Overview Features Coding ApolloOS Performance Forum Downloads Products Order Contact

Welcome to the Apollo Forum

This forum is for people interested in the APOLLO CPU.
Please read the forum usage manual.
Please visit our Apollo-Discord Server for support.



All TopicsNewsPerformanceGamesDemosApolloVampireAROSWorkbenchATARIReleases
Performance and Benchmark Results!

160 MIpspage  1 2 3 

Gunnar von Boehn
(Apollo Team Member)
Posts 6207
27 Aug 2017 14:49


Vampire



Vojin Vidanovic

Posts 770
27 Aug 2017 14:59


V4 tests?
 
  Really nice, chip mem speed, overall MIPS.
 
  What CANT be seen is V4 clock rate, and HOPE to see Caches ON,
  MMU active and at least emulated FPU in the future.
 
  Great work, keep the improvements on.


Roger Shimada

Posts 30
27 Aug 2017 16:00


Though you are teasing us by not mentioning the core, Vampire model or clock multiplier, that is an excellent achievement!


Gunnar von Boehn
(Apollo Team Member)
Posts 6207
27 Aug 2017 16:06


Roger Shimada wrote:

  Though you are teasing us by not mentioning the core, Vampire model or clock multiplier, that is an excellent achievement!
 

 
Its a Vampire-V2

Of course SYSINFO is as we all know a bad test.
Especially on Super-Scalar CPUs as SYSINFO is written not to take any benefit of Super-Scalar CPUs.
So we all know that the Sysinfo-Score both on 68060 and 68080 CPU is basically only 50% of the real power of the CPU.

So if Sysinfo shows 35 MIPS on 68060 - the 68060 in fact has 70 Mips
Same here .. the 160 MIPS on the 68080 could reality be 320 MIPS.


Vojin Vidanovic

Posts 770
27 Aug 2017 16:07


Gunnar von Boehn wrote:

  its a Vampire-V2
 

 
  Nice boost, more power to the cores!
 
  But we all want to know of V4
 
Gunnar von Boehn wrote:
  Of course SYSINFO is as we all know a bad test.  So if Sysinfo shows 35 MIPS on 68060 - the 68060 in fact has 70 Mips  Same here .. the 160 MIPS on the 68080 could reality be 320 MIPS.

Its 160MIPS in SysInfo :-) Well, maybe we need some updated benchmark tool, coders where are you? Since team isnt going to do everything some nice documentation on 080/SAGA programming should be teams 2018 goal :-)


Niclas A
(Apollo Team Member)
Posts 219
27 Aug 2017 16:30


*pft* ok score........nawh cant lie.. Bloody great :D


Vojin Vidanovic

Posts 770
27 Aug 2017 16:48


Niclas A wrote:

*pft* ok score........nawh cant lie.. Bloody great :D

Wonder what MiniBench would say on 2.7 V2 and 3+ core V4

EXTERNAL LINK 
Very nice page!


Thierry Atheist

Posts 644
27 Aug 2017 17:44


wooooo HOOOOO!!!!!!!!

Well..... call her Lady GAGA!!!

I see, I see, the Raspberry π 3 in the rear view mirror!!!!!!

(Darn, the code won't display properly.)


Michael Nurney

Posts 283
27 Aug 2017 18:01


awesome speed increase! ....we'll be play blu-ray soon on a 500


Thomas Blatt

Posts 200
27 Aug 2017 18:24


x17 Core?


Gregthe Canuck

Posts 274
28 Aug 2017 02:22


Very cool!

Would be interested in an updated Minibench score.


Keith Matthews

Posts 39
28 Aug 2017 07:33


Brilliant work, well done Gunnar and team. 8-)


Benutzer A4K

Posts 24
28 Aug 2017 20:37


When I use minibench, I get CPU: 10046, MEM: 2399 :)
 
But the Sysinfo result is very impressive. I cannot get to 100k Dhrystones :(
 
Olli.


Gunnar von Boehn
(Apollo Team Member)
Posts 6207
28 Aug 2017 20:46


Hallo Olli,
 
 
Benutzer A4K wrote:

  When I use minibench, I get CPU: 10046, MEM: 2399 :)
 

  MINIBENCH is unfortunately very inaccurate
  on a fast machine when called with small test parameter size (default) please set to biggest to get correct scores.
 
 
Benutzer A4K wrote:

  But the Sysinfo result is very impressive. I cannot get to 100k Dhrystones :(
  Olli.
 

 
  Its a pity that on AMIGA many people look at Sysinfo.
  As its technically a not really good test.
  Sysinfo scores very low (like halve score) on 68060 and 68080 - as its written fully sequential and basically only uses 50% of the advanced CPUs.
 
 


Roger Shimada

Posts 30
29 Aug 2017 01:18


As much as Dhrystones are a Meaningless Indicator of Processor Speed, it was widely available, so better than nothing.

These are some Dhrystone 2.1 results from UNIX systems I had access to in the 1990s.

118483 HP 9000 712/80i (80 Mhz)
121654 Unisys U/6000 500 (60 Mhz Pentium)
152905 Sun SPARCstation 20 (75 Mhz MicroSPARC)
158982 HP 9000 887 (x50) (96 Mhz PA 7100)
169491 IBM RS/6000 J30 (75 Mhz 601)
182815 HP K200 (100 Mhz PA 7200)
199401 HP T520 (120 Mhz PA 7150)



Gunnar von Boehn
(Apollo Team Member)
Posts 6207
29 Aug 2017 09:14


SYSINFO is a fake.
Sysinfo claims to measure Dhrystone results but does not implement the  dhrystone test.


Szyk Cech

Posts 191
29 Aug 2017 09:49


Gunnar von Boehn wrote:

SYSINFO is a fake.
  Sysinfo claims to measure Dhrystone results but does not implement the  dhrystone test.

So: what is the dhrystone result in SysInfo?!? Some random numbers?!?


Gunnar von Boehn
(Apollo Team Member)
Posts 6207
29 Aug 2017 10:29


Szyk Cech wrote:

 
Gunnar von Boehn wrote:

  SYSINFO is a fake.
    Sysinfo claims to measure Dhrystone results but does not implement the  dhrystone test.
 

  So: what is the dhrystone result in SysInfo?!? Some random numbers?!?
 

 
SYSINFO has some selfmade instruction test -

it measures how long it takes and prints out the Score as MIPS and dhrystone - but this test is not Dhrystone at all and has nothing to do with it.
 
The Sysinfo numbers are fake - and can not be compared to any real MIPS or DHRYSTONE numbers



Peter Heginbotham

Posts 214
29 Aug 2017 12:34


Very nice, what was the hardware spec running this?


Thellier Alain

Posts 141
29 Aug 2017 13:30


Hello

Does
EXTERNAL LINK
more a real Dhrystone test ?

Alain

posts 43page  1 2 3