@tlosm

Quote:

it is GlQuakeMos it is the same of GlQuakeWos and the GlQuakeAos version for make sys comparison is more equal do with the same build..

I used this one and after issuing gl_videosync 0 command I got 200 FPS in demo2:
http://aminet.net/package/game/shoot/GLQuake

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook

@tlosm

But even with the good old GLQuakeMOS I get 77fps in 1024x768 and 163fps in 640x480 with a standard Mac mini 1.5GHz in MorphOS 3.3. I have to run it in window mode, otherwise VSYNC limit will come on my road, limiting to 75 fps at both resolutions. Still, the numbers I get are in line with Fab's and well above those of tlsom...
(for comparison I get 211fps in 640x480, 111 in 1024x768, 75 in 1280x1024 and 55 in native fullscreen 1680x1050 on my PowerBook 1.67GHz with MorphOS 3.3 and GLQuakeMOS, but of course that's with an R300 based Radeon9700. I might try FodQuake as well sooner or later, just for fun!)

Kind regards,
Andrea

@itix

ok i will test it i found my vesion on Morphos files or meta i dont remember...
i any case thanks

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32;
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
#nomorea-eoninmyhome

It will be interesting to see 366Mhz & 400Mhz Cyberstorm ragemem benchmark results.

Last edited by delshay on 12-Oct-2013 at 06:03 PM.
Last edited by delshay on 12-Oct-2013 at 06:03 PM.

_________________
The Machine: Bride Of The Pin•Bot by Williams Electronics

@delshay

you will find it on sysmon on Os4.1 . it is my 604e375@366 result.

or here you can find it and the other machine bench :)
http://amiga.ikirsector.it/forum/viewtopic.php?f=37&t=17550&hilit=benchmark+os+4

my 4000 604e 366 con Radeon 9200 256mb PCi e 640Mb Ram
sdl bench (when i had the mediator kit)

Pitch = 320
Hardware surfaces avail = 1
Window manager avail = 1
Blitter hardware = 1
Colorkey blit hardware = 0
Alpha blit hardware = 0
Software->Hardware accel = 0
Video memory = 0

320x240 320x240 640x480 640x480
software hardware software hardware
Slow points (frames/sec): 0.475568 30.4183 0.0599637 7.5543
Fast points (frames/sec): 36.5453 15.1381 9.27167 3.82769
Rect fill (rects/sec): 1393.67 70620.7 338.54 64000
32x32 blits (blits/sec): 6400 40960 6196.67 41373.7

Last edited by tlosm on 12-Oct-2013 at 11:45 PM.
Last edited by tlosm on 12-Oct-2013 at 11:44 PM.

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32;
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
#nomorea-eoninmyhome

@tlosm

Quote:

or here you can find it and the other machine bench :) http://amiga.ikirsector.it/forum/viewtopic.php?f=37&t=17550&hilit=benchmark+os+4

Whereby, clearly the winner of the day is Peg2. Sadly, PA Semi could
be great for mobile phones or tablets, and for desktops only if it
was scaled to full 8 cores and scaled up in frequency.

Mind that under OS4 all X1000 results are single core.
Under Linux its like Sempron or Celeron with similar freq (1.5-1.8Ghz)
single core.

It seems AmigaOS users wil also cry Genesi is out of the game ...

Last edited by vox on 13-Oct-2013 at 12:00 AM.

_________________
Future Acube and MOS supporter, fi di good, nothing fi di unprofessionals. Learn it harder way!

@vox

Genesis aka bplan was Phase5 ... and you know it was the best hardware maker for the amiga platform... yes is normal cry for they not making more hardware for amigaos or mos too.

But for sure i know one thing we dont see what X1000 can do, first because is "semi" dual core and second the pcie 16x was never utilized for a 3d application because there are not drivers.
I dont know how run linux on it ... but im sure go great.

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32;
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
#nomorea-eoninmyhome

@vox

Quote:

Whereby, clearly the winner of the day is Peg2.

You mean in MIPS values of SysMon? That benchmark show really nothing.
Even your own benchmark shows single core PA6T 1.8 GHz is as fast as G4 1.4 GHz in CPU intensive applications (C-Ray benchmark).
PA6T has great advantage over classic G4 in superior memory performance - eg. delivers G4 2 GHz speed in MPlayer benchmark (single core).
In FFT based applications PA6T is comparable to 3.6 GHz G4 (FPU FFT HardInfo benchmark, again single core).

One thing I have notice here is I can achieve higher video benchmarks result by lowering the screenmode before starting benchmark test. I do stick to one screenmode whenever doing benchmark test. So it seems users can achieve higher result by using a lower screenmode. Perhaps this benchmark utility can be updated to show screenmode that was used to do the test.

Just for the record I always use 1024x768 32bit.

_________________
The Machine: Bride Of The Pin•Bot by Williams Electronics

@pavlor

Quote:

You mean in MIPS values of SysMon? That benchmark show really nothing. Even your own benchmark shows single core PA6T 1.8 GHz is as fast as G4 1.4 GHz in CPU intensive applications (C-Ray benchmark). PA6T has great advantage over classic G4 in superior memory performance - eg. delivers G4 2 GHz speed in MPlayer benchmark (single core). In FFT based applications PA6T is comparable to 3.6 GHz G4 (FPU FFT HardInfo benchmark, again single core).

That is under Linux where both cores are employed
In FPU intensive, memory intensive and Altivec enabled operations, I do believe what you say,
we just need that kind of apps / optimizations.

I will redo Linux test on lighter Linux (MintPPC) and with updated kernel.
That should not drastically change the results, but improve it a bit.

Last edited by vox on 13-Oct-2013 at 05:35 PM.

_________________
Future Acube and MOS supporter, fi di good, nothing fi di unprofessionals. Learn it harder way!

@vox

Quote:

That is under Linux where both cores are employed

Your C-Ray result is cca 2x faster than G4 1.4 GHz result, so single core performance is the same.

@pavlor

Quote:

Your C-Ray result is cca 2x faster than G4 1.4 GHz result, so single core performance is the same.

Nice, on Linux forums people used to compared it to a single core AthlonXP 1.5Ghz and we were badly beaten When it was downclocked to 1350Mhz, it was outperformed only in FPU tests.
Only achievement was doing it at 7W, so it is PowerEfficient.

_________________
Future Acube and MOS supporter, fi di good, nothing fi di unprofessionals. Learn it harder way!

@vox

Quote:

When it was downclocked to 1350Mhz, it was outperformed only in FPU tests.

Both Athlon XP 1500+ (1333 MHz) and PA6T 1.8 GHz could have about the same single core performance in most applications.

You can´t expect miracles from 6 years old SoC.

@pavlor

No, I am not. I expected a bit higher performance from 2x1.8Ghz anyway, but hope
RadeonHD can save the day in multimedia and gfx intensive apps / games.

Last edited by vox on 13-Oct-2013 at 07:32 PM.

_________________
Future Acube and MOS supporter, fi di good, nothing fi di unprofessionals. Learn it harder way!

nice the site Openbenchmark :)

however i don't know how it works with C-Ray, but it's very strange that a Q6600 (4 core) is so slower than Pa-Semi (2 core) and much soloer than other dual core cpu..

_________________

@Seiya

Quote:

however i don't know how it works with C-Ray, but it's very strange that a Q6600 (4 core) is so slower than Pa-Semi (2 core) and much soloer than other dual core cpu..

Lower result is better.

@guruman

I am sorry guys GLQuake doesn‚t work on X1000 with RadeonHD.
Will try with Warp3D software library later.

_________________
Future Acube and MOS supporter, fi di good, nothing fi di unprofessionals. Learn it harder way!

@vox

you can try using a old radeon instead :)
Radeon 9250 with 1,8 Ghz cpu should be give you very high fps

_________________

@Seiya

Quote:

you can try using a old radeon instead :) Radeon 9250 with 1,8 Ghz cpu should be give you very high fps

Please no jokes. I had Radeon 7000 and Radeon 9200SE years ago,
I prefer to wait for RadeonHD 3D driver.

_________________
Future Acube and MOS supporter, fi di good, nothing fi di unprofessionals. Learn it harder way!

@vox

Its useless to run this tests, as I understand it wa are NOT using optimized kernels on X1000, I read this in the tread about decoding raw audio to mp3 the other day.
We do not have 3D to test.

The memory benchmarks are just super, and I believe it will have greater impact on real life applications.

Last edited by NutsAboutAmiga on 15-Oct-2013 at 11:03 PM.
Last edited by NutsAboutAmiga on 15-Oct-2013 at 09:04 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS