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!

IDE How Fast Can You Go?page  1 2 

Gunnar von Boehn
(Apollo Team Member)
Posts 6207
28 Jun 2018 12:41


How fast can Vampire IDE go?




Peter Heginbotham

Posts 214
28 Jun 2018 12:45


Nice, I'm guessing this is on an Amiga 500 vampire, the most i got when testing on a beta version last year was about 6 MB/Sec


Miroslav Parvanov

Posts 5
28 Jun 2018 12:58


Holly Molly ... great job!


Saladriel Amrael

Posts 166
28 Jun 2018 12:59


Well, with that speed most games/apps will be loaded into RAM within a couple of seconds or less.

Very interesting


Vojin Vidanovic
(Needs Verification)
Posts 1916/ 1
28 Jun 2018 15:34


Gunnar von Boehn wrote:

How fast can Vampire IDE go?

18 mb/s on ffs - great!



Leigh Russ

Posts 151
28 Jun 2018 20:13


Still waiting to see if FastIDE will ever be implemented again for V600 users after the beta test of it some time ago. Following that it all went quiet and was never mentioned again


Gunnar von Boehn
(Apollo Team Member)
Posts 6207
28 Jun 2018 20:34


Leigh Russ wrote:

Still waiting to see if FastIDE will ever be implemented again for V600 users after the beta test of it some time ago. Following that it all went quiet and was never mentioned again

The IDE interface on the V500 is fully under our control.
The Vamp support up to 30 MB/sec peak speed.
Depending on drive and driver you might not fully reach this.
The team is currently tweaking drivers to improve the speed.

On the A600 you use the original Amiga chips.
Those original Amiga chips were designed for a speed up to 3.5 MB/sec. We know some tricks to get more out of those chips, 6 MB/sec for example.
The problem is here what you can get out of your A600 it not as reliable as the V500 and it does variate from board to board.
Some board might reach 9MB/sec, some 6 MB/sec, others only 4.5 MB/sec.

I see the following problem with the A600 IDE.
People can get more speed out of the 30 year old chips.
But if they select to high speed - they could get wrong data on their drives. In worst case they might loose data.
I'm a little bit afraid of people accidentally using to high speed on their A600 and the risk for them to loose data.




Andy Hearn

Posts 374
28 Jun 2018 21:38


wow! great stuff! but this is only for compact flash drives right?


Gunnar von Boehn
(Apollo Team Member)
Posts 6207
29 Jun 2018 05:50


Andy Hearn wrote:

wow! great stuff! but this is only for compact flash drives right?

The Faster IDE modes also works on SSD and on Hard Drives


Vojin Vidanovic
(Needs Verification)
Posts 1916/ 1
29 Jun 2018 10:56


Gunnar von Boehn wrote:
  The Faster IDE modes also works on SSD and on Hard Drives

 
 
  Its nice archievement for itself. In speed is highest IDE performance as UDMA ATA33 on PC, unimaginable at old ATA standards.
 
  Hope even combinations like IDE DVD burners and IDE SSDs will be possible, especially on standalone/v4.
 
  SCSI.device without SCSI devices since A300 :-)
 
 


Martin Soerensen

Posts 232
29 Jun 2018 12:39


Really cool to see these kinds of speeds on a Vampire. I assume that this is an unmodified V500 board? Now you are venturing into SCSI-controller performance which was maybe the last barrier towards having the 68080 compete with PPC board performance since the CPU performance already surpasses the PPC in many/most areas. :)


Aksel Andersen

Posts 120
29 Jun 2018 20:07


Sweet..

Is this core 3.x related or will we see this in 2.x as well?


Gregthe Canuck

Posts 274
29 Jun 2018 23:01



Are there plans for DMA support on the IDE port?


Gunnar von Boehn
(Apollo Team Member)
Posts 6207
30 Jun 2018 07:47


gregthe canuck wrote:

Are there plans for DMA support on the IDE port?

Yes this is correct.
The Vampire has IDE-DMA and the team is working on drivers using it atm.


Mallagan Bellator

Posts 393
30 Jun 2018 09:48


Correct me if I’m wrong, but this scsi.device is and will be patched into the ROM on the vampire, right?


Gunnar von Boehn
(Apollo Team Member)
Posts 6207
30 Jun 2018 10:49


Mallagan Bellator wrote:

Correct me if I’m wrong, but this scsi.device is and will be patched into the ROM on the vampire, right?

You are correct, when the DMA driver is finished then it will be included in the update - and will work right out of the box.


Mallagan Bellator

Posts 393
30 Jun 2018 11:33


ROMs can be larger than 512KB, right? I mean, there’s no real limit to that, save for the space limit of the chip itself, right?
If my memory serves me well, the 3.0 ROM and higher, has support for the AGA features of the A1200/4000. My guess is that could be extended to the SAGA chipset as well, if the ROM and FPGA supports it


Gunnar von Boehn
(Apollo Team Member)
Posts 6207
30 Jun 2018 11:55


Mallagan Bellator wrote:

ROMs can be larger than 512KB, right?

Yes all Vampires come with 1 MB Rom.


Leigh Russ

Posts 151
03 Jul 2018 21:42


Gunnar von Boehn wrote:

Leigh Russ wrote:

  Still waiting to see if FastIDE will ever be implemented again for V600 users after the beta test of it some time ago. Following that it all went quiet and was never mentioned again
 

 
  The IDE interface on the V500 is fully under our control.
  The Vamp support up to 30 MB/sec peak speed.
  Depending on drive and driver you might not fully reach this.
  The team is currently tweaking drivers to improve the speed.
 
 
  On the A600 you use the original Amiga chips.
  Those original Amiga chips were designed for a speed up to 3.5 MB/sec. We know some tricks to get more out of those chips, 6 MB/sec for example.
  The problem is here what you can get out of your A600 it not as reliable as the V500 and it does variate from board to board.
  Some board might reach 9MB/sec, some 6 MB/sec, others only 4.5 MB/sec.
 
  I see the following problem with the A600 IDE.
  People can get more speed out of the 30 year old chips.
  But if they select to high speed - they could get wrong data on their drives. In worst case they might loose data.
  I'm a little bit afraid of people accidentally using to high speed on their A600 and the risk for them to loose data.

Totally understand the reasoning behind it. However could it not be re-enabled for the A600 and just set to "off" as default. And then leave it up to the users to decide if they want to enable it or not? Even a small increase is better than nothing at all afternoon. The majority of people wouldn't even touch the feature, but for some who like to experiment and push stuff to its limits, or who don't have any data they are worried about losing then it would be a good feature to try out



Stefano Briccolani

Posts 586
04 Jul 2018 06:14


I totally agree. Even a 4.5 mb/s speed on ide is better than 3. Leave the feature off then give instructions on how to enable/find optimal speed. This would be the perfect solution.


posts 25page  1 2