
BDTI Solution Benchmark™ for H.264 Decoders
HOME << SERVICES << BENCHMARKING AND ANALYSIS <<

BDTI Solution Benchmark™ for H.264 Decoders

About the BDTI Solution Benchmark™ for H.264 Decoders

The BDTI Solution Benchmark for H.264 Decoders is a BDTI Solution Benchmark™ designed to enable credible and reliable performance evaluations of H.264 video decoders. Use of the BDTI Solution Benchmark for H.264 Decoders involves two phases: (1) a vendor configures an existing H.264 video decoder in a manner compliant with BDTI’s specification and runs BDTI’s test streams; and (2) BDTI certifies the performance of the solution.

All performance metrics reported for the BDTI Solution Benchmark for H.264 Decoders are for proprietary BDTI H.264 bitstreams. The BDTI H.264 bitstreams are created from proprietary BDTI video clips encoded with typical parameter settings. The BDTI video content is somewhat more computationally demanding than average for a typical full-length video clip (although it is not intended to uncover the worst case processing requirements for a solution).

Energy and area results for licensable silicon IP solutions are based on the BDTI Core Conditions, which specify a set of uniform conditions (fabrication process, temperature, voltage, etc.) used to derive the clock speed, area, and power consumption for each solution.

To obtain certification for your processing engine on the BDTI Solution Benchmark for H.264 Decoders, please contact BDTI.

Operating Points

The Operating Points (“OP”) associated with the BDTI Solution Benchmark for H.264 Decoders are shown in Table 1.

While the parameters listed in Table 1 (H.264 profile, H.264 level, image resolution, frame rate, and bit rate) have the largest impact on performance metrics, numerous additional parameters also affect performance. These parameters are fully defined in BDTI's proprietary bit streams.

OP Number	OP Type	Profile	Resolution	Frame Rate	Maximum Bit Rate
1	Primary	Baseline	D1 (720x480)	30 fps	1.5Mbps
2	Secondary	Baseline	D1 (720x480)	30 fps	384 kbps
3	Secondary	Baseline	D1 (720x480)	30 fps	10 Mbps
4	Secondary	Baseline	VGA (640 x 480)	30 fps	768 kbps
5	Secondary	Baseline	CIF (352 x 288)	30 fps	768 kbps
6	Secondary	Baseline	CIF (352 x 288)	15 fps	384 kbps

Table 1. BDTI H.264 Operating Points

The “OP Type” in the second column indicates whether the Operating Point is a Primary Operating Point or a Secondary Operating Point. The single Primary Operating Point is the most processing-intensive of the Operating Points. BDTI assumes that a solution capable of supporting the Primary Operating Point will also support the Secondary Operating Points. The Secondary Operating Points are included because:

It is desirable to test the functionality of several configurations and the Secondary Operating points represent scenarios that may be encountered in typical applications.
It is desirable to obtain performance metrics at multiple Operating Points to show tradeoffs and to instill higher degree of confidence in the performance results.

BDTI Certified™ results for the BDTI Solution Benchmark for H.264 Decoders at the Primary Operating Point are available free of charge on BDTI’s web site. For more detailed results, including certified results at Secondary Operating Points, please contact BDTI.

Performance Metrics

The following metrics are reported for the BDTI Solution Benchmark for H.264 Decoders.

Required Solution Clock Rate(MHz):

The required solution clock rate is reported as the minimum clock speed required for real-time operation. For solutions requiring external memory, solution clock rate requirements are reported for a number of external memory access times.

It is common for H.264 video decoder solutions to buffer output frames. Buffering output frames allows the processing required to decode a frame to be spread over multiple frames. Since the processing load varies considerably per frame due to changes in the video content, adding “delay buffers” to buffer output frames significantly reduces the required clock rate—at the expense of increased latency. For this reason, BDTI also reports the minimum clock rate required for real-time operation for a number of delay buffers as shown in Figure 1.

Figure 1: Block diagram showing the placement of a “delay buffer”.

External Memory Bandwidth Utilization (MB/s):

For solutions requiring an external memory device, the minimum external memory bandwidth required for real-time operation is reported when buffering a varying number of output frames in a manner similar to CPU utilization as discussed above.

Total Memory Usage:

The program, static data, and dynamic data memory usage of the solution is reported, in bytes.

Cost:

The cost of the solution is reported as follows:

For chips: Quantity 10,000
For licensable cores: Silicon area

For solutions requiring external memory, the characteristics of the external memory device are reported and should be considered when evaluating the overall solution cost.

Energy/Power Usage:

The energy usage of the solution is reported in joules per frame. The reported energy usage result may be an estimate based on simulation or, optionally, an actual physical energy measurement. The disclosure results note whether energy estimates or energy measurements are provided.

To have BDTI certify your H.264 decoder solution on the BDTI Solution Benchmark, please contact BDTI.

Top of page