Overview Features Coding Performance Forum Downloads Products OrderV4 Contact

Welcome to the Apollo Forum

This forum is for people interested in the APOLLO CPU.
Please read the forum usage manual.



All TopicsNewsPerformanceGamesDemosApolloVampireAROSWorkbenchATARIReleases
Performance and Benchmark Results!

POVRay - Come to Fish Togetherpage  1 2 3 4 

Xtro 1975

Posts 3
22 Jan 2021 22:29


Standalone V4+ CoffinOS R58 and R5 core:
 
    720 x 486
    ----------------------------------------------------------------------------
    Pixels:          350640  Samples:          844624  Smpls/Pxl: 2.41
    Rays:          1908305  Saved:                20  Max Level: 5/5
    ----------------------------------------------------------------------------
    Ray->Shape Intersection          Tests      Succeeded  Percentage
    ----------------------------------------------------------------------------
    Box                            3785275        2011310    53.14
    Cone/Cylinder                  146697          19266    13.13
    CSG Intersection              2718041          372275    13.70
    CSG Union                      1391422          282357    20.29
    Plane                          2146539        1328721    61.90
    Quadric                        3677206        1032301    28.07
    Sphere                        37021717        8377031    22.63
    Clipping Object                228368          216628    94.86
    Bounding Box                  35109513        19724500    56.18
    Light Buffer                  55664771        32066613    57.61
    Vista Buffer                  15731137        11544558    73.39
    ----------------------------------------------------------------------------
    Calls to Noise:            818335  Calls to DNoise:        2734252
    ----------------------------------------------------------------------------
    Shadow Ray Tests:          6286136  Succeeded:              964438
    Reflected Rays:            1047784  Total Internal:              52
    Refracted Rays:              15672
    Transmitted Rays:              225
    ----------------------------------------------------------------------------
    Time For Parse:    0 hours  0 minutes  1.0 seconds (1 seconds)
    Time For Trace:    0 hours 23 minutes  9.0 seconds (1389 seconds)
        Total Time:    0 hours 23 minutes  10.0 seconds (1390 seconds)


Nick Fellows

Posts 111
23 Jan 2021 17:35


Is this version of POVRay optimised for vampire ?


Andy Hearn

Posts 366
23 Jan 2021 17:52


not that i was aware of... BTW, V1200 isn't far behind,  Gold2.13 RC3 build 7717 x12 85Mhz (vanilla R58 image) turns in 24 minutes and 3 seconds  will see what the V500 can do later tonight if i can


Darren Eveland

Posts 51
24 Jan 2021 03:22


Andy Hearn wrote:

not that i was aware of... BTW, V1200 isn't far behind,  Gold2.13 RC3 build 7717 x12 85Mhz (vanilla R58 image) turns in 24 minutes and 3 seconds  will see what the V500 can do later tonight if i can

My Vampire 500 got 27 minutes and change.  Gold 2.12 x12 (85MHz).  I was doing some other stuff in the background while it was running, but not much.


Darren Eveland

Posts 51
24 Jan 2021 05:00


Darren Eveland wrote:

Andy Hearn wrote:

  not that i was aware of... BTW, V1200 isn't far behind,  Gold2.13 RC3 build 7717 x12 85Mhz (vanilla R58 image) turns in 24 minutes and 3 seconds  will see what the V500 can do later tonight if i can
 

 
  My Vampire 500 got 27 minutes and change.  Gold 2.12 x12 (85MHz).  I was doing some other stuff in the background while it was running, but not much.

Just tested my two Amiga 4000s, so I'll add them to my test list:

Vampire 500 V2+, Core 2.12 x12 85Mhz : 27 minutes.

A4000, CS PPC 68060 @50MHz, GREX 4000: 50 minutes, 1 second.

A4000, CS MK2 68060 @100MHz, Mediator: 46 minutes, 42 seconds.

Note the CS PPC memory is much faster than the CS MK2 memory.

Darren


Darren Eveland

Posts 51
24 Jan 2021 23:13


Doing some more comparisons here:

Mac Mini 1.5GHz G4: 4 minutes 13 seconds (impressive!)
X1000 1800MHz Pa Semi, OS4: hangs at line 222....not sure why.


Andy Hearn

Posts 366
25 Jan 2021 09:33


@Darren
NICE! i did wonder about the CS mk2 clocked up to 100, thanks for that info. i guess the 64bit memory interface on the CS PPC really helps out.

I'm guessing you've got a different set of FPU libs running? - as i get 1hr 5 mins with my CS Mk3 060@50Mhz powered A3k with generic libraries


Xtro 1975

Posts 3
25 Jan 2021 11:47


I did a test on my Windows PC, i7 6700k overclocked to 4300mhz with WinUAE 64bit
 
No JIT result: 0 hours 19 minutes  45.0 seconds (1185 seconds)
 
JIT result: 0 hours  1 minutes  9.0 seconds (69 seconds)
 
So a Vampire is more or less like WinUAE with JIT OFF


Darren Eveland

Posts 51
25 Jan 2021 17:11


Andy Hearn wrote:

@Darren
  NICE! i did wonder about the CS mk2 clocked up to 100, thanks for that info. i guess the 64bit memory interface on the CS PPC really helps out.
 
  I'm guessing you've got a different set of FPU libs running? - as i get 1hr 5 mins with my CS Mk3 060@50Mhz powered A3k with generic libraries

I'm not sure actually, but I can check for you what I'm running.  the ram speed setting may help - mine is likely set to 60ns in the Cyberstorm setup screen (press ESC when powering on).



Kymon Erec Zonias

Posts 7
26 Jan 2021 10:02


Dear Andy, nice to see your first rendering results on an real 68060.
Iím a big 3D and Renderfan myself. Just a tip to get better results with the 68060.
Use at least the Cyberpatcher that comes with the Cyberstorms or better use the
latest Oxyron Patcher (3.14 iirc).

Regards
Kymon


Andy Hearn

Posts 366
26 Jan 2021 10:28


Thanks for the tips guys,
i'll try the ram speed settings - after eyeballing exactly what i've got on my Mk3 - i think i'm keeping it "safe" and sitting with conservative settings to not crash more than i have to, and i'll definitely see what Oxyron can do :)

I'd like to get my BlizzardPPC/040 upgraded to an 060, but with getting a V1200, there seems to be little point in the effort of doing that beyond seeing just how much faster a V1200 is than any top end hardware from "back-in-the-day"... :D


Henryk Richter
(Apollo Team Member)
Posts 128/ 1
27 Jan 2021 09:56


A good old 68060 is still able to give a good punch.

Example: A4000, 68060@100MHz, Matze TK060 (in Beta)

720 x 486
----------------------------------------------------------------------------
Pixels:          350640  Samples:          841040  Smpls/Pxl: 2.40
Rays:          1906080  Saved:                4  Max Level: 5/5
----------------------------------------------------------------------------
Ray->Shape Intersection          Tests      Succeeded  Percentage
----------------------------------------------------------------------------
Box                            3858829        2021269    52.38
Cone/Cylinder                  146871          19321    13.16
CSG Intersection              2641040          363189    13.75
CSG Union                      1301867          274165    21.06
Plane                          2056369        1286824    62.58
Quadric                        3834218        1046256    27.29
Sphere                        35901752        8462559    23.57
Clipping Object                229128          217392    94.88
Bounding Box                  34884678        19654541    56.34
Light Buffer                  54814093        31929341    58.25
Vista Buffer                  15593114        11463479    73.52
----------------------------------------------------------------------------
Calls to Noise:            799182  Calls to DNoise:        2735506
----------------------------------------------------------------------------
Shadow Ray Tests:          6262714  Succeeded:              963540
Reflected Rays:            1049312  Total Internal:              66
Refracted Rays:              15557
Transmitted Rays:              171
----------------------------------------------------------------------------
Time For Parse:    0 hours  0 minutes  1.0 seconds (1 seconds)
Time For Trace:    0 hours 25 minutes  53.0 seconds (1553 seconds)
    Total Time:    0 hours 25 minutes  54.0 seconds (1554 seconds)

EXTERNAL LINK


Andy Hearn

Posts 366
27 Jan 2021 13:12


O.O

that's fantastic, i need that in my A4000's life! :D



Henryk Richter
(Apollo Team Member)
Posts 128/ 1
27 Jan 2021 14:40


Well, as impressive these projects for new 060 cards are: full 68060's are hard to come by. The final die-shrink Rev.6 68060's are even more difficult to find. To get hold of one among those to run fully stable at 100 MHz is like chasing a unicorn.

FPGA re-implementations of the 68k series won't have this shortage.


Andy Hearn

Posts 366
27 Jan 2021 15:04


true. i have some 060's in a box that aren't doing anything... one rev6 i can allocate to a card that's going to run it at over the original 50Mhz . otherwise, i'll probably sit the rev5's on my blizzard cards if i ever get them swapped, as i'll never really want to run original hardware overclocked.

but generally yes, FPGA is going to be the way forward...


Gunnar von Boehn
(Apollo Team Member)
Posts 5474
27 Jan 2021 15:20


Henryk Richter wrote:

full 68060's are hard to come by. T

Don't be pessimistic - there might be some hope!
I just spotted a fresh pile of 68Ks in China:

EXTERNAL LINK


Antony Coello

Posts 123
27 Jan 2021 20:11


Henryk Richter wrote:

Well, as impressive these projects for new 060 cards are: full 68060's are hard to come by. The final die-shrink Rev.6 68060's are even more difficult to find. To get hold of one among those to run fully stable at 100 MHz is like chasing a unicorn.

Yep. This.

Even the non fake higher clocked 030s are getting scarce now.

It seems everyone and his dog are making 680x0 accelerators these days.

Its like someone deciding to make a new graphics card to sell that uses 20 x 3DFX Voodoo 5 GPUs per card.  :/



Kymon Erec Zonias

Posts 7
27 Jan 2021 21:09


Hey Henryk, nice to see first results of Matzeís 060TK. Did you run the Test with or without the OxyronPatcher 3.14? If not you may even beat your time dramatically.
  Would love to know :and see ;-)
 
  Thanks and regards
  Kymon


Andy Hearn

Posts 366
28 Jan 2021 09:08


Gunnar von Boehn wrote:

 
  Don't be pessimistic - there might be some hope!
  I just spotted a fresh pile of 68Ks in China:
 
  EXTERNAL LINK 

*laughing* yeah yeah ok. i was thinking of having dig in my kids lego draw to make an equally legit '060 and post that on ebay - see what i can get... ;)



Darren Eveland

Posts 51
29 Jan 2021 02:49


Henryk Richter wrote:

A good old 68060 is still able to give a good punch.

      Total Time:    0 hours 25 minutes  54.0 seconds (1554 seconds)
 
  EXTERNAL LINK 

Interesting this result is more than 2x faster than my 100MHz 060 on my Cyberstorm MK2.  It must be that povray is very memory intensive, because with the same clock speed the only difference is that the Matze TK060 has much faster memory interface than a CS MK2 with the 100MHz hack (memory running @ 50MHz).  Does the TK060 run it's memory @ 100MHz?


posts 79page  1 2 3 4