First Amiga 600 FPGA Accelerator - Vampire 600 - Page 28

kipper2k · 22 October 2013, 13:44

Quote:

Originally Posted by dJOS

Cheers, the state I live in (South Australia) is very bad for IT jobs so looks like I'll be moving my little family to Victoria shortly - another big life adventure!

Don't do what i done... Be ruthless, if you don't need it, sell it or chuck it. I've got to the point where i have kept too much "Just in case" and chucked it when it arrived on this end

dJOS · 22 October 2013, 23:24

Quote:

Originally Posted by kipper2k

Don't do what i done... Be ruthless, if you don't need it, sell it or chuck it. I've got to the point where i have kept too much "Just in case" and chucked it when it arrived on this end

Good advice, we are chucking stuff like cheap ikea furniture the can easily be replaced (costs more to ship than replace) to fit our stuff into 1 shipping container.

tnt23 · 28 October 2013, 08:11

My Vampire 600 has just arrived! Can't wait to try it tonight. Majsta, great job

majsta · 28 October 2013, 09:07

I hope that board is not damaged. it was long journey from me to you exactly 27 days. Too long. Anyway I think that you will participate in project in another way, not just as a user since you can understand complete concept and you are able to understand core. So when I publish it I expect from you a lot

Also please clean the board because at the time I didn't have proper cleaning liquid. Now, I have so other boards sent are much better looking.

28 October 2013, 11:50

Hello everyone! I am one of the lucky ones to have received a Vampire, and Majsta asked me to report my findings here.

I tried it with an A600 rev 1.1 with Gayle 1, and rev 1.3 with Gayle 2, and it works on both! (020 & fast ram recognized) Also, Indy A603 chipram works, and some older chip expansion on the other machine works.

I swapped the Gayle 1 mobo from a friend who had an early ACA-620, that was not working with it. It was fixed by a firmware upgrade, and they said it was a Gayle problem. But whatever, the Vampire does not seem to have that problem.

I get a crash in da Jormas: Major Release (source: http://www.jormas.com/download/JRm-MjRl.s ) in the beginning of the "colortunnel" part (I think).

majsta · 28 October 2013, 11:58

Hi

I uploaded new core much slower, with less memory but more compatible. Try that. I will try also to build faster and more compatible core today.

alexh · 28 October 2013, 12:21

Quote:

Originally Posted by deetsay

I swapped the Gayle 1 mobo from a friend who had an early ACA-620, that was not working with it. It was fixed by a firmware upgrade, and they said it was a Gayle problem.

Do you have any links to discussions about this?

28 October 2013, 12:34

Quote:

Originally Posted by alexh

Do you have any links to discussions about this?

It has to be this:
http://eab.abime.net/showthread.php?t=65087&page=18

Instability hmmm... I thought it was just not working at all, but I could be wrong. Right now I can't remember which mobo I've been using more. Might need some more testing!

lukassid · 28 October 2013, 14:34

I got A300 mobo rev1. Very first A600. My serial is 000584. Vampire works fine same as fixed ACA620. I have compability issues with Classic Workbench 3.9. I will try slower core today or tomorrow and report back.

majsta · 01 November 2013, 15:17

Hi

I want to inform everyone that everything goes as planned so next logical stage was to open source complete code. I didn't finished documentation yet because there is lot job left to be done but for those who are familiar with VHDL publishing code should be enough. Version I have published is for core 0.1 but it should provide enough information how everything is done. Top level design file contains only 600 lines of code

http://majsta.com/modules.php?name=N...article&sid=80

billt · 16 November 2013, 19:28

Quote:

Originally Posted by majsta

Hi

I want to inform everyone that everything goes as planned so next logical stage was to open source complete code. I didn't finished documentation yet because there is lot job left to be done but for those who are familiar with VHDL publishing code should be enough. Version I have published is for core 0.1 but it should provide enough information how everything is done. Top level design file contains only 600 lines of code

http://majsta.com/modules.php?name=N...article&sid=80

Very cool news! I'm hopeful to buy a zedboard or an sockit board and play around with those fpga chips for something like this. perhaps sockit as it has better video for a minimig port.

billt · 20 November 2013, 04:33

Does your involvement with Apollo mean that this core will become open-sourced at some point as well?

majsta · 20 November 2013, 11:12

hi

I can't be certain since we didn't discuss that. At some point maybe. Apollo(Phoenix) version is so complex thing that so far I had hard time to track what those brilliant minds created. My participation in this was too small so it is not fair for me to decide about anything. I found some bugs, proposed solutions, and only write about 500 lines of code so my role was small for now. I know that few people are writing documentation about the core, detailed documentation is written only if you want to sell something or to open source it so we will see.

wawa · 12 January 2014, 16:31

seems majsta has given up as expected.
http://www.majsta.com/
hopefullyat least documentation has been released to preserve the effort just in case someone would like to pick up (not that i think it likely)

chaos · 12 January 2014, 17:37

Quote:

Originally Posted by wawa

seems majsta has given up as expected.
http://www.majsta.com/
hopefullyat least documentation has been released to preserve the effort just in case someone would like to pick up (not that i think it likely)

You're not a very positive person, are you?

AFAIK kipper2k is already making these boards, so nothing is lost IMO.

alexh · 12 January 2014, 19:04

Quote:

Originally Posted by wawa

seems majsta has given up as expected.
http://www.majsta.com/
hopefullyat least documentation has been released to preserve the effort just in case someone would like to pick up (not that i think it likely)

there's no point in majesta doing bitcoin in an FPGA. we've seen ASICs doing 100x performance per watt over FPGA. he'll be back

majsta · 12 January 2014, 19:41

I m not gone. Everything goes as planned so don't spread false rumors

@alexh
All the time in my life someone is telling me that I can't do something and that it can't be done. Do you recall Vampire project

lukassid · 12 January 2014, 20:29

Yes we recall Vampire project. It's vey important to finish what you doing before jumping on something new. Don't get me wrong I'm always on your side, just please don't disappear for too long mate.

alexh · 12 January 2014, 23:20

a lot of minds far smarter than you or I are making full ASIC bitcoin mining accelerators using ULP. Only another ASIC will match them. 3 years ago an FPGA bitminer would have made you a lot of money. Today there are just too many miners and the algorithm means you'll struggle to make even enough money to cover the electricity.

NovaCoder · 12 January 2014, 23:56

This is one of the few technical posts on EAB where I have no idea what anyone is talking about

28 October 2013, 09:07	#544
majsta www.majsta.com Join Date: Jun 2010 Location: Banjaluka/Republic of Srpska Age: 43 Posts: 448	I hope that board is not damaged. it was long journey from me to you exactly 27 days. Too long. Anyway I think that you will participate in project in another way, not just as a user since you can understand complete concept and you are able to understand core. So when I publish it I expect from you a lot Also please clean the board because at the time I didn't have proper cleaning liquid. Now, I have so other boards sent are much better looking. Last edited by majsta; 28 October 2013 at 09:14.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Vampire 600 and floppy issues	Firestone	support.Hardware	15	22 March 2017 18:11
Vampire 600 more cores..	Turran	support.Hardware	48	14 January 2015 17:39
Vampire 600 wierd issues	Retro	support.Other	4	05 September 2014 22:36
Vampire 600 troubles	Viserion	support.Hardware	21	10 December 2013 20:28
WTB: Amiga 600 Accelerator	Gordon	MarketPlace	4	21 February 2009 16:06

28 October 2013, 08:11	#543
tnt23 Registered User Join Date: Feb 2008 Location: Saint-Petersburg / Russia Posts: 324	My Vampire 600 has just arrived! Can't wait to try it tonight. Majsta, great job

28 October 2013, 11:50	#545
deetsay Posts: n/a	Hello everyone! I am one of the lucky ones to have received a Vampire, and Majsta asked me to report my findings here. I tried it with an A600 rev 1.1 with Gayle 1, and rev 1.3 with Gayle 2, and it works on both! (020 & fast ram recognized) Also, Indy A603 chipram works, and some older chip expansion on the other machine works. I swapped the Gayle 1 mobo from a friend who had an early ACA-620, that was not working with it. It was fixed by a firmware upgrade, and they said it was a Gayle problem. But whatever, the Vampire does not seem to have that problem. I get a crash in da Jormas: Major Release (source: http://www.jormas.com/download/JRm-MjRl.s ) in the beginning of the "colortunnel" part (I think).

28 October 2013, 11:58	#546
majsta www.majsta.com Join Date: Jun 2010 Location: Banjaluka/Republic of Srpska Age: 43 Posts: 448	Hi I uploaded new core much slower, with less memory but more compatible. Try that. I will try also to build faster and more compatible core today.

28 October 2013, 14:34	#549
lukassid Registered User Join Date: Oct 2012 Location: Surrey Posts: 390	I got A300 mobo rev1. Very first A600. My serial is 000584. Vampire works fine same as fixed ACA620. I have compability issues with Classic Workbench 3.9. I will try slower core today or tomorrow and report back.

01 November 2013, 15:17	#550
majsta www.majsta.com Join Date: Jun 2010 Location: Banjaluka/Republic of Srpska Age: 43 Posts: 448	Hi I want to inform everyone that everything goes as planned so next logical stage was to open source complete code. I didn't finished documentation yet because there is lot job left to be done but for those who are familiar with VHDL publishing code should be enough. Version I have published is for core 0.1 but it should provide enough information how everything is done. Top level design file contains only 600 lines of code http://majsta.com/modules.php?name=N...article&sid=80

20 November 2013, 04:33	#552
billt Registered User Join Date: Sep 2010 Location: Maryland, USA Posts: 82	Does your involvement with Apollo mean that this core will become open-sourced at some point as well?

20 November 2013, 11:12	#553
majsta www.majsta.com Join Date: Jun 2010 Location: Banjaluka/Republic of Srpska Age: 43 Posts: 448	hi I can't be certain since we didn't discuss that. At some point maybe. Apollo(Phoenix) version is so complex thing that so far I had hard time to track what those brilliant minds created. My participation in this was too small so it is not fair for me to decide about anything. I found some bugs, proposed solutions, and only write about 500 lines of code so my role was small for now. I know that few people are writing documentation about the core, detailed documentation is written only if you want to sell something or to open source it so we will see.

12 January 2014, 16:31	#554
wawa Registered User Join Date: Aug 2007 Location: berlin/germany Posts: 1,054	seems majsta has given up as expected. http://www.majsta.com/ hopefullyat least documentation has been released to preserve the effort just in case someone would like to pick up (not that i think it likely)

12 January 2014, 19:41	#557
majsta www.majsta.com Join Date: Jun 2010 Location: Banjaluka/Republic of Srpska Age: 43 Posts: 448	I m not gone. Everything goes as planned so don't spread false rumors @alexh All the time in my life someone is telling me that I can't do something and that it can't be done. Do you recall Vampire project

12 January 2014, 20:29	#558
lukassid Registered User Join Date: Oct 2012 Location: Surrey Posts: 390	Yes we recall Vampire project. It's vey important to finish what you doing before jumping on something new. Don't get me wrong I'm always on your side, just please don't disappear for too long mate.

12 January 2014, 23:20	#559
alexh Thalion Webshrine Join Date: Jan 2004 Location: Oxford Posts: 14,331	a lot of minds far smarter than you or I are making full ASIC bitcoin mining accelerators using ULP. Only another ASIC will match them. 3 years ago an FPGA bitminer would have made you a lot of money. Today there are just too many miners and the algorithm means you'll struggle to make even enough money to cover the electricity.

12 January 2014, 23:56	#560
NovaCoder Registered User Join Date: Sep 2007 Location: Melbourne/Australia Posts: 4,400	This is one of the few technical posts on EAB where I have no idea what anyone is talking about

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)