A perfect example is what has happened with TV. We started with black and white TVs that required rabbit ears to acquire the signal. In the 80s, there were the competing standards of IDTV (improved definition TV), and HDTV (high definition TV). Over time, IDTV has dropped out of view, while HDTV gained popularity. Even with that, HD services are still not universal and have not been adopted globally.
UHDTV is now on the horizon. This comes at a time when the broadcast and production studios are re-thinking how they use technology and what kind of technology and tools they need to adopt to grow, and to meet future audience needs. Continuously evolving standards have contributed to this shift in thinking.
This shift has enabled infrastructure vendors to migrate from vertically integrated proprietary solutions to software defined platforms that can be used for multiple tasks, or workloads. These infrastructure vendors increasingly use server-based hardware solutions in studio, production, contribution and distribution environments. These CPU-based servers offset the challenge posed by rapidly changing broadcast requirements and standards.
Broadcasters can achieve a lower total cost of ownership (TCO) by using a bank of standard servers to create a pool of run-time, reconfigurable general processing engines with an open-source software management layer such as OpenStack for orchestration. The TCO is lower because the overall power consumed by the hardware is limited to the specific processing tasks undertaken at a given time. This commercial off-the-shelf (COTS) hardware will require significantly lower capital expenditure than deploying proprietary custom boxes for every specific video function or task.
Taking this model a step further, broadcasters can incorporate field programmable gate array (FPGA) devices into their servers, for acceleration, real-time processing, and configuration. This adds on-the-fly flexibility and improves compute performance per watt—a model that has been proven in data centers today. The compute-intensive FPGAs, which are extremely power efficient, are used for high speed data transport and processing intensive tasks such as real-time or faster video processing, encoding/ transcoding, and transport.
Today, at least one FPGA-accelerated H.265 encoder exists in the broadcast market. When compared against software-only solutions, this combination delivers a significant improvement in encode performance (4kp30 -> 4Kp60) with a minimal increase in power consumption.
In the future, broadcasters will benefit from the ability to use FPGA-based servers to optimize the workloads by moving compute-intensive video processing tasks from the CPU to the FPGA. This could provide seamless redundant switching to produce a video server that is more efficient in terms of power and deployment.
For example, a flexible and powerful broadcasting platform can be created by combining an Intel® Xeon® processor, and a discrete Arria® 10 FPGA device on a single CPU blade. This heterogeneous processing model enables advanced FPGA and CPU code development and programming with OpenCL.
Two other areas where FPGA technology can help greatly is in meeting the developing standards for moving uncompressed video and handling network synchronization.
The broadcast industry has been focused on the challenging task of finding an efficient means to transport native, uncompressed video to and from standard high-volume server infrastructures. FPGA technology enables more enterprise IT data management equipment and disciplines, and offers true scalability.
Industry associations such as the Video Services Forum, and the Society of Motion Picture & Television Engineers have created the VSF TR03 / TR04 and SMPTE2022 standards to support interoperable video data movement. An FPGA-based server enables flexible migration and bridging from Ethernet to the industry’s serial digital interface (SDI) standard.
Synchronizing network timing is critical in broadcasting and is certainly key for those considering such a fundamental technology change. In SDI, it is common to use a Blackburst generator to generate a time-sync across every piece of equipment in the studio to provide strong clock accuracy. The IEEE 1588 time synchronization standard for video over IP/Ethernet is now published through the SMPTE (SMPTE 2059-1). FPGAs, unlike unchangeable ASICS or ASSPs, can address the latest tweak, or update in a standard, because FPGAs are reprogrammable. This allows studio/broadcast equipment makers to design systems with SMPTE and other future requirements in mind, all with faster time to market.
The broadcast industry is now positioned to embrace technological advances brought on by the combination of lower cost, highly-video-capable, task-agnostic hardware that can be repurposed, as well as firmly established Ethernet networking standards for native video. By incorporating an infrastructure running on standard high volume servers that are powered by Intel FPGAs and CPUs broadcasters can have confidence they can meet the challenges today and tomorrow in studio, production, contribution and distribution environments.
Come see us at the Intel booth #SU621 and Intel Programmable Solutions (Altera) booth #SU9817 at NAB April 18-21 to learn more about these solutions, or read the NAB press release.
Intel, Xeon and the Intel logo are trademarks of Intel Corporation in the United States and other countries. *Other names and brands may be claimed as the property of others.
Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors.
Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit https://www.intel.com/performance.