BoingCube benchmark results

@Raziel

Glad to hear that

I kind of circumvented the bug by redrawing the whole background after activating the double buffered display.

For some reason, on your system (maybe the gfx-card), seems the background doesn't get copied to buffer the first time...

Anyway, good that it works now, thank you very much for letting me know

Screenshot of version 1.30...

@Marko

Quote:
But the background was there, it was always there, but as soon as the cube was drawn the first time (even being the size of a pixel) the background instantly went black (only to be revealed again by the cube flying over).

Strange...but my system tends to show such strange behaviour, had such with my XE as well, not sure why, must be the weather

edit: typo

@MarkoQuote:

Standard 800. I have one of the really early boards with a 7441/7451 Vger chip. I've always been afraid of overclocking it, even with a really good heat sink/fan on the chip.

Paul

Here's my results with no WBStartup stuff.

1.> boingcube -benchmark

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.30 (20151219)
Mode: Windowed NON-vsynced
Hardware acc.: Buff: ON (hardware) Gfx: ON (hardware)
Iterations: 659 frames
Duration: 10007 ms (10.007 seconds)
Rendering time: 86 ms (0.086 seconds) [lower is better]
D. frame-rate: 65.853902268412 FPS (acquired displayed frames/second)
Result: 7662.7906976744 FPS (rendering frame-rate) [higher is better]

1.> boingcube -benchmark -gofullscreen

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.30 (20151219)
Mode: Fullscreen NON-vsynced
Hardware acc.: Buff: ON (hardware) Gfx: ON (hardware)
Iterations: 650 frames
Duration: 10009 ms (10.009 seconds)
Rendering time: 84 ms (0.084 seconds) [lower is better]
D. frame-rate: 64.941552602658 FPS (acquired displayed frames/second)
Result: 7738.0952380952 FPS (rendering frame-rate) [higher is better]

1.> boingcube -benchmark -noaccgfx -noaccbuffer

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.30 (20151219)
Mode: Windowed NON-vsynced
Hardware acc.: Buff: OFF (software) Gfx: OFF (software)
Iterations: 591 frames
Duration: 10001 ms (10.001 seconds)
Rendering time: 781 ms (0.781 seconds) [lower is better]
D. frame-rate: 59.094090590941 FPS (acquired displayed frames/second)
Result: 756.72215108835 FPS (rendering frame-rate) [higher is better]


Much better results than with WBStartup stuff, there I got:
61.7 and 4262.1
56.6 and 3026.9
55.5 and 678.9


Sam440ep-flex, 800Mhz, 1GB RAM, Radeon 9250 and AOS4.1 FE.

My overclocked Pegasos2 with radeon9000 pro

7.RAM Disk:BoingCube> BoingCube -benchmark

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.31 (20151224)
Mode: Windowed NON-vsynced
Hardware acc.: Buff: ON (hardware) Gfx: ON (hardware)
Iterations: 1123 frames
Duration: 10005 ms (10.005 seconds)
Rendering time: 81 ms (0.081 seconds) [lower is better]
D. frame-rate: 112.24387806097 FPS (acquired displayed frames/second)
Result: 13864.197530864 FPS (rendering frame-rate) [higher is better]

7.RAM Disk:BoingCube> BoingCube -benchmark -gofullscreen

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.31 (20151224)
Mode: Fullscreen NON-vsynced
Hardware acc.: Buff: ON (hardware) Gfx: ON (hardware)
Iterations: 1119 frames
Duration: 10006 ms (10.006 seconds)
Rendering time: 76 ms (0.076 seconds) [lower is better]
D. frame-rate: 111.83290025984 FPS (acquired displayed frames/second)
Result: 14723.684210526 FPS (rendering frame-rate) [higher is better]

Here's my latest results:

X1000, Radeon HD 7770

> BoingCube -benchmark -hardware

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.40 (20160403)
Mode: Windowed NON-vsynced
Hardware acc.: Buff: ON (hardware) Gfx: ON (hardware)
Iterations: 1624 frames
Duration: 10001 ms (10.001 seconds)
Rendering time: 116 ms (0.116 seconds) [lower is better]
D. frame-rate: 162.38376162384 FPS (acquired displayed frames/second)
Result: 14000 FPS (rendering frame-rate) [higher is better]

> BoingCube -benchmark -software

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.40 (20160403)
Mode: Windowed NON-vsynced
Hardware acc.: Buff: OFF (software) Gfx: OFF (software)
Iterations: 1470 frames
Duration: 10003 ms (10.003 seconds)
Rendering time: 1150 ms (1.15 seconds) [lower is better]
D. frame-rate: 146.95591322603 FPS (acquired displayed frames/second)
Result: 1278.2608695652 FPS (rendering frame-rate) [higher is better]


PS. You don't have to untick everything in wbstartup, just hold down the right shift key while booting to disable wbstart.

Although the results in 1.31 and 1.40 should be pretty much same (I haven't touch anything involved in the benchmark itself).

Anyway, here is my latest result without WBStartup stuff:

1.> boingcube -benchmark -hardware

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.40 (20160403)
Mode: Windowed NON-vsynced
Hardware acc.: Buff: ON (hardware) Gfx: ON (hardware)
Iterations: 652 frames
Duration: 10002 ms (10.002 seconds)
Rendering time: 85 ms (0.085 seconds) [lower is better]
D. frame-rate: 65.186962607478 FPS (acquired displayed frames/second)
Result: 7670.5882352941 FPS (rendering frame-rate) [higher is better]

1.> boingcube -benchmark -software

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.40 (20160403)
Mode: Windowed NON-vsynced
Hardware acc.: Buff: OFF (software) Gfx: OFF (software)
Iterations: 585 frames
Duration: 10004 ms (10.004 seconds)
Rendering time: 784 ms (0.784 seconds) [lower is better]
D. frame-rate: 58.476609356257 FPS (acquired displayed frames/second)
Result: 746.17346938776 FPS (rendering frame-rate) [higher is better]


Version 1.40:

@Severin

Quote:
Hmm, right shift during booting, I tried that but nothing happend??

@Marko

Maybe you held it down too soon or too late. I press it when the harddrive led starts flashing like mad as it's loading workbench but after the light comes on for usb.

@Severin

No sorry, still nothing happens, I tried over 20 times at different times, still the WBStartup stuff loads up anyway... :S

@Marko

http://wiki.amigaos.net/wiki/Using_th ... kbench#Boot_modifier_keys

afaik that works fine.

regards...
michael

@Severin & MichaelMerkel

Yes I found the sweet spot, it works

One second after the AmigaOS logo has faded in, there's the sweet spot on my system.

I tested on an X1000 with Radeon HD5450 during normal system operation which is a complete boot with Dopus4 & Odyssey running on seperate screens. The tests were performed on 16bit and 32bit Workbench screens, which produced surprising results. The 'hardware' benchmark results are much lower on a 16bit screen than a 32bit screen and the 'software' benchmark results are much higher on a 16bit screen than on a 32bit screen. That would seem to indicate that games etc. that don't run on a 32bit screen would perform better in 'software' mode on my system. My tests were performed with BoingCube 1,40.

32 bit WB screen:

boingcube -benchmark -hardware

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.40 (20160403)
Mode: Windowed NON-vsynced
Hardware acc.: Buff: ON (hardware) Gfx: ON (hardware)
Iterations: 1572 frames
Duration: 10002 ms (10.002 seconds)
Rendering time: 294 ms (0.294 seconds) [lower is better]
D. frame-rate: 157.16856628674 FPS (acquired displayed frames/second)
Result: 5346.9387755102 FPS (rendering frame-rate) [higher is better]

boingcube -benchmark -software

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.40 (20160403)
Mode: Windowed NON-vsynced
Hardware acc.: Buff: OFF (software) Gfx: OFF (software)
Iterations: 1540 frames
Duration: 10003 ms (10.003 seconds)
Rendering time: 602 ms (0.602 seconds) [lower is better]
D. frame-rate: 153.95381385584 FPS (acquired displayed frames/second)
Result: 2558.1395348837 FPS (rendering frame-rate) [higher is better]

16 bit WB screen:

boingcube -benchmark -hardware

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.40 (20160403)
Mode: Windowed NON-vsynced
Hardware acc.: Buff: ON (hardware) Gfx: ON (hardware)
Iterations: 1652 frames
Duration: 10002 ms (10.002 seconds)
Rendering time: 453 ms (0.453 seconds) [lower is better]
D. frame-rate: 165.16696660668 FPS (acquired displayed frames/second)
Result: 3646.7991169978 FPS (rendering frame-rate) [higher is better]

boingcube -benchmark -software

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.40 (20160403)
Mode: Windowed NON-vsynced
Hardware acc.: Buff: OFF (software) Gfx: OFF (software)
Iterations: 1651 frames
Duration: 10003 ms (10.003 seconds)
Rendering time: 435 ms (0.435 seconds) [lower is better]
D. frame-rate: 165.05048485454 FPS (acquired displayed frames/second)
Result: 3795.4022988506 FPS (rendering frame-rate) [higher is better]

@Severin
It makes perfect sense that your Radeon HD 7770 would show faster 'hardware' benchmark results than my Radeon HD 5450 but I don't understand why your 'software' results would be lower. We're both using an X1000 with OS4.1FE.

thanks for another update Marko! will try out v1.4 when I'm back home

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.40 (20160403)
Mode: Windowed NON-vsynced
Hardware acc.: Buff: ON (hardware) Gfx: ON (hardware)
Iterations: 741 frames
Duration: 10007 ms (10.007 seconds)
Rendering time: 159 ms (0.159 seconds) [lower is better]
D. frame-rate: 74.048166283601 FPS (acquired displayed frames/second)
Result: 4660.3773584906 FPS (rendering frame-rate) [higher is better]

Sam460 1ghz, Radeon HD7770, AOS4.1FE.

*** BoingCube benchmark ***
Name: -benchmark10 (Counts iterations per ten seconds.)
Version: 1.40 (20160403)
Mode: Fullscreen NON-vsynced
Hardware acc.: Buff: ON (hardware) Gfx: ON (hardware)
Iterations: 641 frames
Duration: 10006 ms (10.006 seconds)
Rendering time: 133 ms (0.133 seconds) [lower is better]
D. frame-rate: 64.061563062163 FPS (acquired displayed frames/second)
Result: 4819.5488721805 FPS (rendering frame-rate) [higher is better]
Cheers

@xenic

Could be because I'm running two graphics cards at 1920x1080x32, Keeping the other screen 'alive' during testing must use some cpu time. The 7770 is the main workbench screen so I can run the test on that without a screen being open on the 5450 but I cannot do it the other way round.

It's all fairly pointless anyway, A monitor can only display upto it's refresh rate (61Hz for mine) so at 122 fps half the frames aren't displayed or produce tearing in some programs. much better to turn vsync on if available and give the cpu/gpu a rest :)