English Amiga Board Amiga Lore


Go Back   English Amiga Board > News

 
 
Thread Tools
Old 08 October 2016, 14:29   #181
grond
Registered User

 
Join Date: Jun 2015
Location: Germany
Posts: 184
Quote:
Originally Posted by jarp View Post
There's your problem, you make an assumption which may or may not be true.
There is no problem because I clearly marked that as an assumption. Of course, the number of people interested in an 060 card could still be higher than the number of cards on offer even with the vampire available.
grond is online now  
AdSense AdSense  
Old 08 October 2016, 14:32   #182
eXeler0
Registered User

eXeler0's Avatar
 
Join Date: Feb 2015
Location: Sweden
Age: 43
Posts: 1,391
Quote:
Originally Posted by britelite View Post
No, it will have to be 100% compatible WITH the 060 for people to trade their 060-cards. This is pretty crucial considering it's being marketed as a replacement for the 060, ALL old 060-software has to work on it.
I know you are assembler guru, so yeah, you're in the "I'm keeping my 060" category. But look at it from a different user perspective and things aren't as obvious anymore.
Quote:
Originally Posted by idrougge View Post
People who have 060 cards are people who tend to have more than one Amiga, too. Like me (except I don't have an 060 card). I have an Apollo 3040 in my A3000. If i got a V3000, that Apollo would move into my A4000, thereby pushing an A3640 onto the market.
And I think that's what would generally occur; Vampire 1200 cards would result in 1260 card being put into the second A1200 of Vampire owners, thereby pushing a lot of tired old 1230 cards onto the second-hand market.
Maybe, but it could also be that you'd keep the 030 card because pretty much everything released back (90%++) then works fine, fast enough and with very few compatibility issues, whereas the other "power Amiga" would be the one you use for the other 10% like more recent ports of PC games and various other power hungry stuff.
And if this other would be a Vampire based Amiga, it would also run WB in HD True color and do other nice stuff. So looking at it from a practical perspective, most users might find themselves storing their 1260 cards in the drawer..

Oh well, speculation, speculation. As for the new Phake5 offers, I have a relaxed "wait and see" approach, but I am not really interested regardless.. it fills no gap I have.



Skickat från min HTC One via Tapatalk
eXeler0 is offline  
Old 08 October 2016, 14:55   #183
AJCopland
Registered User

 
Join Date: Sep 2013
Location: Beeston, Nottinghamshire, UK
Posts: 51
If there was an open-source or otherwise DIY 68060 card for the A1200/CD32/A4000 then all of this would be less of an issue because those of us interested could build our own.

There isn't one though, so it's a card from Jens which he cannot get enough CPUs for to make a production run available, a Vampire deriviant which doesn't exist yet, or the Phase5 announcement.

How about, since this is a thread about Phase5, we give them time to release something and THEN judge what they have released rather than trash talking the new owners of the Phase5 name/rights/etc?

Just an idea.
I mean if they release a load of crap supported by a pre-order scheme then flame away but right at the moment they've not really done anything wrong and the "community" is already attacking them for what the owners of the name did TWO companies previously.
AJCopland is offline  
Old 08 October 2016, 17:00   #184
britelite
Registered User
 
Join Date: Feb 2010
Location: Espoo / Finland
Posts: 225
Quote:
Originally Posted by eXeler0 View Post
But look at it from a different user perspective and things aren't as obvious anymore.
I'm of course looking at this from a demoscene centric point of view, but why would I as a user want to trade a known working platform (060-accelerator) for something that's not compatible enough (the vampire), if that results in a lot of demos not working anymore. Especially as the Vampire most definitely won't be getting any decent demos of it's own.

(EDIT) I do agree with others though, that there's a better possibility that people will get a Vampire to complement their collection. But the Vampire won't make 060-accelerators redundant in the scale you seem to suggest.

Last edited by britelite; 08 October 2016 at 17:10.
britelite is offline  
Old 08 October 2016, 18:11   #185
grond
Registered User

 
Join Date: Jun 2015
Location: Germany
Posts: 184
Quote:
Originally Posted by britelite View Post
why would I as a user want to trade a known working platform (060-accelerator) for something that's not compatible enough (the vampire)
There are no 060 accelerators for either a500 or a600 and there is no vampire for any of the amigas that did get their 060 accelerators. Hence, all of this is hypothetical. But as for compatibility, the level of 68k compatibility of the 080 is actually better than that of the 060 because the 080 has all the instructions the 060 has plus several ones the 060 is missing. The FPU will be 882-compatible and not stripped down like that of the 060.
grond is online now  
Old 08 October 2016, 18:18   #186
Locutus
Registered User

 
Join Date: Jul 2014
Location: Finland
Posts: 612
Note, 060 demo prods essentially means AGA+060, the classic compo machine is a A1200 + Blizz 1260.

(yes i know theres some demos that have CGX support)
Locutus is offline  
Old 08 October 2016, 19:40   #187
eXeler0
Registered User

eXeler0's Avatar
 
Join Date: Feb 2015
Location: Sweden
Age: 43
Posts: 1,391
Quote:
Originally Posted by britelite View Post
I'm of course looking at this from a demoscene centric point of view, but why would I as a user want to trade a known working platform (060-accelerator) for something that's not compatible enough (the vampire), if that results in a lot of demos not working anymore. Especially as the Vampire most definitely won't be getting any decent demos of it's own.

(EDIT) I do agree with others though, that there's a better possibility that people will get a Vampire to complement their collection. But the Vampire won't make 060-accelerators redundant in the scale you seem to suggest.
I know you're not a fan of the Vamp project but I don't understand why you feel the need to tell how it's not as compatible. First of all, like Grond pointed out, it has a very complete 68k instruction set, e the FPU is still missing - well have to wait and see about how good that one will be. Secondly, the Vamp brings a lot of other useful features to the table. I bet a lot of ppl would trade 100% compatibility for stuff like 24 bit color HD WB, HDMI out + other planned features. But most importantly, it's an FPGA, don't you think it will gradually improve over time to become faster and even more compatible? Ppl at the forums report bugs, and core is improved.. 👍

And while it "most definitely" will not be getting any 080 optimized demos from you, I wouldn't be so confident in predicting the future for everyone else in that matter. 😜

Regarding the "scale" at which ppl will replace their 060.. well that's obviously guesswork on my part, but it only needs to be as big as the amount of ppl who would today run and buy a 060 ACA or P5 accelerator for €400-500++
Difficult to say how many that would be..

I think it's a bit like with Oculus Rift. For a while everyone had the impression it would be in tight supply for a long time because of preorders, General hype and signs of big interest, it turns out after the eager and loud early adopters got their gear, the sales slowed down a lot.

http://metro.co.uk/2016/09/06/vr-sal...es-up-6112288/


Skickat från min HTC One via Tapatalk
eXeler0 is offline  
Old 08 October 2016, 20:39   #188
trixster
Guru Meditating
trixster's Avatar
 
Join Date: Jun 2014
Location: York
Posts: 493
Whilst the prospect of new phase5 stuff is interesting, unless the price and immediately availability is unbeatable then I'll waitout for Vampire 1200 and / or vampire 4000, or buy a secondhand accelerator if the itch becomes unbearable.
trixster is offline  
Old 08 October 2016, 21:20   #189
matthey
Registered User
 
Join Date: Jan 2010
Location: Kansas
Posts: 1,006
Quote:
Originally Posted by grond View Post
There are no 060 accelerators for either a500 or a600 and there is no vampire for any of the amigas that did get their 060 accelerators. Hence, all of this is hypothetical. But as for compatibility, the level of 68k compatibility of the 080 is actually better than that of the 060 because the 080 has all the instructions the 060 has plus several ones the 060 is missing.
I agree. The Apollo-core CPU should be more compatible with 68k CPUs before the 68060 than the 68060. Write through caches especially should help with self modifying code. Compatibility with the 68060 specifically may have some problems.

Quote:
Originally Posted by grond View Post
The FPU will be 882-compatible and not stripped down like that of the 060.
Where do you see this? I keep seeing this claim but never anything to back it up. From everything I have read, the Apollo-core FPU will not have the 6888x instructions which were removed from the 68060 FPU. It should still be compatible with the 6888x by using software support just as the 68040 and 68060 FPU is. I could be wrong here but all 6888x instructions would take a huge amount of resources and logic to implement and store them in the FPGA.
matthey is online now  
Old 08 October 2016, 22:11   #190
grond
Registered User

 
Join Date: Jun 2015
Location: Germany
Posts: 184
Quote:
Originally Posted by matthey View Post
Where do you see this? I keep seeing this claim but never anything to back it up. From everything I have read, the Apollo-core FPU will not have the 6888x instructions which were removed from the 68060 FPU. It should still be compatible with the 6888x by using software support just as the 68040 and 68060 FPU is. I could be wrong here but all 6888x instructions would take a huge amount of resources and logic to implement and store them in the FPGA.
Only the 040 FPU instructions will be implemented in hardware. The 882-only instructions will cause an internal trap that will result in emulation code being executed from the FPGA-ROM. This emulation code basically corresponds to what is in the 040.library for the same purpose. However, since this code is in the internal ROM (not kickrom or LIBS: ) the FPU will be 882-compatible immediately after power-on.
grond is online now  
Old 08 October 2016, 22:36   #191
matthey
Registered User
 
Join Date: Jan 2010
Location: Kansas
Posts: 1,006
Quote:
Originally Posted by grond View Post
Only the 040 FPU instructions will be implemented in hardware. The 882-only instructions will cause an internal trap that will result in emulation code being executed from the FPGA-ROM. This emulation code basically corresponds to what is in the 040.library for the same purpose. However, since this code is in the internal ROM (not kickrom or LIBS: ) the FPU will be 882-compatible immediately after power-on.
Ok, this sounds reasonable and similar to placing the 68040.library or 68060.library in the kickstart. I would hope they would implement the FINT/FINTRZ instruction in hardware as these are common and there is no fast and precise way to avoid them (huge mistake of the 68040). The FPU sadly seems to be a much lower priority than the SIMD unit which will get much less use.
matthey is online now  
Old 08 October 2016, 23:14   #192
Locutus
Registered User

 
Join Date: Jul 2014
Location: Finland
Posts: 612
Quote:
Originally Posted by matthey View Post
Ok, this sounds reasonable and similar to placing the 68040.library or 68060.library in the kickstart. I would hope they would implement the FINT/FINTRZ instruction in hardware as these are common and there is no fast and precise way to avoid them (huge mistake of the 68040). The FPU sadly seems to be a much lower priority than the SIMD unit which will get much less use.
Kek Kek Kek, that will be the killer feature of h265 on vampire! >;-)
Locutus is offline  
Old 08 October 2016, 23:41   #193
grond
Registered User

 
Join Date: Jun 2015
Location: Germany
Posts: 184
Quote:
Originally Posted by matthey View Post
The FPU sadly seems to be a much lower priority than the SIMD unit which will get much less use.
I believe part of the reason for this is that Gunnar does the SIMD while Chris did all or most of the FPU and is currently busy with something else.
grond is online now  
Old 09 October 2016, 00:24   #194
eXeler0
Registered User

eXeler0's Avatar
 
Join Date: Feb 2015
Location: Sweden
Age: 43
Posts: 1,391
Quote:
Originally Posted by grond View Post
I believe part of the reason for this is that Gunnar does the SIMD while Chris did all or most of the FPU and is currently busy with something else.
And as Gunnar explained, AMMX was simple to implement since its:
1. Pretty much a copy of intel MMX, so not a huge challenge
2. Well documented by Intel, advantage for coders
3. there are image and video codecs that are already optimized for MMX wich gives immediate benefits for Amiga usage as well..
eXeler0 is offline  
Old 09 October 2016, 01:59   #195
Locutus
Registered User

 
Join Date: Jul 2014
Location: Finland
Posts: 612
Quote:
Originally Posted by eXeler0 View Post
And as Gunnar explained, AMMX was simple to implement since its:
1. Pretty much a copy of intel MMX, so not a huge challenge
2. Well documented by Intel, advantage for coders
3. there are image and video codecs that are already optimized for MMX wich gives immediate benefits for Amiga usage as well..
Except:

1: The opcodes don't function like MMX
2. How does documentation for a different ISA help?
3: How is that code portable?
Locutus is offline  
Old 09 October 2016, 05:09   #196
idrougge
Registered User
 
Join Date: Sep 2007
Location: Stockholm
Posts: 2,098
What's really sad about the Amibay sales is that many of those 68060 CPUs end up not on fresh 060 accelerators, but on adapters on tired old 3640 cards, postponing any new 060 cards.
idrougge is offline  
Old 09 October 2016, 09:23   #197
britelite
Registered User
 
Join Date: Feb 2010
Location: Espoo / Finland
Posts: 225
Quote:
Originally Posted by eXeler0 View Post
And while it "most definitely" will not be getting any 080 optimized demos from you, I wouldn't be so confident in predicting the future for everyone else in that matter. 😜
Great to hear that you're going to be making demos for the Vampire, looking forward to seeing them.
britelite is offline  
Old 09 October 2016, 10:16   #198
grond
Registered User

 
Join Date: Jun 2015
Location: Germany
Posts: 184
Demos are useless stuff anyway, wasting the time of the coder and the audience. Rather go code a game with a storyline which can entertain many hours and not just a few minutes.
grond is online now  
Old 09 October 2016, 11:06   #199
britelite
Registered User
 
Join Date: Feb 2010
Location: Espoo / Finland
Posts: 225
Quote:
Originally Posted by grond View Post
Rather go code a game with a storyline which can entertain many hours and not just a few minutes.
And how many games have you coded?
britelite is offline  
Old 09 October 2016, 11:42   #200
grond
Registered User

 
Join Date: Jun 2015
Location: Germany
Posts: 184
Quote:
Originally Posted by britelite View Post
And how many games have you coded?
None, I only coded useless demo stuff.
grond is online now  
AdSense AdSense  
 


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

Similar Threads
Thread Thread Starter Forum Replies Last Post
Elbox Products Back In Stock At AmigaKit.com amigakit.com MarketPlace 1 29 November 2012 02:27
New GVP Products now available Softhutjoe News 33 05 June 2010 11:09
Ok ---Im ready to be transported back.back.back to the 90s. Check out what Ive got. nc88keyz Amiga scene 9 12 January 2008 23:58
The 50 Best Tech Products of All Time rbelk Retrogaming General Discussion 19 04 April 2007 19:25
Asimware products... Amiga_Rob request.Apps 7 12 November 2004 19:45

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +2. The time now is 09:46.


Powered by vBulletin® Version 3.8.8 Beta 1
Copyright ©2000 - 2017, vBulletin Solutions, Inc.
Page generated in 0.26204 seconds with 12 queries