| ▲ | stefan_ 8 hours ago | |
I suppose its unintentional comedy that they picked a 1080p H264 video playback as the benchmark. Because of course the chip in the Raspberry Pi 1 was literally designed for that! The only thing it asks of you is that you make use of the fixed function blocks that take up much of its silicon space. So no wonder that utterly fails with modern software - we need to go all the way to RPi 5 to smother the problem with enough generic computing power to overcome the careless people that spearhead much of browser development. | ||
| ▲ | FrostViper8 5 hours ago | parent | next [-] | |
I don't think it is really a fair test because I doubt the browser is going to use any of the video decoding hardware that maybe available. I suspect if you used something native then any video decoding hardware would be used correctly. | ||
| ▲ | vardump 8 hours ago | parent | prev | next [-] | |
Yeah. Except it's not even fixed function blocks, it's the 12 core VideoCore IV GPU running software that does the decoding. VideoCore is the real Raspberry Pi, the ARM block running Linux was just a subprocessor that VC controls. | ||
| ▲ | dividuum 7 hours ago | parent | prev | next [-] | |
Yep. A Pi 1 can almost play 1080p60 with a proper zero-copy decoding setup. Pi 2 and beyond have no issue with that. As you said: The Pi5 has enough CPU power, so even the H264 decoding itself now uses software as it no longer has a hardware decoder. Oh well. | ||
| ▲ | epolanski 4 hours ago | parent | prev [-] | |
To be fair it was the most indicative test. | ||