However, I just made a nice little industrial project called "right piece of shit" that takes around 20% of the CPU. Probably because I don't have the VS2013 runtimes? I have installed VC++ 2005,2008,20 runtimes. Unfortunately, I couldn't test the VSE2013 compiled version because Windows rebel that "reaper.exe is not a valid win32 application". GA-890XA-UD3 Gigabyte motherboard with the latest XP drivers. Graphic card is Gigabyte Nvidia GT240 with the v196.46 driver. My OS for audio is Windows XP SP3 and the audio computer is based on AMD Phenom 965BE CPU with 4GB of CL6 G.Skill RAM Audio driver for RME HDSP is the last v3.40. The FX CPU use jumped considerably with smaller sample block sizes, as expected.įirst, the session overview with the navigator, minus video and bus tracks:Ĥ6.9% - Standard (subsequent re-tests push that figure up to 52%)Ģ56 Samples (includes CPU and RT-CPU graphs) All data is for the Windows 圆4 version, and all are portable installs with settings imported from my main installation of Reaper 圆4.ĬPU is i7 920 2.66 GHz running Win7 圆4 with an SSD as a system disk, from which Reaper is also being run.ħ20p H264 MP4 video with the Master/Parent deactivated and the track fader at -inf.ġ60 effect plugins, almost all of them fully automated in small section chunks, including several instances of Aether. I ran one of my larger mixing projects as a benchmark, the result of which can be viewed here (). Thanks in advance! Or if you're busy, please don't bother :) The results of these tests will not have an affect on the immediate development, it is more for long term planning. Then describe the tests made and the relative performance of the versions you tested. Intel Core 2 Duo 3.06GHz, or AMD FX-8350 4.0GHz)
To help in this, I've made builds of REAPER 4.611 using a few different compiler setups, so we can decide whether it makes sense to modernize some of our compilers for the release builds.įor this to be useful, perform performance comparisons between the different builds (mainly the first two, which represent the current release setup and the more modern but non-ICC equivalent) for a particular platform and architecture, and let us know the results of your tests.ĬPU brand type and speed (e.g.
I've been looking at our development toolchain for some time, and want to reevaluate some of the assumptions we've made.
0 kommentar(er)
