My Hades and benchmark test

All things related to the Hades and Milan clones

Moderators: Mug UK, moondog/.tSCc., [ProToS], lp, Moderator Team

Kroll
Captain Atari
Captain Atari
Posts: 248
Joined: Fri Mar 09, 2012 10:07 am

My Hades and benchmark test

Postby Kroll » Mon Mar 07, 2016 8:40 pm

Hi,
I would like to say also about the results of speed tests, using the program Kronos 2.02. As a reminder, this is the Hades with a problem with the PCI (see other thread). First run Hades with CTRL (clean TOS with 2 colors) and the results are surprising:

Processor: 4.9
FPU: 177.3
Memory and video bus speed: 10.7
VDI 3.2
Disk: 10.5
3D open GL: 26.6

The second test I've done at 800x600 resolution with 256 colors

Processor: 4.7
FPU: 162.7
Memory and video bus speed: 15.4
VDI: 13.8
Disk: 10.5
3D open GL: 13.8

Very low is the speed of the processor.

How do you think whether it can cause problems with PCI?
I do not remember and I can not find previous a test but are there any standard for Hadesa060 and results are definitely better.

Regards
Kroll

OL
Captain Atari
Captain Atari
Posts: 366
Joined: Fri Apr 01, 2005 6:59 am
Contact:

Re: My Hades and benchmark test

Postby OL » Sun Mar 13, 2016 11:20 am

Hello,
Can you post, the result by saving result as ABH file.

A fast Hades 060 should give something near this for CPU:

CPU_MEMOREAD=42.228; (Mo/sec)
CPU_MEMOWRITE=33.121; (Mo/sec)
CPU_BLOCKMOVE=20.030; (Mo/sec)
CPU_BITSHIFT=119.509; (Minst/sec)
CPU_MULDIV=5264.623; (Kinst/sec)
CPU_OVERLAP=60.379; (Minst/sec)
CPU_RAWPOWER=119.509; (Minst/sec)

I have often seen in the past lower results often link to cache configuration.

And a Milan 040 25Mhz:
CPU_MEMOREAD=34.940; (Mo/sec)
CPU_MEMOWRITE=21.635; (Mo/sec)
CPU_BLOCKMOVE=14.047; (Mo/sec)
CPU_BITSHIFT=12.467; (Minst/sec)
CPU_MULDIV=1951.687; (Kinst/sec)
CPU_OVERLAP=24.937; (Minst/sec)
CPU_RAWPOWER=24.937; (Minst/sec)

If the reference is the default I propose with Kronos, so your results looks slow (even for an Hades 040 (I never seen test on it))

Olivier
OL

Kroll
Captain Atari
Captain Atari
Posts: 248
Joined: Fri Mar 09, 2012 10:07 am

Re: My Hades and benchmark test

Postby Kroll » Wed Mar 30, 2016 9:25 pm

Hi,

I send you benchmark as ABH file from Kronos 2.02. Below you can find.
All test I send you as atachments. This test was made with the problem of PCI still.
I dont know why, but in program I see 9.8 MHz (very strange)

CPU_TIMER_ERROR=0;
CPU_MEMOREAD=16.058; (Mo/sec)
CPU_MEMOWRITE=16.530; (Mo/sec)
CPU_BLOCKMOVE=8.770; (Mo/sec)
CPU_BITSHIFT=9.714; (Minst/sec)
CPU_MULDIV=838.390; (Kinst/sec)
CPU_OVERLAP=9.876; (Minst/sec)
CPU_RAWPOWER=9.650; (Minst/sec)

FPU_TIMER_ERROR=0;
FPU_MULTIPLICATION=506732.098; (inst/sec)
FPU_SQRT=296800.841; (op/sec)
FPU_LOG=23326.737; (op/sec)
FPU_ADDITION=495571.129; (inst/sec)
FPU_SINCOS=19853.378; (op/sec)
FPU_FLOATDOUBLE=860349.564; (inst/sec)

HADESPCI.TXT


Regards
Kroll
You do not have the required permissions to view the files attached to this post.

OL
Captain Atari
Captain Atari
Posts: 366
Joined: Fri Apr 01, 2005 6:59 am
Contact:

Re: My Hades and benchmark test

Postby OL » Thu Mar 31, 2016 8:32 pm

Hello

I see your results, for an Hades 060 the result is not normal, looks like your cache are OFF.

Results should be around 6 time faster!

Olivier


Kroll wrote:Hi,

I send you benchmark as ABH file from Kronos 2.02. Below you can find.
All test I send you as atachments. This test was made with the problem of PCI still.
I dont know why, but in program I see 9.8 MHz (very strange)

CPU_TIMER_ERROR=0;
CPU_MEMOREAD=16.058; (Mo/sec)
CPU_MEMOWRITE=16.530; (Mo/sec)
CPU_BLOCKMOVE=8.770; (Mo/sec)
CPU_BITSHIFT=9.714; (Minst/sec)
CPU_MULDIV=838.390; (Kinst/sec)
CPU_OVERLAP=9.876; (Minst/sec)
CPU_RAWPOWER=9.650; (Minst/sec)

FPU_TIMER_ERROR=0;
FPU_MULTIPLICATION=506732.098; (inst/sec)
FPU_SQRT=296800.841; (op/sec)
FPU_LOG=23326.737; (op/sec)
FPU_ADDITION=495571.129; (inst/sec)
FPU_SINCOS=19853.378; (op/sec)
FPU_FLOATDOUBLE=860349.564; (inst/sec)

HADESPCI.TXT


Regards
Kroll
OL

Kroll
Captain Atari
Captain Atari
Posts: 248
Joined: Fri Mar 09, 2012 10:07 am

Re: My Hades and benchmark test

Postby Kroll » Thu Mar 31, 2016 9:30 pm

OK, thanks for comment, I saw it that CACHE is OFF, but how can I turn ON ?
Do I have to use in AUTO folder Cacheon.prg?

Regards
Kroll

OL
Captain Atari
Captain Atari
Posts: 366
Joined: Fri Apr 01, 2005 6:59 am
Contact:

Re: My Hades and benchmark test

Postby OL » Sat Apr 02, 2016 7:49 am

Hello

as I remember there was an acc or a CPX to fix this, it's old

Olivier

Kroll wrote:OK, thanks for comment, I saw it that CACHE is OFF, but how can I turn ON ?
Do I have to use in AUTO folder Cacheon.prg?

Regards
Kroll
OL

User avatar
lp
Fuji Shaped Bastard
Fuji Shaped Bastard
Posts: 2171
Joined: Wed Nov 12, 2003 11:09 pm
Location: GFA Headquarters
Contact:

Re: My Hades and benchmark test

Postby lp » Sat Apr 02, 2016 8:49 am

Just run cacheon.prg from the desktop as a test, then start the benchmark. The results should be better. ;)

OL
Captain Atari
Captain Atari
Posts: 366
Joined: Fri Apr 01, 2005 6:59 am
Contact:

Re: My Hades and benchmark test

Postby OL » Sat Apr 02, 2016 6:44 pm

lp wrote:Just run cacheon.prg from the desktop as a test, then start the benchmark. The results should be better. ;)


There is 2 mode cache on 68060: CACHE_WRITETHROUGH and CACHE_COPYBACK

the second one is less compatible but far faster, I'm near sure there is a tool to choose the mode

There is too a branch prediction that can be or not enabled.


Olivier
OL

User avatar
lp
Fuji Shaped Bastard
Fuji Shaped Bastard
Posts: 2171
Joined: Wed Nov 12, 2003 11:09 pm
Location: GFA Headquarters
Contact:

Re: My Hades and benchmark test

Postby lp » Sat Apr 02, 2016 7:53 pm

I always used the cacheon.prg supplied by Ozk's mmutools archive. However, I think you can get the old ACC from Fedi's page at:
http://www.medusacomputer.com/thes-download.html

Kroll
Captain Atari
Captain Atari
Posts: 248
Joined: Fri Mar 09, 2012 10:07 am

Re: My Hades and benchmark test

Postby Kroll » Sat Apr 02, 2016 9:19 pm

Hi,

thanks a lot, I install Hades_B3.ACC and I thing the test is much beeter, results is in file HADESPC3.TXT.
The test was made in resolution 640x480x256K and system was clean, only driver for ISA card. What do you think about it now ?
The question is what is different between Hades_B3, hades_B4, Hades_B6.ACC ?
I have to turn off set_mmu.prg in AUTO folder. All memory is as ST RAM, IF I can undedrstand I have to change set_mmu.inf ?

HADESPC3.TXT


Regards
Kroll
You do not have the required permissions to view the files attached to this post.

OL
Captain Atari
Captain Atari
Posts: 366
Joined: Fri Apr 01, 2005 6:59 am
Contact:

Re: My Hades and benchmark test

Postby OL » Sun Apr 03, 2016 9:33 am

Kroll wrote:Hi,

thanks a lot, I install Hades_B3.ACC and I thing the test is much beeter, results is in file HADESPC3.TXT.
The test was made in resolution 640x480x256K and system was clean, only driver for ISA card. What do you think about it now ?
The question is what is different between Hades_B3, hades_B4, Hades_B6.ACC ?
I have to turn off set_mmu.prg in AUTO folder. All memory is as ST RAM, IF I can undedrstand I have to change set_mmu.inf ?

HADESPC3.TXT


Regards
Kroll


Your CPU results are good and normal now! your computer should run far faster.

On FPU you can do better do you have selected the 68060 test or the 68881?

Olivier
OL

Kroll
Captain Atari
Captain Atari
Posts: 248
Joined: Fri Mar 09, 2012 10:07 am

Re: My Hades and benchmark test

Postby Kroll » Sun Apr 03, 2016 5:59 pm

Hi,

Yes now computer run far faster, but still I have another problem with SET_MMU, but I will describe a new thread.
Test with 68060 but it is not much better than 68881. I have seen test is much better if I used in AUTO folder FPU060.PRG than FPU__3 PRG.

Regards
Kroll


Social Media

     

Return to “Hades / Milan”

Who is online

Users browsing this forum: No registered users and 1 guest