Amiga.org
Amiga computer related discussion => Amiga Hardware Issues and discussion => Topic started by: Tenacious on December 07, 2007, 03:07:28 AM
-
I've often wondered why none of the remaining hardware makers have come out with a Zorro III memory card that could be populated with modern memory modules. It can't be that difficult.
I would buy two in a second.
Would anyone else buy in?
-
I agree, why?
My guess is too small a market, but a small run of em should sell. 100-200 easy.
I'd pay $100 for a 2GB SD-RAM board. ;)
-
I think Jens Schoenfeld is the guy to make such a board like this.
Why noone think about this before?
-
I've been pondering this since the day I bought my A3000. Back before they put the SIMMs on the CSPPC when DIMMs were the standard!
I know I'll get flamed again for saying it again, but 146M of RAM is just not enough for me! I do use it up all the time. I'd break the bank for 1G. Heck, I'd live with 256M. This "you don't need 128M on an Amiga" thing holds as much water with me as "640k ought to be enough for anyone". Just like, you don't need SCSI, and 3 reboots on every cold start is acceptable.
-
Anyone running OS4 on a A3/4000/4000T could definately benefit from this board.
I have OS4 on 96MB of FAST ram and after booting there is not a lot of ram free. Need some more breathing room...
If that card wasn't too expensive - like 199 CAD ...and let's say capable of putting 4 DDR or SDRAM modules on it I would definately buy it. Even though Z3 ram will be much slower, at least it will be ram. :)
Having said that, OS4 is an absolute amazing upgrade vs. OS 3.9 and I would highly recommend it to anyone who has a PPC (or even doesn't but can get one) that continues to enjoy the use of the classic Amiga. It definately makes the machine feel way more modern, yet keeps the Amiga feeling. Really amazing product.
-
This opens up another can of worms. We need a new PPC card! If anyone is going to make anything this would be it. A PPC G4 card with SD-RAM Slots on it for 512MB to 2GB RAM would ROCK!!!
-
Up to two years ago, I definitely would have bought into that. Memory was the thing I most needed on my A3K. I wanted to find a DKB3128, but those boards are hard to come by!
Now I've moved to Amithlon and memory is not a problem for me anymore. Now the problem is that my PC is too slow to run hardware-banging games and demos properly in E-UAE. But that means that my classic hardware is still useful.
I agree with other posters that the thing the Classic market really needs at the moment is more PPC cards. Those should take modern memory modules.
The Zorro III memory card might still be a good idea. I don't think it should be difficult to make one. How big is the market?
-
The Mediator, Prometheus, and G-Rex can take PCI cards with plenty of FAST graphics memory. Even though there would be no drivers for these "newer" gfx cards, it shouldn't be too difficult to find there memory address and add it to the free memory. I'm surprised no one has done this. I thought about doing it at one time as I have a Mediator but I would have to deal with Elbox and sign their exclusive non-disclosure agreement for their super top secrets. It's much more important to keep the competition at bay in this huge Amiga market than have lots of software for their products after all. Too bad they have this closed minded attitude as they make excellent hardware.
-
I doubt any FAST ram on a mediator would be faster than regular zorro3 ram. The CSPPC reaches the mediator through zorro3 interface AFAIK. as for the BPPC and mediator, I have no idea. GRex is another matter though, it's independent of zorro3 bus.
-
Sure, the Mediator would be limited by Zorro3 speed. The speed of the memory on a gfx card would not be the bottleneck is what I was trying to say. The memory would be faster than almost all existing Zorro memory cards. It would be relatively slow for a PowerPC and maybe even a 68060. Faster CPU local memory would be used first. It would certainly be faster than any kind of virtual memory.
-
The Mediator, Prometheus, and G-Rex can take PCI cards with plenty of FAST graphics memory. Even though there would be no drivers for these "newer" gfx cards, it shouldn't be too difficult to find there memory address and add it to the free memory.
All PCI memory "solutions" you list up here have three common problems:
(1) memory on gfx cards is optimized for writing. It is dead slow on reading, as usually you don't read it in normal operation. Reading on PCI may lead to delayed transactions, with penalty in time and the risk of Buster timeouts (can be seen when accessing PCI-PCI bridges or FlashROM on gfx cards).
(2) no gfx mem access without gfx processor initialization. The gfx mem is only accessible via the GPU, and this beast needs to be setup correctly.
(3) Endianess. Some gfx cards change the byte order of the memory when changing screen modes. In case you switch to a new screen, your memory gets scrambled. Bad idea.
Michael
-
mboehmer_e3b wrote:
(1) memory on gfx cards is optimized for writing. It is dead slow on reading, as usually you don't read it in normal operation. Reading on PCI may lead to delayed transactions, with penalty in time and the risk of Buster timeouts (can be seen when accessing PCI-PCI bridges or FlashROM on gfx cards).
Slow is relative. Zorro3 speeds are slow compared to what a gfx card can move around. The Mediator already uses a part of gfx memory for 10/100 nic cards which is read when it's transferred to the Amiga memory. The 10/100 nic cards aren't blazing speed but faster than Zorro II 10mbit cards in most cases. Do the buster problems occur with a Revision 11 buster?
(2) no gfx mem access without gfx processor initialization. The gfx mem is only accessible via the GPU, and this beast needs to be setup correctly.
Are all gfx cards this way?
(3) Endianess. Some gfx cards change the byte order of the memory when changing screen modes. In case you switch to a new screen, your memory gets scrambled. Bad idea.
I was talking about using a 2nd gfx card just for the memory on it. No screens = no byte order swapping. Some gfx cards probably wouldn't work well as an Amiga "memory card". Choose ones that do. E-bay has tons of used PCI gfx cards for next to nothing.
-
@HammerD
Anyone running OS4 on a A3/4000/4000T could definately benefit from this board.
OS4Classic does not use motherboard memory (except ChipRAM)
and Zorro memory. It would be slow and the new memory
system is no longer designed for multiple blocks of memory
with different speeds.
-
OS/4 wouldn't be able to use it I'd imagine. But, no reason it would not work. I have a DDR2 core if someone has a Zorro III core I could mate it up to.
-
This must be possible on the Amiga because its already been done (http://www.czuba-tech.com/CT60/english/welcome.htm) on other 68K platforms.
-
Slow is relative. Zorro3 speeds are slow compared to what a gfx card can move around.
Relative is not the problem. If your PCI gfx card delays your PCI read just a few times, you are lost as the Buster will timeout the Zorro III cycle, leading to a bus error.
As you want to go to more memory I guess we are talking about modern PCI gfx cards and not the 4MB S3 Virge models.
Most PCI gfx cards nowadays are using PCI-AGP bridges, and so the problem gets worse. Those integrated bridges will add more delays, and the problem of write / read combining.
The bridge will follow certain rules when different read / write operations are pending, and this may lead to bad situations in real life (like: several writes are pending inside the bridge, followed by a read, so all writes will have to be completed before the read is completing -> long delayed read).
Do the buster problems occur with a Revision 11 buster?
The timeout is not a "problem". It's intended to break deadlocks when a Zorro III access fails.
Or do you mean the other Buster "features"? Which one, as there are quite a few of them, and they are available in different tastes in all Buster revisions :-(
Michael
-
The reason there is no ZIII RAM card is because there is very little demand and certainly no investment.
If 200 people paid $200 up front, with a lead time of six months, I'd make one. Hell, any of the Amiga hardware developers would. It aint going to happen though.
Look at the costs: a design would cost about $4k make the first, and about $20-30 each to make there after with an MOQ of 200.
At $200 each, the break even is 60. If you sold all 200, it's $20k profit, not bad, probably worth the risk.
If you sell at $100, break even is 142 units and even if you sell them all you only make a profit of $4k. Hardly even worth the effort? There is also a very high chance of loosing the majority of your investment ($10k) :-(
The fact is that if you need more RAM you get an existing accelerator on the second hand market, Cyberstorm, Apollo etc.
If you are asking "How come there are no new Accelerators with RAM?" now that is a different question.
Someone said "136Mbytes on a classic Amiga is not enough". Really??
I think 99.9% of Zorro III capable Amiga owners would think a RAM board was cool, but when it came to putting their hand in their pockets, not buy one. Probably save up for a CSPPC, or a DENEB ;-)
-
Gee wiz, guys! I wanted a count of people interested in a new memory card for the A3000 and A4000 (a better version of the DKB 3128). This seems so simple to make and if there is enough interest, maybe we can get some movement on this.
I love a lively discussion and all the talk about OS 4.0, new PPC cards, Mediators, and video memory is fascinating. However, I would rather that discussion to move to it's own thread and this thread show a level of interest in a modern version of the 3128.
I probably picked a bad time, after the release of OS 4, to bring this up. Everyone is so excited. As for me, I will probably never buy OS 4 unless it's ported to some in-expensive hardware, like PPC Mac.
Thanks for the interest.
-
Someone said "136Mbytes on a classic Amiga is not enough". Really??
Yes. Really. No kidding. My Amiga multitasks, and so do I. I use screens! I do not run one app at a time on the Workbench screen.
I have a Voodoo 5/5500, and I run 1280x1024x24. At any given time I can have Provector, IBrowse, ArtEffect, and HSPascal running on seperate screens. There are times when I use many more screens than this. I am very productive this way. I run FPC which is a memory hog(& serious fragger). I have AmigaAmp running. I generally have 3 or more browsers open in IBrowse. I am (intermittently) working on a multiplayer game that can easily use 256M of RAM. I have a busy webserver running. I have a shell running on a remote terminal (which I am constantly using).
It's not that I'm too lazy to close windows/screens. I am making full use of them. I have always pushed my Amigas to their limits. This is why I am an Amiga user to this day. Try working a peecee this hard sometime.
Sorry if I sound :madashell: but I've been severely flamed on multiple occasions by people (granted not anyone currently here I think) who think I am wrong in the way I use my Amiga.
-
I dont think you are wrong, I think you are the minority / only one who needs that much RAM.
I am sure there are a few people who will always want more RAM, but IMHO finding a second hand ZIII RAM card would be more practical than trying to get a new RAM card manufactured.
Despite what has been posted about availability, I managed to pick up a DKB 3128 (http://amigahardware.mariomisic.de/cgi-bin/showhardware_en.cgi?HARDID=989)(£85) and a FASTLANE (http://amigahardware.mariomisic.de/cgi-bin/showhardware.cgi?HARDID=1231)(£65) relatively easy.
http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=220155723947
-
Well, isn't the CS-PPC-SCSI a magnitude faster than Z3 (even if you get it work 100% perfectly) ?
Write a pager, get a fast drive (maybe even a solid-state one), and instant access to plenty of "RAM".
Mind you, I still don't see why anybody could use that much on a bottlenecked CPU running at 233MHz or less ...
-
If 200 people paid $200 up front, with a lead time of six months, I'd make one. Hell, any of the Amiga hardware developers would.
Allow me to be the first. I'll buy two. Only 198 to go.
I have a Voodoo 5/5500, and I run 1280x1024x24. At any given time I can have Provector, IBrowse, ArtEffect, and HSPascal running on seperate screens. There are times when I use many more screens than this. I am very productive this way. I run FPC which is a memory hog(& serious fragger). I have AmigaAmp running. I generally have 3 or more browsers open in IBrowse. I am (intermittently) working on a multiplayer game that can easily use 256M of RAM. I have a busy webserver running. I have a shell running on a remote terminal (which I am constantly using).
Me too. Ibrowse alone can eat that much RAM after about 30 minutes of cruising ebay. Hollywood/Designer uses a lot. Pagestream uses a lot, Imagine5.17, SoundFX, CubicIDE, StormC, etc, etc..
-
@ alexh
I've been looking for a DKB 3128 for years in the US. If they are easy for you to get, maybe you could send one my way. Grin.
-
Allow me to be the first. I'll buy two. Only 198 to go.
Put me in for two. 196 to go.
Mind you, I still don't see why anybody could use that much on a bottlenecked CPU running at 233MHz or less ...
What the devil does the CPU speed have to do with RAM usage? Give me half a chance and I'm sure I could find a legitimate application for a 6502 that could use 512M. I may have been the only person to regularly run out of RAM on a C64 too.
-
I have a DKB 3128 in my A4000 with stock 040, as it ran out of memory all the time in iBrowse. Now it's ok, but I would love a new memory card! Heck if they were $100 each i would buy 4!
-
I need 1024 MB of ram in my A1200T.
Everyone needs more ram.
-
Hi, yall
Have a look on Dave Haynie web site ( Google for it )
There is a reference design for a Zorro3 Big ram ..
maybe of some use...
Wol.
-
ChaosLord wrote:
I need 1024 MB of ram in my A1200T.
Shame you dont have ZorroIII then ;-)
-
Tenacious wrote:
I've been looking for a DKB 3128 for years in the US.
You cannot have been looking very hard. I've seen at least one advertised on ebay.com
Set up a range of searches, get ebay to email you when it is advertised and you'll easily find one in under a year.
-
A little late in the reply gamy, but hey, it piqued my interest as memory modules are dirt-cheap these days.
@matthey:
Very interesting thoughts/points!
@mboehmer_e3b:
Interesting counter-points! I realize you're an expert in hardware design, but I have a couple of comments and a couple of questions to bounce off of you, so I can understand better the real issues behind Matth's proposal.
Can you change the Buster time-out values programmatically, or only via hardware hacks?
I'm no PCI/AGP guru, but the conditions you describe for writes being blocked for a pending read, or the opposite, it's clearly a deadlock situation which no bridge or circuitry would allow. It's theoretically possible that no extra read will ever take place, thus those writes are stalled forever. Now, if there's a time-out, after which the writes take place, then everything would continue, but the whole process is delayed, which is what you meant when you said Buster will timeout waiting for that. I guess it could be alleviated by 'dummy reads' or 'dummy writes' interspersed by the software driver.
BTW, isn't it possible to disable delayed transactions? I thought that's what is available for certain PCI configurations which need ISA style timings.
As for read speeds, that has been my experience from the older days programming VGA cards, but I'm not sure if that is still the case for modern cards. For one thing, bit-blitting does reads, so internally the bandwidth for reading IS there, it was the interface bandwidth that was problematic (either smaller bus, like PCI/VLB being 32bit, or ISA being 16bit or 8bit, or smaller and slower bus, all compared to the internal busses which are 256bit or even larger in some cases, and much faster).
As for card setup, you're right, it could be an issue, but my understanding is to get the memory mapped properly you don't need to touch the GPU (and if you do, there's open source code in Linux/BSD for even the most modern cards, since it doesn't touch the propriatery 3D registers), but it's all PCI/AGP register setup. All the necessary code should be in the agpgart.c and related code (PCI, mttr.c, etc).
Lastly for byte-ordering can be an issue too, but as matthey says, this would be a "dumb card", not to be used for display.
@alexh:
I think the only viable way to get hardware to happen is have people pre-pay. Unfortunately that's hard because there aren't many trust-worthy Amiga hardware developers (thinking Ainc, ACT, etc here). I personally want a PPC card or even a CT-60 modified to work on a classic Amiga, and would pre-pay, but I only trust AmigaKit's Matthew for this, because if the project bombs, I feel he'll return me the cash, unlike those others mentioned...
@Kronos:
Interesting alternative. As solid state drives become the norm and get faster, a paged store/virtual memory subsystem could easily be the solution, although limited for IDE based SSDs. But throw in a fast SCSI PPC accelerator (the other can of worms - let's keep it closed for now) and you have a viable memory expansion!
I'm patiently waiting for SSDs to drop in price - I hate HDDs and CDs and their unreliability. We should have had SSDs since the last century... Ugh. I still remember reading articles back in those days of IBMs crystal lattice storage devices. Too bad nothing sellable came out of those projects :-(
-
"Memory is like an orgasm. It's a lot better if you don't have to fake it."
-- Cray Seymore, on virtual memory.
-
Hahahaha, that's an awesome quote, Caius, but then again I believe I've read that Mr. Cray Seymour lived a more lavish lifestyle than many of us Amigans, even if generally secluded, which means he didn't have to fake memory. Don't know what he did in the orgasms department :-p
-
da9000 wrote:
I hate HDDs and CDs and their unreliability.
WTF do you do to your HDD's to make them unreliable? I've never had a failure within the lifetime (5-6 years) of a hard drive. (Touch wood)
If you use modern SMART tools you can predict most drive failures well in advance.
-
alexh wrote:
WTF do you do to your HDD's to make them unreliable? I've never had a failure within the lifetime (5-6 years) of a hard drive. (Touch wood)
If you use modern SMART tools you can predict most drive failures well in advance.
Note I was including CDs, which are obviously much more prone to failure. But since you asked I'll quote a few:
One time I was turning on and off a computer while trying different combinations (jumpers/dips). I was careful to leave it a few seconds off each time. After about 20 times (it's been a while so I don't recall exactly, but it certainly wasn't 100), the hard drive died. How do you like that?
This one was the weirdest: after having a (very well taken care of) Mac Mini running almost always on for a whole year, the brand new internal hard drive, not only did crash and burn, but it made sure it scratched the CRAP out of the metal-coated glass platter, so much so that I could see THROUGH to the other side! Seen that before? Thanks Seagate, you idiots, for while you can detect when the HDD is falling, you can't detect the head smashing onto the platter and stopping the motor at the very least. I've recovered many hard drives (not only with software, but opening them up and doing things to them :-)), but this one was definitely out of my reach!
The sucky thing was that it died 2 days before a *VERY* important meeting, and I had documents needed as evidence on that drive. Luckily I had made a temporary backup a couple days before and even though it was intended to be deleted, it wasn't, which saved my butt. Unfortunately some of the files were missing bytes at the end, I don't know why, so some of them were damaged.
Anyways, I don't abuse my drives (in fact I'm so paranoid I don't even stack them vertically without spacers or about 10cm space in between - keep silicon baggies near them, etc, etc), yet I've had a tremendous amount of failure. But, I'm a power user. I don't keep them in one case, and I've got tons of them all over the place, so perhaps that might explain the high numbers (as opposed to someone having 1 computer, vs 20-30, and never removing that one HDD, and never keeping he machine on constantly, but turning it off every night).
BTW, most of my failed drives have been IDEs. SCSIs seem to hold true to their higher quality.
Speaking of HDDs: I got a lot of 9.5GB 10,000RPM SCSI SCA-U160 drives. These are tested good, hehe. Anyone want to buy any? PM me.
-
da9000 wrote:
One time I was turning on and off a computer while trying different combinations (jumpers/dips). I was careful to leave it a few seconds off each time. After about 20 times (it's been a while so I don't recall exactly, but it certainly wasn't 100), the hard drive died. How do you like that?
20 seems a lot. You were kinda asking for failure. Power On, AutoPark and Spin Up are the 3 most likely points of failure for a drive. Hard drives are less prone to failure if they are hardly ever switched off and that includes spinning down. EXCESSIVE power saving (i.e. Turn off hard drives after: 5 minutes) can kill drives in weeks.
never keeping he machine on constantly, but turning it off every night).
As I said, drives last longer if they never switched off and never power/spin down. It is also interesting to note that higher (but not insanely high) temperatures have negligable effect on lifespan, neither does excessive utilisation.
I should know, I work as an engineer at a company which designs them. But dont just listen to me, listen to the experts:
http://labs.google.com/papers/disk_failures.pdf
"One of our key findings has been the lack of a consistent pattern of higher failure rates for higher temperature drives or for those drives at higher utilization levels."
-
Well granted that start-stop is not the best thing for drives, but come-on! Just 20 times. Anyways...
As for constant usage, I meant not being subjected to any sorts of shock while spinning. I do agree with you that the start-stop is a major source of failure.
As for temperature, it's interesting that you brought it up. I have always believed that it can't be good, and even to the point of failure. There are a few reasons for this, other than simple logic. One major reason is that many times contraction-expansion or other chemical reactions will actually cause crappy solder joints to fail. And while one may say "but it's always on, temperature is constant", that's not true, not only due to load, but ambient temperature (I'm sure you already know this, but not everyone does).
20GB Fujitsu, which I own (excerpt from mhdd.com):
Manufacturer: Fujitsu
MPG3xxxAT/AH drive family
Malfunction signs: Quite unexpectedly for user and user data a drive is no longer identified in PC BIOS.
We should note that this very drive model has broken all records of failures, which happen in most cases after a year of operation, just after completion of the warranty period. The main cause of the malfunction was in the Cirrus Logic CL-SH8671-450E chip. It can hardly be replaced with a working chip because those microcircuits were produced for a special Fujitsu order and the manufacture of that drive family was discontinued long ago. However, there is a method of "revival" and "revitalizing" a malfunctioning chip which allows extending HDD life a little. However, if you ignore drive "hangings" and do not take due steps (at least backup valuable data) the table of S.M.A.R.T. logs in firmware zone will be gradually overfilled and the drive will additionally corrupt its modules in firmware zone, which cannot be restored without specialized software.
One of the versions explaining the cause of problems with those chips is the use of a new polymer compound during production of chip case. The compound decomposes under the influence of increased temperature in humid conditions producing phosphoric acid. But it is just a version; we may never learn whether it is so or not. However, one thing is known for sure: if you unsolder that chip, remove old solder from its pins and contact pads on the board, flush the location for the chip and then solder it back the drive will begin to work properly.
But the more interesting reason for me is that when my Mac Mini HDD crashed and burned, it was during the hottest days of the summer and the office it was in had reached over 100+ F (I forget right now, but I know it was extremely hot), and I have a feeling that it was the temperature that caused the problem.
No matter, what you write is very intriguing and interesting since it's contradictory to logic. I'll read that article you posted and see if they can convince me to remove all fans from my drives :-)
BTW, one distinction that needs to be made with is is "what's hot enough". I personally think any disk over 45-50C is becoming hot.
-
TjLaZer wrote:
This opens up another can of worms. We need a new PPC card! If anyone is going to make anything this would be it. A PPC G4 card with SD-RAM Slots on it for 512MB to 2GB RAM would ROCK!!!
Elbox took care of this one, TJ. Haven't you heard of the Shark?
*snort*
:roflmao:
-
Hahaha. Ainc, Ack, Elbox...
OK, enough. It's really my fault for taking this off topic. So I'll bump it back to topic by asking the hardware gurus here (E3B, alexh, etc - hint hint) to tackle my questions/comments above (before we veered offtopic).