Your support is needed and is appreciated as Amigaworld.net is primarily dependent upon the support of its users.
|
|
|
|
Poster | Thread | Kronos
|  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 19-Mar-2023 10:09:52
| | [ #881 ] |
| |
 |
Elite Member  |
Joined: 8-Mar-2003 Posts: 2713
From: Unknown | | |
|
| @dipsomania
Quote:
dipsomania wrote: maybe 386 PCs + VGA gfx card but at the price of A4000s (the latter was much more elegant than any grey PCs anyway).
|
Funny thing when the A4000 case was just a slightly changed version of the one C= used for their PCs at that time (which didn't really stand out from the grey box masses)._________________ - We don't need good ideas, we haven't run out on bad ones yet - blame Canada |
| Status: Offline |
| | kolla
|  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 19-Mar-2023 19:56:11
| | [ #882 ] |
| |
 |
Elite Member  |
Joined: 20-Aug-2003 Posts: 3352
From: Trondheim, Norway | | |
|
| | Status: Offline |
| | Kronos
|  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 19-Mar-2023 20:24:31
| | [ #883 ] |
| |
 |
Elite Member  |
Joined: 8-Mar-2003 Posts: 2713
From: Unknown | | |
|
| @kolla
The A1000 was an oddball design with it's kbd-garage The A2000 shared it's design with never released C9000 The A3000 was only one that got a proper "design" design 3000T was just butt ugly and for the A4000t they didn't even bother with adapting the case and went the other way round instead. _________________ - We don't need good ideas, we haven't run out on bad ones yet - blame Canada |
| Status: Offline |
| | kolla
|  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 19-Mar-2023 22:08:10
| | [ #884 ] |
| |
 |
Elite Member  |
Joined: 20-Aug-2003 Posts: 3352
From: Trondheim, Norway | | |
|
| @Kronos
Yes, which I figured out just a tad too late, riddened myself with an old AT tower just months before I realized that the A4000T motherboard + various I/O boards I had packed away were in fact AT compatible - I had just presumed all along that they were custom as well. Sigh! :) Last edited by kolla on 19-Mar-2023 at 10:08 PM.
_________________ B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC |
| Status: Offline |
| | bhabbott
|  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 23-Mar-2023 8:16:35
| | [ #885 ] |
| |
 |
Cult Member  |
Joined: 6-Jun-2018 Posts: 509
From: Aotearoa | | |
|
| @kolla
I don't care which came first, the white front with clean lines and silver name badge looks very classy IMO, compared to typical PC clones. This same theme was carried over to the A600 and A1200, so it wasn't just about copying their 'big box' PCs. Kudos to whoever it was in Commodore who thought of it.
|
| Status: Offline |
| | bhabbott
|  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 23-Mar-2023 8:26:54
| | [ #886 ] |
| |
 |
Cult Member  |
Joined: 6-Jun-2018 Posts: 509
From: Aotearoa | | |
|
| @Kronos
Quote:
Kronos wrote: The A1000 was an oddball design with it's kbd-garage The A2000 shared it's design with never released C9000 The A3000 was only one that got a proper "design" design 3000T was just butt ugly and for the A4000t they didn't even bother with adapting the case and went the other way round instead. |
The A1000 was great.
The A3000 was a bad design. Compact, but very heavy. The power supply fan was annoyingly loud. There was no space for a 5.25" drive. The sloping disk eject button was not ergonomic. |
| Status: Offline |
| | Hammer
 |  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 5-Jul-2024 5:15:20
| | [ #887 ] |
| |
 |
Elite Member  |
Joined: 9-Mar-2003 Posts: 6161
From: Australia | | |
|
| @matthey
Quote:
Don't be fooled by the 4 byte/cycle instruction fetch of the 68060 even though it is half of the less efficient Pentium and PPC 603e competition. The key is the decoupled fetch (IFP) and execution (OEP) pipelines with a FIFO between them. The IFP consistently fetches, predecodes instructions and places them in the FIFO when the OEPs are stalled or executing multi-cycle instructions. This is often enough that the FIFO rarely empties with just a 4 byte/cycle fetch and the FIFO often contains enough instructions for multi-issue into the OEPs. The IFP may spend multiple cycles fetching a long instruction but it doesn't matter unless the FIFO is nearly empty. Too many large instructions together can starve the FIFO as can a large instruction after a pipeline refill but these are rare as the average instruction length is less than 3 bytes. An 8 byte/cycle fetch may have increased performance slightly but it is not worthwhile until other improvements are made, especially cache size increases.
|
SysInfo says Hi. AC68080 fixes Motorola's 68060 4-byte fetch per cycle from the instruction L1 cache mistake.
Quote:
The 64 bit data bus increased cost of not just the CPU due to more pins but also memory required for it while providing a small performance benefit.
|
False.
For Quake, Pentium Overdrive 83 Mhz with 16+16KB L1 cache is nearly on par with Pentium 75's 8+8KB L1 cache
https://thandor.net/benchmark/33 Intel Pentium 100 Mhz = 26.70 AMD K5 PR166 at 116 Mhz = 24.40 fps AMD K5 PR133 at 100 Mhz = 22.90 fps Intel Pentium 75 Mhz = 20.00 fps Intel Pentium Overdrive 83 Mhz = 16.20 fps
https://www.youtube.com/watch?v=0_dW-21gdkw Warp1260 with 68060 Rev 6 @ 100Mhz and RTG (without Super Baster bottleneck) has about Pentium 75 results.
Warp1260 has DDR3 memory and 64KB L2 cache.
Software rendering needs high memory bandwidth since the CPU is being used for the software GPU use case.
100 Mhz capable SDR memory modules (e.g. PC100) in 1995, you're in dreamland.
Quote:
A good compromise would have been to allow 32 or 64 bit memory like the PPC 603(e) allowing for cheaper or higher performance memory but it still requires more pins/multiplexing and logic. It's more compelling to use a 64 bit data bus today as 64 bit memory is more common and cheaper. Most high performance CPUs are integrated into a SoC where as much memory bandwidth as possible is more advantageous.
|
64-bit wide bus allows for low effective clocked memory modules instead of the non-existent 100 Mhz SDR. It's EDO DRAM from 1995 to 1997.
_________________ 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 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB |
| Status: Offline |
| | Hammer
 |  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 5-Jul-2024 5:22:59
| | [ #888 ] |
| |
 |
Elite Member  |
Joined: 9-Mar-2003 Posts: 6161
From: Australia | | |
|
| @Kronos
Quote:
Kronos wrote: @kolla
The A1000 was an oddball design with it's kbd-garage The A2000 shared it's design with never released C9000 The A3000 was only one that got a proper "design" design 3000T was just butt ugly and for the A4000t they didn't even bother with adapting the case and went the other way round instead.
|
A3000, Commodore's Jeff Porter forgotten Video Toaster "professional" use case.
During A3000's case design, somebody at Commodore was out of touch i.e. focusing on style over substance.
A3000's front design style is okay.Last edited by Hammer on 05-Jul-2024 at 05:26 AM.
_________________ 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 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB |
| Status: Offline |
| | Hammer
 |  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 5-Jul-2024 5:33:41
| | [ #889 ] |
| |
 |
Elite Member  |
Joined: 9-Mar-2003 Posts: 6161
From: Australia | | |
|
| @agami
Quote:
No it's not. It's about how good or bad the AGA chipset was in 1992/1993. |
Yes, it is.
For technology, the good or bad judgment is relative to the competition e.g. A500 was reasonably good in its time.
Quote:
As I've commented before, you seem to only be able to make such determinations based on comparative assessments.
It would seem you have developed an inferiority complex somewhere along the way, and it actually blocks you form making up your own mind about the value of a thing.
|
This is NOT a personality debate. If you want a flame war, I'll will give you a flame war.
Last edited by Hammer on 05-Jul-2024 at 05:34 AM.
_________________ 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 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB |
| Status: Offline |
| | bhabbott
|  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 7-Jul-2024 2:26:47
| | [ #890 ] |
| |
 |
Cult Member  |
Joined: 6-Jun-2018 Posts: 509
From: Aotearoa | | |
|
| @Hammer
Quote:
Hammer wrote:
A3000, Commodore's Jeff Porter forgotten Video Toaster "professional" use case.
During A3000's case design, somebody at Commodore was out of touch i.e. focusing on style over substance. |
The A3000 was intended to be a 'workstation' like eg. the SGI Indy. None of the engineers liked the A2000's boxy shape, even though it had far more utility than the A1000 (which they preferred). I also bought the A1000 based more on the case style than utility.
The Toaster issue was unfortunate, but again occurred because they were focusing on the A3000 being a compact workstation. They weren't to know that Newtek would try to maximize the space available in the A2000 (A3000 was released in June 1990, Video Toaster in December 1990). The mistake here of course was discontinuing the A2000.
BTW at that time PCs were moving towards 'slimline' cases too. Many of them had barely enough room to fit a single hard drive and perhaps one or two full-length ISA bus cards. This trend reversed in the Pentium era due to the massive motherboards and huge fans required to keep the CPU cool, as well a CD-ROM drive etc. A slimline desktop case didn't have enough room for all that.
Quote:
A3000's front design style is okay. |
Yeah it's OK. Just a pity they had to slope the floppy drive knob. I would have preferred a flat front like the A1000 had.
|
| Status: Offline |
| | Hammer
 |  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 7-Jul-2024 3:20:58
| | [ #891 ] |
| |
 |
Elite Member  |
Joined: 9-Mar-2003 Posts: 6161
From: Australia | | |
|
| @bhabbott
Quote:
The A3000 was intended to be a 'workstation' like eg. the SGI Indy. |
A3000's 25Mhz 68030 doesn't have the compute power of SGI Indy's 100 MHz R4000 CPU launched configuration.
The Video Toaster includes its custom graphics DSP and 24-bit display frame buffer.
https://en.wikipedia.org/wiki/SGI_Indy Quote:
The performance of the 100 MHz R4000 in conjunction with 500 KB of secondary cache, this cache not being provided on the base model, was described as broadly comparable to Intel's 66 MHz Pentium, at least in terms of published benchmark results
|
100 MHz R4000 is about "40 times the performance of a machine with a 68030" https://web.archive.org/web/20040117173253/http://www.futuretech.blinkenlights.nl/pcw9-93indy.html
Quote:
BTW at that time PCs were moving towards 'slimline' cases too. Many of them had barely enough room to fit a single hard drive and perhaps one or two full-length ISA bus cards.
|
Full-length ISA slots are enough for typical SVGA cards e.g. ET4000AX or ET4000W32xx.
https://www.youtube.com/watch?v=X991tn4Ophg ESCOM 386DX-40 PC in a black slim case from 1992 has two 5-inch bay support. At 2:37 in the YouTube video, it's running Doom. This is the machine type that the A1200/020 @ 14Mhz (Q4 1992) and A4000/030 @ 25Mhz (1993) are facing against in Germany's 1992 and 1993 market. This youtube poster has an A500 in display storage.
Amitech's A2200 clone with 68030 @ 40 Mhz should been facing against ESCOM 386DX-40 PC.
Commodore's black case Amiga CDTV desktop computer is style without substance. Many A500 owners purchased fast 386DX-40 PCs in XMas 1992. "More than six months" wasted time on ECS A1000Jr and ECS A600-related mandate led to AGA being late.
After "More than six months", you have a frozen AA3000+ revision 1's AGA chipset being dropped into ECS A3400, hence creating the A4000.
Mehdi Ali fired Bill Sydnes.
Quote:
This trend reversed in the Pentium era due to the massive motherboards and huge fans required to keep the CPU cool, as well a CD-ROM drive etc. A slimline desktop case didn't have enough room for all that.
|
Did you forget laptop PCs? Hint: PCMCIA?
There are P5 Pentium mobile variants starting from Model 4 Stepping 3, 2.285V–2.665V. Model 4 Stepping 3, 2.10V–2.34V. Model 8 Stepping 1, 1.850V–2.150V. Model 8 Stepping 1, 1.665V–1.935V.
https://www.cpu-world.com/CPUs/Pentium/TYPE-Mobile%20Pentium.html
P5 Pentium mobiles are lower than the standard 3.3V Pentium Socket 5.
Intel was extensively separating Pentium's quality for their intended target markets.Last edited by Hammer on 08-Jul-2024 at 02:39 AM. Last edited by Hammer on 07-Jul-2024 at 03:54 AM. Last edited by Hammer on 07-Jul-2024 at 03:43 AM. Last edited by Hammer on 07-Jul-2024 at 03:41 AM. Last edited by Hammer on 07-Jul-2024 at 03:37 AM. Last edited by Hammer on 07-Jul-2024 at 03:34 AM. Last edited by Hammer on 07-Jul-2024 at 03:32 AM. Last edited by Hammer on 07-Jul-2024 at 03:24 AM.
_________________ 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 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB |
| Status: Offline |
| | Hammer
 |  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 7-Jul-2024 4:16:39
| | [ #892 ] |
| |
 |
Elite Member  |
Joined: 9-Mar-2003 Posts: 6161
From: Australia | | |
|
| @kolla
Quote:
With better out-of-the-box CPUs e.g. Cyrix Cx486 DX40.
No out-of-the-box 68040 @ 33 Mhz or 68040 @ 40 Mhz for the A4000 from Commodore.
https://forums.atariage.com/topic/261516-the-amiga-why-did-it-fail-so-hard-in-the-united-states/page/14/#comment-4915711 Quote:
Aha! I have pinpointed exactly when the affordable PC compatible came to Sweden, and in the long run helped kill off the Amiga.
It actually turns out to be Commodore's doing all along. On September 1, 1992 Commodore Sweden decided to more or less halve their PC prices over night:
C386SX-25 (2 MB RAM, 40 MB HDD): from 9370 SEK incl. VAT to 5369 SEK (excl. monitor)
DT386-33C (4 MB RAM, 120 MB HDD): from 19875 SEK to 8119 SEK
DT486SX-25C (4 MB RAM, 120 MB HDD): from 19875 SEK to 8119 SEK
DT486-33C (4 MB RAM, 120 MB HDD): from 23625 SEK to 11995 SEK
Apparently there were already other PC compatibles on the market, with prices far below Commodore's so they decided it was time to cut theirs. Also note that some of those prices in September were slightly lower than the advertised prices in December, three months later.
That lead to several Amiga resellers bringing in Commodore PCs in their selection, because suddenly those were viable to sell. It also lead to some resellers offering other brands of PCs decided to finally begin to advertise in Amiga oriented magazines. Even if they had sold PC's at those prices before, not until Commodore themselves lead the way, they thought it was worthwhile to advertise in the enemy's territory.
The reason I can pinpoint this so precisely is that in the issue for the end of August, there is not a single reseller advertising any PC systems, but in the following issue for the beginning of September, suddenly several of the regular advertisers as well as some new ones had several different PC's to sell.
|
In Sweden, Commodore led the full 32-bit PC steamrolling the Amiga themselves.
Commodore's resistance against CPU accelerated A1200 ("out of the box" 68030 @ 40Mhz) is to protect Commodore's "32-bit" PC business.
Last edited by Hammer on 07-Jul-2024 at 04:24 AM.
_________________ 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 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB |
| Status: Offline |
| | Kronos
|  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 7-Jul-2024 7:07:10
| | [ #893 ] |
| |
 |
Elite Member  |
Joined: 8-Mar-2003 Posts: 2713
From: Unknown | | |
|
| @Hammer
Quote:
"Introduced July 12, 1993; 30 years ago"
Amiga3000
"Release date June 1990"
3 years difference back than was like almost a decade now, so yeah a 030 was still "workstation class" in 1990.
Back then the school I went to had a CAD lab with some 030 based HPs that were still current. AFAIK these were purchased on cooperation with a local business who could use the room in the afternoon._________________ - We don't need good ideas, we haven't run out on bad ones yet - blame Canada |
| Status: Offline |
| | OneTimer1
|  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 7-Jul-2024 21:43:14
| | [ #894 ] |
| |
 |
Super Member  |
Joined: 3-Aug-2015 Posts: 1140
From: Germany | | |
|
| It is true, the Amiga cases where not very inspiring.
A1000 was much to small, you needed an external case if you wanted an HD (like on the A500) A3000 was also seen as to small, when optical drives where needed people bought needed a tower case, I had an external SCSI tower for my A2000.
And the keyboard cases where even worse, in a time when 3,5" HDs where considered cheap, you needed nearly double the price for the internal 2,5" on the A1200.
Well the A600 was something else, normally a computer with SMD should be cheaper, but its extensions where incompatible, the PCMCIA was nearly useless and it was nothing that could get Amiga new customers.
I believe C= should have switched to motherboard instead of full equipped desktop models and AGA just wasn't good enough. I have seen the numbers from Hammer, if I have been a game developer I would have made games first for VGA and than maybe for AGA.
But when games where not made for sprites or not designed around the copper, the reduction in resolution and colors only resulted in bad quality.
|
| Status: Offline |
| | Hammer
 |  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 8-Jul-2024 1:55:53
| | [ #895 ] |
| |
 |
Elite Member  |
Joined: 9-Mar-2003 Posts: 6161
From: Australia | | |
|
| @Kronos
Quote:
Kronos wrote: @Hammer
Quote:
"Introduced July 12, 1993; 30 years ago"
Amiga3000
"Release date June 1990"
3 years difference back than was like almost a decade now, so yeah a 030 was still "workstation class" in 1990.
Back then the school I went to had a CAD lab with some 030 based HPs that were still current. AFAIK these were purchased on cooperation with a local business who could use the room in the afternoon.
|
Reminder, SGI also used 68K CPUs with Weitek Floating Point accelerator boards under IRIS 2000 and 3000 series.
FYI, Weitek co-designed HP's PA-RISC's FPU. https://en.wikipedia.org/wiki/Weitek
In 1988, SGCS superseded the Professional IRIS line with the higher-end Power IRIS line with Symmetric Multi-Processing (SMP) that can scale up to 8 CPU cores. Professional IRIS has MIPS-based R2300 CPUs.
In October 1991, MIPS announced the first commercially available 64-bit microprocessor, the R4000. SGI used the R4000 in its Crimson workstation line.
Microsoft made sure Windows NT 3.x also supports Symmetric Multi-Processing. https://www.youtube.com/watch?v=qIbeV1BYyOo Running Windows NT 3.51 on modern hardware: AMD Ryzen 9 3900X Zen 2 and Intel i3-13100F 13th Gen RaptorLake bare metal.
https://en.wikipedia.org/wiki/SGI_Crimson
SGI's MIPS-based workstation survived up to 2003's Tezro. Around this time, AMD64 K8, IBM PowerPC 970 (PowerMac G5, 64bit PowerPC for desktops), and Itanium IA-64 were gaining attention.
In the 2000s, RISC CPUs such as MIPS and Alpha families were victims of the X86 Ghz race and X86-64's evolution.
68K's evolution pace couldn't keep up with RISC CPUs.
MIPS CPU family was a candidate RISC CPU for the majority of PC clone's https://en.wikipedia.org/wiki/Advanced_Computing_Environment in place of Intel X86 CPU. Intel's Pentium release in 1993, rapid clock speed improvements, and 1995's Pentium Pro (P6) release have reduced MIPS advantage.
Wintel PCs have a "bigger fish to fry" and it's not the Amiga or 68K. My high school's art rooms had A2000s and they were replaced by 256 color Macs. Due to aging ECS, A3000 didn't replace my high school's A2000s.
Last edited by Hammer on 09-Jul-2024 at 05:51 AM. Last edited by Hammer on 08-Jul-2024 at 02:22 AM. Last edited by Hammer on 08-Jul-2024 at 02:19 AM. Last edited by Hammer on 08-Jul-2024 at 02:17 AM. Last edited by Hammer on 08-Jul-2024 at 02:16 AM. Last edited by Hammer on 08-Jul-2024 at 02:03 AM.
_________________ 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 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB |
| Status: Offline |
| | bhabbott
|  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 8-Jul-2024 21:25:36
| | [ #896 ] |
| |
 |
Cult Member  |
Joined: 6-Jun-2018 Posts: 509
From: Aotearoa | | |
|
| @OneTimer1
Quote:
OneTimer1 wrote: It is true, the Amiga cases where not very inspiring.
A1000 was much to small, you needed an external case if you wanted an HD |
Hard drives were very expensive and physically large back then (the original IBM PC 'full height' hard drive took up two 5.25" bays!). A1000 concept drawings show a separate box on top of the main one for expansion. At least one 3rd party expansion for the A1000 did that.
But the Mac and ST used external SCSI/ACSI drives, so this was standard practice at the time and we hardly expected the Amiga to be any different (few preferred a huge ugly box like the PC had!).
The A1000 case looks very nice to me - that's why I bought one. Only problem is the technology of the day used up all the space inside. An updated model with more modern chips and no WCS would have plenty of room to add a hard drive and CD-ROM. How do I know? because I did it - pulled out the A1000 motherboard and replaced it with an A600, plus an IDE hard drive and CD-ROM drive (mounted where the front RAM expansion went).
Quote:
A3000 was also seen as to small, when optical drives where needed people bought needed a tower case. |
The A3000 was designed to be a 'thin' workstation, which typically didn't have drive bays. If you wanted drive bays you bought the A3000T, which was released soon after.
Quote:
And the keyboard cases where even worse, in a time when 3,5" HDs where considered cheap, you needed nearly double the price for the internal 2,5" on the A1200. |
That's a lie. Firstly 2.5" drives were not 'nearly double the price". Secondly, the much lower cost of the in-keyboard design and smaller power supply made up for the higher price of the drive. OEM cost of the A1200 case and PSU:- $11 (~$20 retail), cost of the A4000 case and PSU:- $45 (over $100 retail).
If it was such a bad idea then why did ICD do it 2 years before Commodore, and why did the Atari Falcon and Acorn Archimedes A3000 also use 2.5" drives? Answer:- it wasn't such a bad idea. Amiga fans just got upset because it was different.
BTW I now have so many high capacity 2.5" drives that I don't know what to do with them. Bought several old laptops dirt cheap - with drives ranging from 2-4GB. Now I have to decide if it's worth the hassle to replace my existing Amiga drives.
Quote:
Well the A600 was something else, normally a computer with SMD should be cheaper, but its extensions where incompatible, the PCMCIA was nearly useless and it was nothing that could get Amiga new customers. |
Except it did.
Commodore sold all the A600s they produced in a little over a year. It was designed to replace the A500 in a smaller form factor that put it closer to game consoles without compromising its computing capabilities, and it achieved that. The vast majority of A500 owners never expanded their machine beyond putting in trapdoor RAM, and the A600 was aimed at that market. However if you wanted a hard drive and/or FastRAM the A600 did it cheaper and more ergonomically than the A500.
GVP got upset because their HD-8 and A530 wouldn't plug into it. But few people were buying the A530 and GVP could have produced PCMCIA expansion devices if they wanted to (others did). They just weren't moving the times, expecting no change to the expansion connector and case design after 5 years!
We see a 'parallel' to this in the Sony PlayStation. The original model had parallel bus expansion and serial I/O connectors (as well as RCA AV sockets). The slightly shrunk down PSone removed all that. When Sony released the PS2 they dropped the original PlayStation and introduced the PSone, which initially sold more units than the PS2! Clearly fans were not upset by removal of the parallel and serial ports. Why not? They weren't Amiga fans... Quote:
I believe C= should have switched to motherboard instead of full equipped desktop models and AGA just wasn't good enough. I have seen the numbers from Hammer, if I have been a game developer I would have made games first for VGA and than maybe for AGA. |
Makes sense. PCs had 90% of the market and this would not change, so they were the logical platform to target. Actually with 90% of the market already covered why bother with the Amiga at all? Unless you were an Amiga fan who was happy just reaching that 10%.
The thing is though, no matter what hardware the Amiga had in it this wouldn't change. PCs had been gathering momentum ever since the 1981 when the industry settled on IBM. PCs were the Borg - Everyone else would eventually be assimilated or killed.
I'm not sure what advantage switching to a 'motherboard' (without AGA chipset) would have. The AGA chipset didn't cost much, and without it you would need other stuff to do the floppy drive, serial port, mouse and joystick ports etc. Having the chipset onboard meant that every model had it, providing a base hardware configuration that could be relied on.
Around this time PCs were starting to come out with integrated graphics etc. too. In the late 90's SIS developed motherboards with 'shared memory' graphics built into the motherboard chipset. Other manufacturers followed suit, and today most machines have it (was a life saver for my Dell Dimension 5150 when a mouse piddled on the AGP port!). Quote:
But when games where not made for sprites or not designed around the copper, the reduction in resolution and colors only resulted in bad quality.
|
Blah. Standard VGA was 320x200, a stretched format that looked quite ugly on high resolution monitors (and even worse on LCD). The Amiga could do more than 320x256 in PAL with overscan (compare that to the SNES which was only 256x224). The only games that looked bad were lazy ports from other platforms, eg. PC games that squashed the colors down to 32 or even 16 colors, with a poor algorithm. We saw it the other way too, with Amiga games getting ported to the VGA and only getting 16 or 32 colors. The worst were ST ports, which only had 16 colors out of a 512 color palette. But such is the nature of ports. A machine should not be judged by them (the Amstrad CPC had some terrible Spectrum ports, but also some very good ones).
I don't understand why Amiga fans constantly denigrate the object of their desire. Each platform has its own character, and the Amiga's certainly isn't bad. Nobody runs down other machines like you guys do the Amiga. Is it PC envy, feeling inadequate because your bitmap has fewer colors, familiarity breeding contempt, hurt because Commodore up and died leaving you in the lurch? Perhaps all of the above. One would have thought that after 30 years you would grow up.
|
| Status: Offline |
| | agami
|  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 9-Jul-2024 1:30:23
| | [ #897 ] |
| |
 |
Super Member  |
Joined: 30-Jun-2008 Posts: 1897
From: Melbourne, Australia | | |
|
| @Hammer
Quote:
Hammer wrote:
68K's evolution pace couldn't keep up with RISC CPUs. |
Motorola didn't keep up the pace of evolution of 68k CPUs with that of RISC CPU makers. With the right amount of focus, I'm pretty sure they could've.
Words have meaning.
Last edited by agami on 09-Jul-2024 at 01:30 AM.
_________________ All the way, with 68k |
| Status: Offline |
| | DiscreetFX
 |  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 9-Jul-2024 4:14:02
| | [ #898 ] |
| |
 |
Elite Member  |
Joined: 12-Feb-2003 Posts: 2543
From: Chicago, IL | | |
|
| I got an AGA based A1200 in 1993 and very much enjoyed it. It was a really nice upgrade to my Amiga 500 which I sold to get the A1200. I enjoyed it so much and always regretted selling it so I just picked up an 68060 A1200 this year and am having a blast with it. I’m still on the look out for a nice A600 since I like the small form factor and a Vampire V4 card can upgrade it beyond AGA with SAGA. Maybe I’ll even get a A500 to celebrate my first Amiga purchase that started my IT career. Soon after getting my A500 I got my first job selling and supporting Commodore computers in the Philippines. Last edited by DiscreetFX on 09-Jul-2024 at 04:43 AM. Last edited by DiscreetFX on 09-Jul-2024 at 04:19 AM. Last edited by DiscreetFX on 09-Jul-2024 at 04:15 AM.
_________________ Sent from my Quantum Computer. |
| Status: Offline |
| | Hammer
 |  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 9-Jul-2024 5:56:21
| | [ #899 ] |
| |
 |
Elite Member  |
Joined: 9-Mar-2003 Posts: 6161
From: Australia | | |
|
| @agami
Quote:
agami wrote: @Hammer
Quote:
Hammer wrote:
68K's evolution pace couldn't keep up with RISC CPUs. |
Motorola didn't keep up the pace of evolution of 68k CPUs with that of RISC CPU makers. With the right amount of focus, I'm pretty sure they could've.
Words have meaning.
|
MIPS R-series has stronger MUL IPC, exploiting the weakness on 68020 and 68030.
If you want a strong MUL IPC from Motorola without purchasing braindead 68EC040, purchase DSP56K.
http://archive.computerhistory.org/resources/access/text/2013/04/102723244-05-01-acc.pdf Data Quest 1991's Estimated Long-Range Microprocessor and Peripheral Price Trends—North American Bookings
68040, $730.38 in 1990 and $538.85 in 1991.
80486-25, $790.39 in 1990 and $551.31 in 1991.
R3000-25, $134.19 in 1991.
29000-25, $131.00 in 1991
Motorola follows Intel's price guide despite lacking Intel's 32-bit CPU market dominance.
https://archive.computerhistory.org/resources/access/text/2013/04/102723262-05-01-acc.pdf Data Quest 1992's Estimated Long-Range Microprocessor and Peripheral Price Trends—North American Bookings
For 1992, 68030-25 CQFP, $108.75 68040-25, $418.52 (68040-40 release in May 1992)
68EC030 doesn't promote Unix use cases. 68EC040 doesn't promote Unix use cases. 68EC040 is brain-dead for desktop 68K computers with DMA.
386DX-25 PQFP, $103.00 80486SX-20, $157.75 80486DX-25, $376.75 80486DX-33, $376.75 80486DX-50, $553.25 80486DX2-50, $502.75 All 32-bit Intel X86 CPUs have Unix-capable MMU to run Xenix or IBM's multi-tasking/memory-protected OS/2 2.x.
AM386-40, $102.50, status quo disruptor. Unix capable with MMU.
MIPS R3000-25, $96.31, this is 68040 class, Unix capable with MMU. SPARC-25, $71.00, this is 68040 class, Unix capable with MMU.
1. Unix workstations exited Motorola's 68K.
2. Motorola's 68030=25 was following the Intel's 386DX-25 price guide. AMD's 386-40 dominated this segment.
3. Motorola's 68040 is not price-competitive.
Last edited by Hammer on 09-Jul-2024 at 06:30 AM. Last edited by Hammer on 09-Jul-2024 at 06:28 AM.
_________________ 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 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB |
| Status: Offline |
| | Hammer
 |  |
Re: How good or bad was the AGA chipset in 1992/1993. Posted on 9-Jul-2024 7:00:26
| | [ #900 ] |
| |
 |
Elite Member  |
Joined: 9-Mar-2003 Posts: 6161
From: Australia | | |
|
| @bhabbott
Quote:
Hard drives were very expensive and physically large back then (the original IBM PC 'full height' hard drive took up two 5.25" bays!).
|
Irrelevant for this topic's 1992-1993 time period.
This topic's title is "How good or bad was the AGA chipset in 1992/1993."
5.25-inch HDD examples are Quantum's Bigfoot series.
Quote:
But the Mac and ST used external SCSI/ACSI drives, so this was standard practice at the time and we hardly expected the Amiga to be any different (few preferred a huge ugly box like the PC had!).
|
For 1993, Apple's 68LC040-25 based Quadra 605, LC 475, and Performa 475 have 3.5 inch HDD.
Apple's Quadra 605, LC 475, and Performa 475 with 68LC040 are fcking cheaper than A4000/030 (68030-25)'s $1599 USD!
Quadra 605/LC 475/Performa 475 recycled Macintosh LC III's case size.
https://retroviator.com/apple-macintosh-lc-iii/ Macintosh LC III's 3.5-inch HDD is located at the front.
https://512pixels.net/2013/10/old-mac-quadra-605/ Quadra 605's 3.5-inch HDD is located at the front.
https://computers.popcorn.cx/apple/macintosh/lc475/lc475-02.jpg LC 475's 3.5-inch HDD is located at the front.
https://wisconsincomputerclub.com/index.php?threads/macintosh-performa-475-in-an-lc-case.484/ Performa 475's 3.5-inch HDD is located at the front.
Performa 475, Quadra 605 and LC 475 share the same motherboard, 68LC040-25 CPU, and case!
Quote:
That's a lie. Firstly 2.5" drives were not 'nearly double the price". Secondly, the much lower cost of the in-keyboard design and smaller power supply made up for the higher price of the drive. OEM cost of the A1200 case and PSU:- $11 (~$20 retail), cost of the A4000 case and PSU:- $45 (over $100 retail).
|
2.5-inch HDD has a higher per MB cost when compared to 3.5-inch.
Quote:
If it was such a bad idea then why did ICD do it 2 years before Commodore, and why did the Atari Falcon and Acorn Archimedes A3000 also use 2.5" drives? Answer:- it wasn't such a bad idea. Amiga fans just got upset because it was different. $45 (over $100 retail).
|
For Atari Falcon and Archimedes A3000, you cited products with worse sales records than the Amiga!
From https://www.riscosopen.org/forum/forums/11/topics/14754
1992 Autumn – 240,000 RISC OS 2.x, 3.00 and 3.1 src: http://chrisacorns.computinghistory.org.uk/docs/Acorn/DN/Acorn_DevNL28.pdf
1996 July – 500,000 RISC OS systems shipped src: http://chrisacorns.computinghistory.org.uk/docs/Acorn/ART/ART_DS013_AcornRISCOS.pdf
https://www.stcarchiv.de/stc1993/10/interview-atari-bob-gleadow 13000 to 14000 Falcons were sold by September 1993, with 4164 Falcons in Atari's European warehouse in the Netherlands.
Quote:
Makes sense. PCs had 90% of the market and this would not change, so they were the logical platform to target. Actually with 90% of the market already covered why bother with the Amiga at all? Unless you were an Amiga fan who was happy just reaching that 10%.
|
With the desktop PC, graphics functions are partitioned.
It only takes a killer app or killer OS or killer game to trigger the tsunami component upgrade cycle on the PC.
Quote:
The thing is though, no matter what hardware the Amiga had in it this wouldn't change. PCs had been gathering momentum ever since the 1981 when the industry settled on IBM. PCs were the Borg - Everyone else would eventually be assimilated or killed.
|
It was the 3rd party apps that drove PC standard i.e. high-resolution text-based Lotus 123, WordStar, and WordPerfect were the 1st wave. Commodore's response was a weak #metoo high-resolution text mode in the C12's 1985 release.
IBM's text-based Display Write was less successful compared to text-based WordStar and WordPerfect.
Thanks to IBM MDA, Lotus 123 displaced VisiCalc.
Apple (Steve Jobs) and Microsoft (Bill Gates) are smart enough to realize that to displace the text-based establishment, GUI Excel and GUI Word are needed.
Microsoft recycled the lessons from the Macintosh and ported Mac's GUI version of Excel and Word for Windows 2.x to displace text-based establishment, hence creating a perfect storm for Windows 3.0's tsunami and Windows 3.1's super tsunami.
Commodore reverted back to #metoo mentality after A500's release with time-wasting on monochrome hi-res Denise and 4-color hi-res Denise. With Amiga graphics upgrades, it's "to be or not to be".
Jeff Porter stated his position on post-A500 upgrades i.e. "8-bit planes with 16 million colors" during the debates. Jeff Porter has forgotten chunky pixels.
Without the ECS Denise, the Amiga platform missed the window of opportunity during Microsoft's Windows 2.X establishment.
Unlike the Amiga, Apple gained enough back office and DTP business customers for the Mac platform's install base.
Last edited by Hammer on 09-Jul-2024 at 07:33 AM. Last edited by Hammer on 09-Jul-2024 at 07:28 AM.
_________________ 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 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB |
| Status: Offline |
| |
|
|
|
[ home ][ about us ][ privacy ]
[ forums ][ classifieds ]
[ links ][ news archive ]
[ link to us ][ user account ]
|