Poster | Thread |
Fairdinkem
| |
Apollo Team announce new accelerators Posted on 5-Jul-2021 5:15:15
| | [ #1 ] |
|
|
|
Cult Member |
Joined: 23-Feb-2010 Posts: 517
From: Victoria, Australia | | |
|
| |
Status: Offline |
|
|
pavlor
| |
Re: Apollo Team announce new accelerators Posted on 5-Jul-2021 5:27:20
| | [ #2 ] |
|
|
|
Elite Member |
Joined: 10-Jul-2005 Posts: 9639
From: Unknown | | |
|
| |
Status: Offline |
|
|
Hammer
| |
Re: Apollo Team announce new accelerators Posted on 5-Jul-2021 6:30:42
| | [ #3 ] |
|
|
|
Elite Member |
Joined: 9-Mar-2003 Posts: 5976
From: Australia | | |
|
| @Fairdinkem
That's nice. _________________ Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68) Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68) Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB |
|
Status: Offline |
|
|
kolla
| |
Re: Apollo Team announce new accelerators Posted on 5-Jul-2021 8:11:56
| | [ #4 ] |
|
|
|
Elite Member |
Joined: 20-Aug-2003 Posts: 3261
From: Trondheim, Norway | | |
|
| @Fairdinkem
The title is misleading, it should really be "The Vampire creator leaves the Apollo Team"
I suppose this means that the A1200 "Raptor" is cancelled, and all previous "roadmaps" are binned. Again. It's now back to "V4 for everybody!!111". Last edited by kolla on 05-Jul-2021 at 08:19 AM.
_________________ B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC |
|
Status: Offline |
|
|
Fairdinkem
| |
Re: Apollo Team announce new accelerators Posted on 5-Jul-2021 9:19:38
| | [ #5 ] |
|
|
|
Cult Member |
Joined: 23-Feb-2010 Posts: 517
From: Victoria, Australia | | |
|
| @kolla
Hi Kolla,
I am not sure what you are referring to as I have not read the forums (I found this on my Facebook feed from Apollo Team) so I am unaware of any politics within the Apollo team.
I just hope to get a powerful accelerator for my Amiga 1200. I would love a WARP 1260 but the chances of me sourcing a 68060 rev6 is next to 0% unless I mortgage my house.
I hope a PiStorm for A1200 magically appears, but that doesn't seem to be likely anytime soon that I am aware of. The Buffee project seems to be only in the alpha stage at best and only for Amiga A500, A600, A1000 etc, so I really hope despite any politics within Apollo Team that these new boards are produced quickly and are of good quality and worth the money. Last edited by Fairdinkem on 05-Jul-2021 at 09:21 AM.
_________________ Amiga A1200T - TF1260 - R9200 - Indivision AGA MK3 Amiga A500 - PiStorm EMU68 Pegasos 2 G4 - AmigaOS 4.1 FE / MorphOS 3.16 |
|
Status: Offline |
|
|
Samurai_Crow
| |
Re: Apollo Team announce new accelerators Posted on 5-Jul-2021 14:09:14
| | [ #6 ] |
|
|
|
Elite Member |
Joined: 18-Jan-2003 Posts: 2320
From: Minnesota, USA | | |
|
| @Fairdinkem
Igor is leaving the team but Thomas Hirsch is back. He developed the original SAGA core back in the NatAmi days. |
|
Status: Offline |
|
|
BigD
| |
Re: Apollo Team announce new accelerators Posted on 5-Jul-2021 15:46:47
| | [ #7 ] |
|
|
|
Elite Member |
Joined: 11-Aug-2005 Posts: 7449
From: UK | | |
|
| @Samurai_Crow
So the Raptor is gone and we now have V4 spec hardware with SAGA, Ethernet and usb joypad support. I assume the price is therefore higher? _________________ "Art challenges technology. Technology inspires the art." John Lasseter, Co-Founder of Pixar Animation Studios |
|
Status: Offline |
|
|
matthey
| |
Re: Apollo Team announce new accelerators Posted on 6-Jul-2021 1:51:43
| | [ #8 ] |
|
|
|
Elite Member |
Joined: 14-Mar-2007 Posts: 2379
From: Kansas | | |
|
| kolla Quote:
The title is misleading, it should really be "The Vampire creator leaves the Apollo Team"
|
Majsta is still shown as an Apollo team member on the forum. No more Vampire branding without Majsta?
Samurai_Crow Quote:
Igor is leaving the team but Thomas Hirsch is back. He developed the original SAGA core back in the NatAmi days. |
SAGA doesn't exist and Thomas Hirsch has no contact with the Apollo team. It's all a hoax.
BigD Quote:
So the Raptor is gone and we now have V4 spec hardware with SAGA, Ethernet and usb joypad support. I assume the price is therefore higher? |
The price will likely go up but the hardware may offer similar value. Using the same sized FPGA on all V4 spec hardware should make development easier and allow for more frequent updates. I wouldn't be surprised if Amiga users with older Vampire hardware get fewer updates.
|
|
Status: Offline |
|
|
agami
| |
Re: Apollo Team announce new accelerators Posted on 6-Jul-2021 7:43:42
| | [ #9 ] |
|
|
|
Super Member |
Joined: 30-Jun-2008 Posts: 1841
From: Melbourne, Australia | | |
|
| I had a feeling about these guys. I see lots of potential and still hope to do business one day.
I already have the V2 600, V2 1200, and the V4 Standalone. I will definitely get the IceDrake V4 1200 and I look forward to the A4000 card. If they ever release one for the CD32, I will get that too.
_________________ All the way, with 68k |
|
Status: Offline |
|
|
DrProcton
| |
Re: Apollo Team announce new accelerators Posted on 6-Jul-2021 7:45:43
| | [ #10 ] |
|
|
|
Member |
Joined: 18-Feb-2017 Posts: 61
From: Unknown | | |
|
| @matthey You should be more careful before telling that 'it's all a hoax'. I'd like to inform you that Thomas is back, click the link and stop speculations if you're not informed, please
Apollo team 2021 Last edited by DrProcton on 06-Jul-2021 at 07:54 AM.
|
|
Status: Offline |
|
|
matthey
| |
Re: Apollo Team announce new accelerators Posted on 6-Jul-2021 11:54:33
| | [ #11 ] |
|
|
|
Elite Member |
Joined: 14-Mar-2007 Posts: 2379
From: Kansas | | |
|
| DrProcton Quote:
You should be more careful before telling that 'it's all a hoax'. I'd like to inform you that Thomas is back, click the link and stop speculations if you're not informed, please
|
Sorry you didn't get it. It was a joke. Certain users here have been trolling me about how SAGA doesn't exist and how Gunnar could never have obtained permission to use SAGA from Thomas. There were all kinds of conspiracy theories circulating which I tried to dispel with my inside knowledge from when I was one of the early members of the Apollo Team (Gunnar, Jens, Chris, Meynaf, Rune and I). In fact, I played a role in Gunnar obtaining SAGA by suggesting adding HDMI/DVI RTG/AGA to accelerators for ECS Amigas to increase value.
|
|
Status: Offline |
|
|
Lou
| |
Re: Apollo Team announce new accelerators Posted on 6-Jul-2021 18:34:37
| | [ #12 ] |
|
|
|
Elite Member |
Joined: 2-Nov-2004 Posts: 4228
From: Rhode Island | | |
|
| @matthey
Quote:
matthey wrote: DrProcton Quote:
You should be more careful before telling that 'it's all a hoax'. I'd like to inform you that Thomas is back, click the link and stop speculations if you're not informed, please
|
Sorry you didn't get it. It was a joke. Certain users here have been trolling me about how SAGA doesn't exist and how Gunnar could never have obtained permission to use SAGA from Thomas. There were all kinds of conspiracy theories circulating which I tried to dispel with my inside knowledge from when I was one of the early members of the Apollo Team (Gunnar, Jens, Chris, Meynaf, Rune and I). In fact, I played a role in Gunnar obtaining SAGA by suggesting adding HDMI/DVI RTG/AGA to accelerators for ECS Amigas to increase value.
|
What did I say that was a hoax? I said Thomas H is who did the chipset on NATAMI and that's why it took so long for the standalone to come out since Gunnar only had access to the cpu when they split. SAGA has NOTHING to do with AGA. Are OCS/ECS/AGA 'registers' in chipram or are they on the cpu? Where are SAGA registers? Creating a signal generator for HD output from fastram does not a chipset make.Last edited by Lou on 06-Jul-2021 at 06:36 PM. Last edited by Lou on 06-Jul-2021 at 06:35 PM.
|
|
Status: Offline |
|
|
QuikSanz
| |
Re: Apollo Team announce new accelerators Posted on 6-Jul-2021 19:07:03
| | [ #13 ] |
|
|
|
Super Member |
Joined: 28-Mar-2003 Posts: 1236
From: Harbor Gateway, Gardena, Ca. | | |
|
| @Lou<
DrProcton said that, not you Lou. Your OK.
|
|
Status: Offline |
|
|
matthey
| |
Re: Apollo Team announce new accelerators Posted on 6-Jul-2021 19:52:20
| | [ #14 ] |
|
|
|
Elite Member |
Joined: 14-Mar-2007 Posts: 2379
From: Kansas | | |
|
| Lou Quote:
What did I say that was a hoax?
|
SAGA
Lou Quote:
I mentioned implementing AKIKO and enhancing it and in so doing found out SAGA is nothing more than a framebuffer for "HDMI out" [RTG] and some software running on the 2nd thread of the cpu. You've presented nothing to dispute this. As I said - it was so stated my at least one developer. Gunnar was supposed to 'open-source' SAGA...where is it? |
https://amigaworld.net/modules/newbb/viewtopic.php?topic_id=44218&forum=14&start=40&viewmode=flat&order=0#842914
Lou Quote:
I said Thomas H is who did the chipset on NATAMI and that's why it took so long for the standalone to come out since Gunnar only had access to the cpu when they split.
|
Gunnar likely had the SAGA code (originally AHDL as I recall but Gunnar likely converted to VHDL) from Natami development. He just had to get permission from Thomas to use it. I practically stated this earlier.
Lou Quote:
SAGA has NOTHING to do with AGA. Are OCS/ECS/AGA 'registers' in chipram or are they on the cpu? Where are SAGA registers? Creating a signal generator for HD output from fastram does not a chipset make. |
SAGA=AGA with enhancements like a chunky RTG layer under all other gfx, more chip memory, a programmable display pixel clock (Thomas enhancement during Natami days), 8 channel 16 bit sound from AAA & AA+, double the number of sprites with more colors and fewer limitations, etc.
I don't believe SAGA is in the CPU cores. SAGA should be separate logic including separate coprocessor cores of the Copper and Blitter although Gunnar likely removed part of the Blitter. The CPU cores, coprocessor cores and SAGA registers would be attached to the memory bus of the memory controller (it would be possible to have separate memory controllers for fast and chip memory but CBM, Natami and Vampires use one as far as I know). My perspective of the CPU and coprocessor cores are of logic units connected to the memory bus driven by a clock while the SAGA registers with logic are generally simpler logic. Take the simple Akiko hardware c2p registers for example. There would be logic to connect to the memory bus including latches (a series of latching flip flops retaining each bit) to lock values written to the registers. This would be connected to logic which converts the data and is attached to another latching register connected to the memory bus on the other side of the logic. No CPU core is needed. I am not a hardware guy so my understanding and especially terminology are likely not the best. There are probably guys with a hardware background on this forum who could explain it better. Feel free to correct me and educate us. Education helps to flush out these conspiracy theories.
Last edited by matthey on 06-Jul-2021 at 11:51 PM. Last edited by matthey on 06-Jul-2021 at 08:02 PM. Last edited by matthey on 06-Jul-2021 at 07:53 PM.
|
|
Status: Offline |
|
|
Lou
| |
Re: Apollo Team announce new accelerators Posted on 6-Jul-2021 20:55:40
| | [ #15 ] |
|
|
|
Elite Member |
Joined: 2-Nov-2004 Posts: 4228
From: Rhode Island | | |
|
| @matthey
Quote:
matthey wrote: Lou Quote:
What did I say that was a hoax?
|
SAGA
Lou Quote:
I mentioned implementing AKIKO and enhancing it and in so doing found out SAGA is nothing more than a framebuffer for "HDMI out" [RTG] and some software running on the 2nd thread of the cpu. You've presented nothing to dispute this. As I said - it was so stated my at least one developer. Gunnar was supposed to 'open-source' SAGA...where is it? |
https://amigaworld.net/modules/newbb/viewtopic.php?topic_id=44218&forum=14&start=40&viewmode=flat&order=0#842914
Lou Quote:
I said Thomas H is who did the chipset on NATAMI and that's why it took so long for the standalone to come out since Gunnar only had access to the cpu when they split.
|
Gunnar likely had the SAGA code (originally AHDL as I recall but Gunnar likely converted to VHDL) from Natami development. He just had to get permission from Thomas to use it. I practically stated this earlier.
Lou Quote:
SAGA has NOTHING to do with AGA. Are OCS/ECS/AGA 'registers' in chipram or are they on the cpu? Where are SAGA registers? Creating a signal generator for HD output from fastram does not a chipset make. |
SAGA=AGA with enhancements like a chunky RTG layer under all other gfx, a programmable display pixel clock (Thomas enhancement during Natami days), 8 channel 16 bit sound from AAA & AA+, double the number of sprites with more colors and fewer limitations, etc.
I don't believe SAGA is in the CPU cores. SAGA should be separate logic including separate coprocessor cores of the Copper and Blitter although Gunnar likely removed part of the Blitter. The CPU cores, coprocessor cores and SAGA registers would be attached to the memory bus of the memory controller (it would be possible to have separate memory controllers for fast and chip memory but CBM, Natami and Vampires use one as far as I know). My perspective of the CPU and coprocessor cores are of logic units connected to the memory bus driven by a clock while the SAGA registers with logic are generally simpler logic. Take the simple Akiko hardware c2p registers for example. There would be logic to connect to the memory bus including latches (a series of latching flip flops retaining each bit) to lock values written to the registers. This would be connected to logic which converts the data and is attached to another latching register connected to the memory bus on the other side of the logic. No CPU core is needed. I am not a hardware guy so my understanding and especially terminology are likely not the best. There are probably guys with a hardware background on this forum who could explain it better. Feel free to correct me and educate us. Education helps to flush out these conspiracy theories.
|
You don't 'believe'. That's all you had to say.
SAGA is still not open-sourced. You linked to a post of me asking where it is. Where is it? I see CPU registers but a DEDICATED CHIP wouldn't have CPU REGISTERS, it would have registers shadowed in a bank of RAM like every other piece of hardware that exists.
I believe Thomas Hirsch has the ability to recreate and enhance AGA, but the 'SAGA' within the Vampire line is nothing more than routines operating on fast-ram in the extra thread of the cpu unused by the OS...so it feels like free processing.Last edited by Lou on 06-Jul-2021 at 08:56 PM.
|
|
Status: Offline |
|
|
BigD
| |
Re: Apollo Team announce new accelerators Posted on 6-Jul-2021 21:02:27
| | [ #16 ] |
|
|
|
Elite Member |
Joined: 11-Aug-2005 Posts: 7449
From: UK | | |
|
| @Lou
That's why the increased cost over the previously announced Raptor matters. We now have feature creep over and above the 080 accelerator and Ram to include the SAGA which may be a damp squib. _________________ "Art challenges technology. Technology inspires the art." John Lasseter, Co-Founder of Pixar Animation Studios |
|
Status: Offline |
|
|
kolla
| |
Re: Apollo Team announce new accelerators Posted on 6-Jul-2021 22:17:45
| | [ #17 ] |
|
|
|
Elite Member |
Joined: 20-Aug-2003 Posts: 3261
From: Trondheim, Norway | | |
|
| @BigD
Quote:
BigD wrote: We now have feature creep over and above the 080 accelerator and Ram to include the SAGA which may be a damp squib. |
This "creep" is going backwards - what they announce now is pretty much what was announced with the original V4... that's why the V4 has 68000 socket pins on the board. So this announcement is like 5 years ago all over again.
(And I am still waiting for someone to point me to video snippet of a V4 playing for example Slamtilt with multiball and interlaced screen mode - there are many videos of Slamtil on even V4, but Vampire owners seem to suck at pinball, so they never reach multiball... heh)Last edited by kolla on 06-Jul-2021 at 10:20 PM.
_________________ B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC |
|
Status: Offline |
|
|
matthey
| |
Re: Apollo Team announce new accelerators Posted on 6-Jul-2021 22:33:38
| | [ #18 ] |
|
|
|
Elite Member |
Joined: 14-Mar-2007 Posts: 2379
From: Kansas | | |
|
| Lou Quote:
You don't 'believe'. That's all you had to say.
|
I don't say for sure what I don't know for sure.
Lou Quote:
SAGA is still not open-sourced. You linked to a post of me asking where it is. Where is it? I see CPU registers but a DEDICATED CHIP wouldn't have CPU REGISTERS, it would have registers shadowed in a bank of RAM like every other piece of hardware that exists.
|
I don't know why the SAGA sources aren't available to the public as I have nothing to do with it. Where do you see CPU registers?
CPU core registers are in the CPU core but usually not directly accessible on the memory bus (some old architectures have low memory addresses available as CPU registers but I don't know if they would be available to another processor using the memory bus). SAGA registers should be external from the CPU core and in SAGA connected to the memory bus. Everything is in one FPGA chip and memory blocks in the FPGA may be near each other so the registers are physically close but the connections should make them independent. This is the problem of an FPGA which is routing inefficiency and mismatched resources. In an ASIC SoC, the CPU core logic would be further from the SAGA logic rather than scattered about and intermingled in an FPGA.
Lou Quote:
I believe Thomas Hirsch has the ability to recreate and enhance AGA, but the 'SAGA' within the Vampire line is nothing more than routines operating on fast-ram in the extra thread of the cpu unused by the OS...so it feels like free processing. |
Recall that Thomas used real 68k CPUs to debug the Natami hardware and SAGA in the beginning where there was no 2nd thread. It would be possible to setup emulation inside an FPGA using a processor to emulate the Amiga custom chips but it would give poor performance as the custom chips are the most difficult part of the Amiga to emulate requiring at least a medium performance CPU core to avoid slow downs and glitches. Since SAGA was already mostly working for Thomas, it is logical to assume that Gunnar would use it instead of emulation on the 2nd CPU thread.
BigD Quote:
That's why the increased cost over the previously announced Raptor matters. We now have feature creep over and above the 080 accelerator and Ram to include the SAGA which may be a damp squib. |
SAGA takes little space in the FPGA. I would guess no more than 20% of the resources and likely less. This would be maybe $5-$10 U.S. to include.
Last edited by matthey on 06-Jul-2021 at 11:18 PM. Last edited by matthey on 06-Jul-2021 at 10:37 PM.
|
|
Status: Offline |
|
|
kolla
| |
Re: Apollo Team announce new accelerators Posted on 7-Jul-2021 0:39:43
| | [ #19 ] |
|
|
|
Elite Member |
Joined: 20-Aug-2003 Posts: 3261
From: Trondheim, Norway | | |
|
| @matthey
If Thomas made SAGA, how could Gunnar announce that he and the team had decided to make SAGA open source? If it REALLY was the work of Thomas, then Gunnar was in no place to make such an announcement.
What if there are several quite different plans and implementations that have been sailing under the "SAGA" name up through the years... similar to the "FPU". Last edited by kolla on 07-Jul-2021 at 12:46 AM.
_________________ B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC |
|
Status: Offline |
|
|
matthey
| |
Re: Apollo Team announce new accelerators Posted on 7-Jul-2021 1:05:54
| | [ #20 ] |
|
|
|
Elite Member |
Joined: 14-Mar-2007 Posts: 2379
From: Kansas | | |
|
| kolla Quote:
If Thomas made SAGA, how could Gunnar announce that he and the team had decided to make SAGA open source? If it REALLY was the work of Thomas, then Gunnar was in no place to make such an announcement.
|
Thomas could have put conditions on Gunnar using SAGA. An obvious one would be that Gunnar return bug fixes and enhancements. It's also possible that Thomas required Gunnar to open source SAGA as well.
kolla Quote:
What if there are several quite different plans and implementations that have been sailing under the "SAGA" name up through the years... similar to the "FPU". |
I expect there are different versions of SAGA for different target hardware with different functionality, features and bugs. Referring to SAGA usually means the most advanced and bug free version available. I would hope we are done with too small of FPGAs but even that is relative as the FPU is still a cut down version in the biggest FPGAs (precision and instructions compared to a 68060).
|
|
Status: Offline |
|
|