Showing posts with label benchmark. Show all posts
Showing posts with label benchmark. Show all posts

Transcoding with VAAPI decoding accelaration using VLC 2.0

I saw that vlc gets the decoded image back from the graphics cards when using hardware acceleration, e.g. VAAPI. So I thought this should also work as encoding accelerator, so the CPU only has to encode, not decode the video and cause at least some speedup, especially for large video files encoded into small files. I'm using the GMA500 poulsbo chipset for this test with 1,6 Ghz Atom(TM) CPU Z530.

Benchmarking VLC Encoding/Transcoding Speed

If you need to benchmark how fast vlc transcodes a certain file with certain settings, it's a bit tough. It doesn't seem to have any way to display the frames per second (fps) during encoding and no benchmark option as mplayer. A good way to work around this is to check how long it takes vlc to encode a certain file for a few seconds, e.g.

time vlc --stop-time 25 (...) vlc://quit.
e.g.  time vlc --stop-time 25 /D/MVI_6615.MOV --sout "#transcode{width=384,height=288,vcodec=h264,acodec=vorbis}:standard{access=file,mux=mkv,dst=test.mkv}}" --sout-x264-crf=24 --sout-transcode-threads=2 vlc://quit

This encodes for 25 seconds and then quits and displays the time it took. Make sure to have vlc://quit at the very end of the command line and after your input file. The "user" value is what matters, this shows how long it kept the cpu occupied. Of course this only works on Unix systems. Also be aware that the second time the input file comes from cache, so only use the values from the second or later encodings for comparison.

What can Governor Adjustments do for your Battery Mileage?

After my previous post about Android saving power with Linux governor tweaking, I did some governor tweak testing on my netbook with a tweaked phoronix test suide (pts) and came up with these results:

The first two show the power usage in Watts, the third shows the frequencies used during the test.

Switching my Main Browser: Opera 10.51 Review

Well, Opera doesn't seem to have convinced everybody, but it certainly convinced me to switch over again with version 10.51. Also in my personal test it was faster than Google Chrome beta and today's Chromium in the sunspider test (705 vs. 820 and 10xx ms). It works flawlessly with Google Mail, Blogger, etc. It consumes less memory than Chrome or Chromium, especially with many tabs open and uses less CPU when it's just idling. How I love the browser wars!

You can download the newest bleeding edge version of Opera from their Desktop Blog page.

Linux Boots Faster!

Check out this live video comparison at TuxRadar:

A Netbook for Linux

While most netbooks use common intel chips, some use the GMA500, which is yet badly supported for linux. An open source driver "may" be there in Q4 2009. But as the video decoding acceleration benchmarks show, it might be worth waiting for it. It's not only good, but also energy-efficient at decoding even HD material. The bad thing is though - there's at the moment no yet any well working Linux driver available. And you can't even know for sure it will ever come. If you're still interested, check out the MSI Wind U110.

Linux Kernels 2.6.24 - 2.6.29 Benchmarked (Phoronix)

I've read the newest Phoronix benchmark of all kernels 2.6.24 through 2.6.29 and if you don't follow the link now, I'll spoil it for you in the next paragraph...

Okay, so besides some normal fluctuations there is a significant speed gain for OpenSSL (+99%) and the OpenCL-backed Graphicsmagick (up to +73% and +80%) and a regression in 7-zip compression (-27%).

Wow - fascinating stuff. I wonder how much potential kernel code optimization has. I still remember the extreme difference between 2.4.x and 2.6.x. I had just bought a new computer because my old one was... getting old. And I felt completely cheated. The new kernel had brought me much more increase in responsiveness and "felt performance" than the new computer (AMD Duron 700 Mhz -> Athlon XP 2500+). If I could I probably would've brought it right back to the store. Well, they don't take returns on self-build computers here.

I really wonder how much difference the intel compiler compiled kernel makes in comparison to gcc. Maybe Phoronix could compare that one, too? Oh yeah "... boost up to 40% for certain kernel parts and an average boost of 8-9% possible" quotes LJ.

Latencytop Explained

I hope you all know what powertop is - at least if you own a Laptop with Linux on it - but I always had trouble with the wonderfully non-intuitive interface to LatencyTop. But now there's an article on explaining it.

Phoronix Compile Benchmarks

  1. Processor: Intel Celeron M 530 @ 1.72GHz (Total Cores: 1) "2008..."
  2. Processor: Intel Celeron 2.60GHz @ 2.62GHz (Total Cores: 1) "compilation"
  3. Processor: AMD Phenom 9850 @ 2.50GHz (Total Cores: 4) "0001"
  4. Processor: Intel Core 2 T7200 @ 2.00GHz (Total Cores: 2) "comp"
Time in seconds, less is better of course. You can see a clear advantage of multiple cores for compiling. I wonder how it looks for audio encoding...