Archive 1997–2024

ASRock Vision HT: Ivy Bridge Carries the SFF HTPC Forward

HTPC Decoding and Rendering Benchmarks

In our Ivy Bridge HTPC review, we had covered the CPU / GPU utilization during playback of various types of clips. In the Vision3D 252B review, we had graphs of CPU and GPU loading with various renderers and codecs. Unfortunately, AMD doesn't provide similar data / sensors for use with their APUs. Hence, we had to resort to power consumed at the wall along with GPU loading in the Trinity HTPC review. In order to keep benchmarking consistent across all HTPC reviews, we will be adopting the Trinity HTPC review methodology in our future HTPC articles.

The tables below present the results of running our HTPC rendering benchmark samples through various decoder and renderer combinations. Entries in bold indicate that there were dropped frames which indicate that the unit wasn't up to the task for those types of streams. The recorded values include the GPU loading and power consumed by the system at the wall. An important point to note here is that the system was set to defaults in the BIOS, but Windows power settings were set for maximum performance instead of the default balanced profile.

madVR

madVR was configured with settings similar to what we used for the Ivy Bridge HTPC review. Full screen windowed mode gaved the best performance in terms of avoiding dropped frames. In our first trial, we configured LAV Video Decoder to use avcodec (software decode). As expected, Intel's GPU / memory bandwidth is not enough for madVR processing of some types of content (namely, 720p60 H.264 and 1080p60 H.264).

LAV Video Decoder Software Fallback + madVR
Stream GPU Usage % Power Consumption
     
480i60 MPEG-2 78 35.46 W
576i50 H.264 64 25.92 W
720p60 H.264** 98 44.52 W
1080i60 H.264 86 47.40 W
1080i60 MPEG-2 84 42.55 W
1080i60 VC-1 86 47.10 W
1080p60 H.264** 98 48.81 W

madVR takes up more than 80% of GPU resources while processing 1080i60 material. However, the unit is quite power efficient on the whole, consuming less than 50 W for 1080i60 material using software decode. In the next trial, we configured LAV Video Decoder to use hawrdware decoding in the form of Quick Sync. Unfortunately, the results are quite similar to what we obtained with software decode (just that the power consumption is slightly lesser than software decode for the HD streams).

LAV Video Decoder QuickSync + madVR
Stream GPU Usage % Power Consumption
     
480i60 MPEG-2 79 36.16 W
576i50 H.264 64 27.45 W
720p60 H.264** 96 43.17 W
1080i60 H.264 90 45.23 W
1080i60 MPEG-2 87 42.86 W
1080i60 VC-1 88 44.66 W
1080p60 H.264** 94 46.32 W

EVR-CP and EVR

With the Enhanced Video Renderer (Custom Presenter) and Enhanced Video Renderer, native DXVA2 acceleration can be used. Here, we are able to process all our test streams without dropped frames. EVR is very efficient in terms of power consumption also.

LAV Video Decoder DXVA2 Native + EVR-CP
Stream GPU Usage % Power Consumption
     
480i60 MPEG-2 32 22.28 W
576i50 H.264 29 21.74 W
720p60 H.264 45 26.01 W
1080i60 H.264 47 26.87 W
1080i60 MPEG-2 47 25.48 W
1080i60 VC-1 43 33.46 W
1080p60 H.264 62 29.42 W

 

LAV Video Decoder Software Fallback + EVR
Stream GPU Usage % Power Consumption
     
480i60 MPEG-2 27 20.75 W
576i50 H.264 24 20.56 W
720p60 H.264 49 25.23 W
1080i60 H.264 35 23.44 W
1080i60 MPEG-2 35 22.23 W
1080i60 VC-1 33 30.45 W
1080p60 H.264 58 27.04 W

 

Originally published at https://www.anandtech.com/show/6443/asrock-vision-ht-ivy-bridge-carries-the-sff-htpc-forward (Wayback Machine).