low power consumption package.

The cards, some as small as a credit card, are capable of delivering 8K-120p video processing with sub-frame latency. Such a modular solution allows today’s manufacturers and system integrators to easily integrate high-processing power FPGA solutions inside their applications.

Vendors can build their own I/O circuit boards or use off-the-shelf cards to interface with the FPGA modules. Combined with high-level synthesis (HLS) coding software it now becomes possible to deliver real-time solutions with the agility and flexibility of more traditional software deployments.

Armed with the proper FPGA solution and a host card, a system integrator can write their own FPGA code in C/C++ or VHDL/Verilog and use one of the many available plug-and-play IP-Cores. The result can deliver real-time video processing, greater flexibility and provide additional user choice in today’s workflows.