Not actually sure if it’s slower or not, but anyway, I’ve been experimienting with Haiku R1A3, and a recent nightly build, and BeOS 5.03 on an Abit IP35 Pro motherboard. I’ve noticed that Chart runs at a much higher framerate in BeOS at 100% CPU usage (290 framrate, fullscreen, bitmap window) than either of the Haiku releases at 100%. In both Haiku versions, the framerate is around 175, and around 290 in BeOS.
In all cases, I kept the BIOS settings the same, and the screen set to 1024x768, 32-bit. The only difference is that I need to run BeOS using a Promise Ultra100 IDE pci card, and Haiku using the onboard SATA in IDE mode, but I don’t see how that would affect processing. Pulse shows all four cores working in BeOS and Haiku. Turning cores on/off one by one in Haiku and BeOS noticably changes the framerate. Not sure if I understand exactly how SMP, threading, etc. works… in Haiku Chart says it’s running at 100% CPU with 175 max framerate, but at the same time ActivityMonitor says each CPU is using only around 30%, and Pulse shows generally the same. Turning on “2 Threads” in Chart brings up the framerate to around 190.
I don’t care about Chart itself, but I’m wondering if this is indicative of the overall processing speed being somehow hindered in Haiku ? Are there any other programs I can try to test out processing speed in Haiku? Also, are there any tweeks I can do in a config file in Haiku to bring it up to BeOS performance? It might not be related to overall processing speed, but rather just something to do with how BeOS and Haiku process video. I noticed that changing the screen settings make a big difference in both Haiku and BeOS, regarding the framerate. When switching to “direct window” in Chart, the framerates are almost equal in BeOS and Haiku.
The computer specs are:
Abit IP35 Pro
Intel Q6600 Core2 Quad, 2.4 Ghz
Radeon X300 SE PCIe 16x video card, 128 mb
512 mb RAM, DDR2 800 MHz