Poster | Thread |
MEGA_RJ_MICAL
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 16-Aug-2022 11:43:15
| | [ #41 ] |
|
|
|
Super Member |
Joined: 13-Dec-2019 Posts: 1200
From: AMIGAWORLD.NET WAS ORIGINALLY FOUNDED BY DAVID DOYLE | | |
|
| |
Status: Offline |
|
|
Karlos
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 16-Aug-2022 12:04:29
| | [ #42 ] |
|
|
|
Elite Member |
Joined: 24-Aug-2003 Posts: 4394
From: As-sassin-aaate! As-sassin-aaate! Ooh! We forgot the ammunition! | | |
|
| |
Status: Offline |
|
|
MEGA_RJ_MICAL
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 16-Aug-2022 13:32:45
| | [ #43 ] |
|
|
|
Super Member |
Joined: 13-Dec-2019 Posts: 1200
From: AMIGAWORLD.NET WAS ORIGINALLY FOUNDED BY DAVID DOYLE | | |
|
| @Karlos
Friend Karlos,
I am sure it was, and IS, very very ample. Gigantic.
It could smother a Baron of Hell. A Mancubus, it could trample.
Offhand. _________________ I HAVE ABS OF STEEL -- CAN YOU SEE ME? CAN YOU HEAR ME? OK FOR WORK |
|
Status: Offline |
|
|
cdimauro
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 16-Aug-2022 18:44:08
| | [ #44 ] |
|
|
|
Elite Member |
Joined: 29-Oct-2012 Posts: 3621
From: Germany | | |
|
| @Hammer
Quote:
Hammer wrote: @cdimauro
Quote:
20FPS is a very good result: very fluid, albeit not full-motion. |
A1200's AGA is capable of full motion video with CSLabs Warp A1260's 68060 @ 100 Mhz. You're bullshit again.
|
Hey, quiche eater: what's you IQ level?
I was commenting HyperX's Amiga 4000 which, as HE stated, can only reach 20FPS in Doom!
So, you jumped-in on the thread, without reading what people said, so completely missing the context, and replying to my message with a total non-sense.
Have I said that Mother Nature hasn't done a good job with you? Ah, yes: already said. It looks like that shes was a very bad stepmother for you.. Last edited by cdimauro on 18-Aug-2022 at 04:49 PM.
|
|
Status: Offline |
|
|
Karlos
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 16-Aug-2022 20:29:52
| | [ #45 ] |
|
|
|
Elite Member |
Joined: 24-Aug-2003 Posts: 4394
From: As-sassin-aaate! As-sassin-aaate! Ooh! We forgot the ammunition! | | |
|
| |
Status: Offline |
|
|
Karlos
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 17-Aug-2022 8:15:42
| | [ #46 ] |
|
|
|
Elite Member |
Joined: 24-Aug-2003 Posts: 4394
From: As-sassin-aaate! As-sassin-aaate! Ooh! We forgot the ammunition! | | |
|
| @cdimauro
It may be my ageing memory but I'm sure I got better than 20fps with my 040/RTG combo. I won't be able to test that any time soon though given the machine is currently missing its HD and is likely full of dust, dried capacitors etc.
-edit- Damn you, autocorrect... Last edited by Karlos on 18-Aug-2022 at 08:25 PM.
_________________ Doing stupid things for fun... |
|
Status: Offline |
|
|
kolla
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 17-Aug-2022 9:43:46
| | [ #47 ] |
|
|
|
Elite Member |
Joined: 20-Aug-2003 Posts: 2859
From: Trondheim, Norway | | |
|
| @Karlos
“missing it’s HD”? Really?! _________________ B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC |
|
Status: Offline |
|
|
Karlos
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 17-Aug-2022 11:36:53
| | [ #48 ] |
|
|
|
Elite Member |
Joined: 24-Aug-2003 Posts: 4394
From: As-sassin-aaate! As-sassin-aaate! Ooh! We forgot the ammunition! | | |
|
| @kolla
Yes, it has a spinning rust drive that I took out some time ago to image with dd. Which I did, but I've not gotten round to putting it back in. I really ought to get something solid state. _________________ Doing stupid things for fun... |
|
Status: Offline |
|
|
Lou
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 17-Aug-2022 16:31:14
| | [ #49 ] |
|
|
|
Elite Member |
Joined: 2-Nov-2004 Posts: 4169
From: Rhode Island | | |
|
| So... It took you guys 30 years to realize a 3.57Mhz system bus speed wasn't going to cut it?
95+% of Vampire performance is DDR3 vs chipram speeds...
I guess the C65 having capabilities somewhat on par with Amiga wasn't a clue either... Last edited by Lou on 17-Aug-2022 at 04:35 PM.
|
|
Status: Offline |
|
|
Hammer
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 18-Aug-2022 5:43:06
| | [ #50 ] |
|
|
|
Elite Member |
Joined: 9-Mar-2003 Posts: 5246
From: Australia | | |
|
| @Karlos
Quote:
Karlos wrote: @Hammer
At up to 7MB/s, AGA should be capable of playing video, especially if it's already in a suitable format and streaming from memory.
However, if what you have is some modernish compressed video stream, with all the mpeg like compression tricks and YUV encoding streaming from disk, then I dunno. That's a big workload even before you decide how to convert your RGB pixels into something renderable on screen.
|
A very fast modern CPU with Fast (CPU) RAM can decode MPEG and translate multiple frames ahead of AGA's display capabilities.
Even the AC68080 solution still drive C= AGA at full motion video frame rates at VCD/VHS resolution.
At a certain point with fast CPUs, C= AGA can be a bottleneck.
https://www.youtube.com/watch?v=yfBOmXOKnKU Graffiti DAC coupled with A1200's 32-bit Chip RAM rendering Doom (DoomAttack) example. This Amiga 1200 has Ventisca 1230 III with 68030/68882 @ 50 Mhz and 32 MB 32-bit Fast RAM.
My Amiga1200/TF1260 has Indivision AGA Mk3 (with Graffiti) and I replaced the built-in RF modulator for Indivision's HDMI.
Last edited by Hammer on 18-Aug-2022 at 03:29 PM.
_________________ Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68) Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68) |
|
Status: Offline |
|
|
bhabbott
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 18-Aug-2022 6:04:57
| | [ #51 ] |
|
|
|
Regular Member |
Joined: 6-Jun-2018 Posts: 330
From: Aotearoa | | |
|
| @Lou
Quote:
Lou wrote: So... It took you guys 30 years to realize a 3.57Mhz system bus speed wasn't going to cut it? |
It 'cuts it' just fine. My A1200 runs Doom as fast as my 386DX-40 used to, and that's fast enough for me.
Quote:
95+% of Vampire performance is DDR3 vs chipram speeds... |
Nonsense. That DDR3 RAM would be useless without an fpga powerful enough to realize its speed (including a fast CPU core etc.). The Vampire is as fast as it is because everything has been tuned for best performance.
But this thread isn't about DDR3 RAM or Vampires. It's about whether packed pixels could have had a significant advantage on 1985-1992 hardware. Many Amiga fans have been saying that it would, some even suggesting the improvement would be huge. Yet Hypex's tests confirm what some of us suspected - with a fast CPU the difference is minimal (and with a slow CPU it's irrelevant). That means a lot of criticism of the Amiga was off-base, and it also means producing some kind of chunky conversion hardware is probably not worth the effort. Maybe now Amiga fans will stop pining for a mythical chunky graphics mode and enjoy the Amiga for what it is.
Who am I kidding? Amiga fans will never be satisfied with what they have.
Quote:
I guess the C65 having capabilities somewhat on par with Amiga wasn't a clue either... |
It would be wonderful if the C65 had "capabilities somewhat on par with Amiga" but sadly it was never released. Of course given what chips were available back then it wouldn't be surprising if the RAM speed was similar on both machines. Why would you expect otherwise?
|
|
Status: Offline |
|
|
Karlos
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 18-Aug-2022 6:52:48
| | [ #52 ] |
|
|
|
Elite Member |
Joined: 24-Aug-2003 Posts: 4394
From: As-sassin-aaate! As-sassin-aaate! Ooh! We forgot the ammunition! | | |
|
| @bhabbott
I don't think your conclusions are correct. Packed pixels even on OCS still carry the advantage that you need fewer read/write cycles for pretty much any CPU access. The most marked difference is in the simplest operation, such as plotting a single pixel. There's nothing hypothetical about requiring N read/write accesses versus 1 (or 2 in the worst case of spanning a word boundary) for that task.
Almost every drawing operation on a N-bit packed pixel display is similar to the work required to draw on a single bit plane. This is true whether or not it's the CPU or the blitter doing it.
As for users never being satisfied, maybe you have a point :)
In my defence though, when it was apparent AGA was just not up to the things I needed it for, I didn't sell out and get a PC. I got an RTG card instead and carried on using my A1200 as my main machine into the nougties. I only caved in and built myself a PC in late 2007. Last edited by Karlos on 18-Aug-2022 at 07:04 AM.
_________________ Doing stupid things for fun... |
|
Status: Offline |
|
|
kolla
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 18-Aug-2022 12:32:15
| | [ #53 ] |
|
|
|
Elite Member |
Joined: 20-Aug-2003 Posts: 2859
From: Trondheim, Norway | | |
|
| @Karlos
I was more concerned about the implications of your spelling of “its”. _________________ B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC |
|
Status: Offline |
|
|
Lou
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 18-Aug-2022 13:10:04
| | [ #54 ] |
|
|
|
Elite Member |
Joined: 2-Nov-2004 Posts: 4169
From: Rhode Island | | |
|
| |
Status: Offline |
|
|
Karlos
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 18-Aug-2022 16:01:45
| | [ #55 ] |
|
|
|
Elite Member |
Joined: 24-Aug-2003 Posts: 4394
From: As-sassin-aaate! As-sassin-aaate! Ooh! We forgot the ammunition! | | |
|
| |
Status: Offline |
|
|
Hammer
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 18-Aug-2022 16:22:06
| | [ #56 ] |
|
|
|
Elite Member |
Joined: 9-Mar-2003 Posts: 5246
From: Australia | | |
|
| @bhabbott
Quote:
Nonsense. That DDR3 RAM would be useless without an fpga powerful enough to realize its speed (including a fast CPU core etc.). The Vampire is as fast as it is because everything has been tuned for best performance.
But this thread isn't about DDR3 RAM or Vampires. It's about whether packed pixels could have had a significant advantage on 1985-1992 hardware. Many Amiga fans have been saying that it would, some even suggesting the improvement would be huge. Yet Hypex's tests confirm what some of us suspected - with a fast CPU the difference is minimal (and with a slow CPU it's irrelevant). That means a lot of criticism of the Amiga was off-base, and it also means producing some kind of chunky conversion hardware is probably not worth the effort. Maybe now Amiga fans will stop pining for a mythical chunky graphics mode and enjoy the Amiga for what it is.
Who am I kidding? Amiga fans will never be satisfied with what they have.
|
Apollo V4's 1GB/s DDR3 would be about 1332 MT/s (666 Mhz) 16 bit from ISSI IS43TR16256A or Micron ?????D9SFD.
V4's 1332 MT/s memory bandwidth would be Pentium IIIS era i.e 133 Mhz x 64 bit = 1064 MB/s.
No PC with a Pentium II class CPU was coupled with Pentium IIIS's 133 Mhz FSB and memory bandwidth. Pentium III motherboards have access to RDRAM (Rambus).
K7 Athlons Slot A has 64 bit DDR FSB e.g. 200 MTs x 64 bit = 1600 MB/s
----- For Doom demo 3 via https://www.vogons.org/viewtopic.php?t=43551 [img]https://www.vogons.org/download/file.php?id=18064&mode=view[/img] Cyrix 5x86 with 486's 40 Mhz 32bit FSB with VIA FIC486-VIP-IO chipset and S3 Trio 64U yields 29.52 fps
With SIS chipset, c5x86 (LS486e@2x66) with Matrox Storm has 58 fps. The VIA chipset seems slow compared to the SIS chipset.
[img]https://www.vogons.org/download/file.php?id=18084&mode=view[/img] Pentium 133 with S3 Trio 64U has 74.17 fps Pentium 133 MMX with S3 Trio 64U has 92.12 fps
Hyperx's CyberGraphics 64's S3 Trio 64U is being bottlenecked by a pipe between CPU and SVGA card.
Last edited by Hammer on 18-Aug-2022 at 04:41 PM. Last edited by Hammer on 18-Aug-2022 at 04:36 PM. Last edited by Hammer on 18-Aug-2022 at 04:27 PM. Last edited by Hammer on 18-Aug-2022 at 04:25 PM. Last edited by Hammer on 18-Aug-2022 at 04:22 PM.
_________________ Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68) Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68) |
|
Status: Offline |
|
|
cdimauro
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 18-Aug-2022 17:35:03
| | [ #57 ] |
|
|
|
Elite Member |
Joined: 29-Oct-2012 Posts: 3621
From: Germany | | |
|
| @Karlos
Quote:
Karlos wrote: @cdimauro
It may be my ageing memory but I'm sure I got better than 20fps with my 040/RTG combo. I won't be able to test that any time soon though given the machine is currently missing it's HD and is likely full of dust, dried capacitors etc |
Then HypeX's system has some problem, since it's able to achieve the same with its 060+RTG.
@Lou
Quote:
Lou wrote: So... It took you guys 30 years to realize a 3.57Mhz system bus speed wasn't going to cut it? |
"A bit" before... Quote:
95+% of Vampire performance is DDR3 vs chipram speeds... |
Absolutely no. It gives a good help, but most is thanks to its 68080. Quote:
I guess the C65 having capabilities somewhat on par with Amiga wasn't a clue either... |
I don't think so. The C65 was still very limited. AFAIR its Blitter has also less functionalities compared to Amiga one.
Anyway, it was never sold, so: who cares!
@bhabbott
Quote:
bhabbott wrote: @Lou
Quote:
Lou wrote: So... It took you guys 30 years to realize a 3.57Mhz system bus speed wasn't going to cut it? |
It 'cuts it' just fine. My A1200 runs Doom as fast as my 386DX-40 used to, and that's fast enough for me. |
Any timedemo benchmark for it (and full config specs)? Quote:
It's about whether packed pixels could have had a significant advantage on 1985-1992 hardware. Many Amiga fans have been saying that it would, some even suggesting the improvement would be huge. |
There's no need of people's opinions: math is and should be enough to prove it. Quote:
Yet Hypex's tests confirm what some of us suspected - with a fast CPU the difference is minimal |
Let me recall why the Amiga was famous and his fans continuously trumpeted it: because of its chipset which was able to perform operations much faster than the CPU.
Whereas the finger pointing was towards PCs because they had nothing like that, and it required a fast CPU to achieve the same goals.
Please, let me know if I recall wrong.
If not, take your sentence and tell me if you recognize some similarity with that... Quote:
(and with a slow CPU it's irrelevant). |
If it's irrelevant then I assume that Doom could run well on a stock Amiga 1200. Or on an expanded a Amiga 500 with its plain 68000.
A slow CPU is irrelevant, right? Quote:
That means a lot of criticism of the Amiga was off-base, and it also means producing some kind of chunky conversion hardware is probably not worth the effort. |
Both wrong. And continuously repeating it don't make it true. Rather, it makes you a parrot... Quote:
Maybe now Amiga fans will stop pining for a mythical chunky graphics mode |
Maybe you can stop writing utterly bullsh*ts, instead.
However it might be that you're too limited to understand the topic. And looking at the non-sense that you continue to write like a broken disk, then it's likely the case.
Computer science isn't something for you: better that you switch to farming (hoping that you don't do disasters there as well)... Quote:
and enjoy the Amiga for what it is. |
That's what we did.
But maybe the topic is a bit different, eh?
And, even more important, the two things are NOT mutually exclusive; elementary logic at the hands. But logic isn't your friend, as we know...
@Karlos
Quote:
Karlos wrote: @bhabbott
I don't think your conclusions are correct. Packed pixels even on OCS still carry the advantage that you need fewer read/write cycles for pretty much any CPU access. The most marked difference is in the simplest operation, such as plotting a single pixel. There's nothing hypothetical about requiring N read/write accesses versus 1 (or 2 in the worst case of spanning a word boundary) for that task.
Almost every drawing operation on a N-bit packed pixel display is similar to the work required to draw on a single bit plane. This is true whether or not it's the CPU or the blitter doing it. |
+2
@Hammer
Quote:
Hammer wrote: Pentium III motherboards have access to RDRAM (Rambus). |
https://en.wikipedia.org/wiki/List_of_Intel_chipsets#Pentium_II/III_chipsets |
|
Status: Offline |
|
|
kolla
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 18-Aug-2022 20:01:03
| | [ #58 ] |
|
|
|
Elite Member |
Joined: 20-Aug-2003 Posts: 2859
From: Trondheim, Norway | | |
|
| @Karlos
Do you also spell “his” as “he’s” and feel awkward about … hohum… female possessive “her” without any s to put your precious apostrophe in front of?
I suppose English isn’t your (notice the absence of possessive apostrophe and s) native language. _________________ B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC |
|
Status: Offline |
|
|
Karlos
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 18-Aug-2022 20:12:12
| | [ #59 ] |
|
|
|
Elite Member |
Joined: 24-Aug-2003 Posts: 4394
From: As-sassin-aaate! As-sassin-aaate! Ooh! We forgot the ammunition! | | |
|
| @kolla
No, that's what I thought you were suggesting because I wrote "its" and not "it's", which would be a verbal contraction of "it is".
-edit- apparently not, lol. Stupid phone.
Would you care to elaborate on what you found concerning about the use of the expression "missing its HD" when talking about my A1200? I don't think computer is a gendered noun in English, though I could be wrong.
On your other point, you are correct. I'm a Geordie (just about) and so English as is commonly thought of may be considered a second language. Our regional dialect has significant divergence. Last edited by Karlos on 18-Aug-2022 at 08:34 PM.
_________________ Doing stupid things for fun... |
|
Status: Offline |
|
|
Karlos
| |
Re: An experimental Doom speed test and feasibility study based on a virtual packed-planar mode Posted on 18-Aug-2022 20:19:59
| | [ #60 ] |
|
|
|
Elite Member |
Joined: 24-Aug-2003 Posts: 4394
From: As-sassin-aaate! As-sassin-aaate! Ooh! We forgot the ammunition! | | |
|
| @kolla
If I did write "it's", then it was almost certainly an autocorrect/autocompletion issue that I failed to notice. Easy to do when typing with two thumbs on a touchscreen. _________________ Doing stupid things for fun... |
|
Status: Offline |
|
|