Amiga.org
Amiga computer related discussion => Amiga community support ideas => Topic started by: bd1308 on June 05, 2008, 05:16:55 PM
-
I am discussing with a friend of mine about possibly re-producing some of these long-out-of-production PPC/68k accelerator cards.
Several pics were pulled from ads, and the feasibility of re-creating these cards seemed promising..
If you have a spare PPC or 68040/68060 card for a A1200 lying around, please let me know on a pm (no ebay ads). I have cash/paypal/money order waiting.
-
You have to be kidding...
Recreating a PPC card from a photo... first off, it would cost you more than you could ever make back...
These cards were VERY complex beasts, 8 layer boards or something... to try and interface two incompatible CPUs (The 040 and the 603) onto an old 68020 bus, is a horrific task... the firmware development alone would probably cost more than your entire budget.
I honestly cant see what advantage this would have over running UAE on something like a Mini/Pico-ITX board...?!?
-
Photographs are not going to tell you what voodoo magic is going on inside the FPGA, the ROM/flash chip, etc. It won't work without that stuff, even if you manage to reproduce the PCB layout and get the electrical characteristics of that layout working well for timing, signal integrity, etc. And to really recreate the PCB, you'll need to sacrifice at least one board, probably more, to get chip pin pads and internal layer routing scanned. Than you have a PCB, an empty FPGA, and missing parts that are no longer available or sellable in a product due to ROHS.
If you're going to go to that much expense/effort to re-engineer such an old and complicated thing, we're all better off if you instead use that time and money to create a new board, ROHS compliant, leave out the 68K chip, and give us some more recent features, modern clock speed, memory, etc.
-
If you have the skills and equipment, why not try to build a new mother board. With 040/060 DDR memory bus and PCI?
I think that would be easier. And parts may still be available to do so...
If i am not wrong the 603/604 cpus are hard to find. and really hard to find in any quantity for mass production...
-
Is there any real advantage to getting a PPC board in the first place other than being able to install AOS 4? I just use my Amiga for games, demos and music... surely I can do all that just with an 680x0 accelerator and WB 3.1/3.5/3.9.
Going by the ridiculous prices these cards fetch on Ebay, there is definitely a demand for them, but as someone else has pointed out, it would be cost prohibitive to produce them for the dwindling stock of Amiga's out there :(
-
bd1308 wrote:
Several pics were pulled from ads, and the feasibility of re-creating these cards seemed promising..
Well, it *COULD* be feasible if they were single or double layer boards... It gets worse when you have 3, 4 maybe 5 layers.
Maybe you can pull it off if you buy a working example, and grind off each layer.
-
Several pics were pulled from ads, and the feasibility of re-creating these cards seemed promising..
allow me to ....
whahahhahahahahahahahahhhhahahaahahahahahwh
ahahhahahahahahahahahhhhahahaahahahahahwhahahhahahahah
ahahahahhhhahahaahahahahahwhahahhahahahaha
hahahahhhhahahaahahahahahwhahahhahah
ahahahahahahhhhahahaahahahahah
sorry... but...
whahahhahahahahahahahahhhhahahaahahahahah
whahahhahahahahahahahahhhhahahaahaha
hahahwhahahhahahahahahahahahhhhahahaahahahahah
whahahhahahahahahahahahhhhahahaahahahahah
ahum
good luck
Tom UK
-
I just watched Star Wars again and I think I'll have a working lightsaber clone next week!
-
A new 060 Accelerator is possibly viable but PPC? Forget it.
-
Without the CPLD data your plan is damned.
You don't even know what a CPLD is? Twice so.
-
Zac67 wrote:
Without the CPLD data your plan is damned.
You don't even know what a CPLD is? Twice so.
Fortunately, I'm not doing the legwork. I am looking for a 040 card to give to the guy, who will dissect the card and build a new one, along with getting the other chips re-created.
And no...I dont know/care about what CPLD is.
I'll stop here.
-
It's an ambitious project - just mapping the pcb alone will be a lengthy process, and if the CPLDs are protected, then I don't know how you'll manage to recreate the programming of them.
I wish you luck though.
Just remember: there'll always be people who say it can't be done. They may be right, they may be wrong, but that's no reason to give up trying.
Regards
Rich
-
All I need to know is this:
If an 040 accelerator came out for lets say--the A1200 @ $300 per unit, would there be a market for them?
The same for an 060 model?
The rest is possible--the rest will happen provided there is some interest (more than 50 people ) in this. Several people have been contacted regarding getting the chips copied, boards scanned, etc.
I'm sharing details, I just want to know if this could happen if the price was around $300......for an 040 or 060 accelerator.
-
I would buy a new 060 accelerator for any of my Amigas. I'd prefer an 060 accelerator for the 4000/3000 or even 2000 but would pimp out my 1200 with an 060 if offered the choice in the same ballpark as your price.
-
Given the price of second-hand hardware (Ebay), then I believe that the $300 for a new 040 or 060 accelerator would be very competitive, assuming it is stable.
As to numbers... who knows? I think 50 is a given, but more than 100 ...?
Some people would buy it partly to support the effort, partly out of curiosity, whereas others would buy it out of a genuine desire/need for an 040/060 accelerator.
If you do manage to get a working prototype and - eventually - a production run, then A1200 accelerator would definitely be the way to go to get the numbers. Personally, I'd like to see an A4000 version aswell though.
Regards
Rich
-
This sounds very ambitious.
I for one would snap up a new ppc card if it were available, even an 040 or 060 would be nice.
There is so much negativity around this, why can't people be positive? Some encouragement wouldn't be a bad thing.
Good luck, I suspect you will need it though!
-
I'm in talks with several overseas people about this.
As long as I get between 50-60 people who will buy an accelerator **FOR THE A1200**, I can begin talking to my people.
No doubt there **WILL*** be a A4000 accelerator--I just figured we would need to start with something readily available--the A1200.
Stay Tuned...this should be real intresting.
I'm in the market for a **working** 060 accelerator and/or a PPC accelerator (i'm indeed starting with the A1200). pm me.
-
Very unlikely that people would want anything less than a fast 060 board (faster than any available today) that would involve creating or licensing (from Rodolphe of Czuba tech) a 68000 SDRAM controller.
With a 100MHz 060 board you would increase your market by a large amount as owners of current 060 boards would consider upgrading.
Unfortunately you'll never source enough 060 chips to do even one production run though, I've been asking around they just dont exist as NOS. Maybe if you could find a good recycling center who recycled a batch of telecoms equipment or something?
The next thing is that a lot of users are used to having an accelerator with a DMA capable SCSI controller. Unfortunately you'll never get these chips ever again. Your only hope is to make your own DMA capable IDE controller... many months work and increased cost as your CPLD/FPGA has to be much bigger. To leave out a "high speed" disk interface would alienate all potential 060 upgraders and most potential new buyers.
The development of any four layer board and a prototype run will cost in excess of $20k+
You will never make a profit. I very much doubt you will break even.
Have you even considered where you will get the edge connectors from? They havent been made for 10's of years. Tooling costs alone will be huge and the MOQ will run into the 1000's!
-
swoslover wrote:
This sounds very ambitious.
I for one would snap up a new ppc card if it were available, even an 040 or 060 would be nice.
There is so much negativity around this, why can't people be positive? Some encouragement wouldn't be a bad thing.
Good luck, I suspect you will need it though!
there is negativity because many have heard dozens of times people coming up with some stupid idea that has zeeerrrooo chance and tries to drum up enthusiasm for it and then it disapears. hang on a sec ill try...
:-D :-D :-D great this is soooo cool i want like five ill take pics of all my amiga stuff and my dog and you can clone that too!!! :-o :-) when is your release date? can i send some money now? :-o
wait, not done yet :banana: :banana: :banana: there.
good luck though :roll:
sorry that was over the top... there are sites were you can get lots of info on 060 cpu stuff including one for the atari that could presumably be reengineered for the miggy.
over the topp ma haaaahahahaha
here have your experts scan this (http://www.czuba-tech.com/CT60/english/welcome.htm) website for real info. try the tech link towards the bottom there are schematics and stuff there
-
I applaud your initiative but I would recommend that you rather looked into the 68020/30 project over at the EAB website.
68020/30 - RAM - Project (http://eab.abime.net/showthread.php?t=36596)
The reason 68060 accelerators go for 300 on ebay is because they are not taken off an assembly line. Flood the market and you will hardly make 300/board. Unless you plan to sell one every month for years to come.
The only "market" in my view is for the casual gamers who just want to play some games and tinker around a bit. the 32bit extra ram and a faster 020 or 030 is more than enough to do this while keeping it all at a reasonable price level. Perhaps around 100 euro.
-
But the "casual gamer" doesn't need any accelerator!?
Just an A1200 and a RAM board.
-
@bd1308
Let me guess, you want preorders?
Anyway 50+ orders would be enough? That'd be $15000, hardly enough to cover the labour costs for single person for couple of months. You gotta be kidding me...
-
@bd1308
The Natami (http://www.natami.net/knowledge.php?b=0) should be available in a year and your accelerator cards will be obsolete. They are going to try to put a 680x0 series core on the same chip as an enhanced reengineered version of the AGA chipset called SuperAGA. Their 100 MHz 68060-based developer boards should be available within a few months.
You are getting behind already. Maybe you should join up with the Natami team instead. By building the graphics chips into the same die as the main processor they will be able to take shortcuts and integrate in ways never before possible on any Amiga.
-
If its SO impossible and out of reach......how did PHASE5 accomplish this impossible task?
What you mean is it's not financialy viable.
Upgrading a 15 year old computer so much to the point that it still doesn't perform the everyday tasks of the hardware around today isn't financialy viable either but we all still jump at the chance of an 030/040/060/PPC for our ageing hardware.
Its not about wether or not the numbers stack up, It's about the feeling and the passion of the hobby...give the guy a chance to accomplish what we all want ffs.
If it pulled off i would be in the market as much as the rest of you would.
-
You could argue all the casual gamer need is AmigaForever 2008.
and wait, if you need 68k speed why not just go for UAE-JIT instead?
-
It seems like an 040/060 board would be the best, easiest option. probably better to go with a new design however, thus getting modern RAM and IDE. new 68060s can still be had from Freescale in various packages and speeds, or you could sell the boards sans CPU. I've got the 060 already lying around here as a spare, waiting for a home :-)
I don't see why this is so unfeasible, I mean, the FPGA hackers around here can make a whole Amiga chipset on one, why couldn't the firmware be made to do an accelerator for those that want it? I'd love to pimp out my 3000 with an 060, heck, I'd even pay to do it for my 500. 040 accelerators were made back in the day that plugged into a 500, and those ran way hotter than an 060.
Come up with a working prototype and the required convincing Youtube videos, and I'm sure you'd have 300 preorders with a quickness.
Get the working proto, BOM and production schedule set up though, before the video comes out. Then you won't run the risk of burning anyone ala' AntiGravity and others.
FIRST product, THEN announcement ;-) Go ahead, buck the trend in Amiga land!
Good luck, my money's waiting!
-
how did PHASE5 accomplish this impossible task?
By actually having a market of thousands rather than 50. By actually having competent developers and designers.
Oh, and they did go under eventually.
-
Still...... :-)
-
I say, if they feel confident doing it, then go for it and walk the Minmig path - open source everything and let people do the dirt work themselves.
I can always get my hands on a 060 and solder everything up myself. Just have someone make me a blank PCB, like the Minimig I'm working on.
-
There is no magic involved here, all you need is money, money you are prepared to lose as it will be a big gamble.
You might be able to license Natami's 060 circuitry to put on your accelerator card.
Instead of three accelerator cards for different amiga models you could have one card with three connectors to fit 1200/2000/3000/4000 to maximise your potential sales with one design.
Another possibility is a whole new replacement mainboard, leaking batteries and capacitors are accelerating the demise of old boards and a new board could incorporate superAGA graphics compatible with new SVGA monitors.
The continued high demand for amiga's on ebay suggests that the potential market is much larger than the membership here would have you believe.
-
Tron2k2 wrote:
new 68060s can still be had from Freescale in various packages and speeds
That (as I have found out) is not true. Whilst "on the books" they havent been made for many years and they are only selling off back stock... in limited numbers.
-
Another possibility is a whole new replacement mainboard, leaking batteries and capacitors are accelerating the demise of old boards and a new board could incorporate superAGA graphics compatible with new SVGA monitors.
This is actually something I have been expecting somehow, a replacement motherboard that fits the A1200 case - I suspect we all have them lying around? :-)
Earlier we needed tower boxes, but today the "everything in a keyboard" is actually quite OK. A motherboard with SATA, USB, SO-DIMM RAM, minipci, ethernet etc, and a way to hook the old A1200 keyboard (keyrah?) and tada...
The open question is what CPU-solution is the best. PPC ofcourse for MorphOS and OS4, but some of us would actually love a more modern m68k board. I've sometimes pondered on how feasable it would be to create a m68k replacement using a x86-variant and m68k emulation "on board" somehow, more or less hidden for any OS running on top of it ("Amithlon" on FLASH?).
Oh well, one can dream and ponder...
I'm typing this from a "everything in the keyboard" mac btw, an old G4 iBook I got cheap, with torn off display (yeah, really - torn off, by force!) :-)
-
I say forget about 040s 060s. A soft 68k core would be much more easier to develop (there is already one developed by tobias günther, albeit 68000 only), dead cheap compared to real 68ks, possible to interface to SD-DDR ram...
-
I'd have thought the logical thing to do was bring out a cheap 1200/4000 based card for the classic users who want to keep their original cases, something like the Dragon (http://www.elbox.com/news_04_12_17.html) but without the busboard (basically an updated Blizzard 030 which supported more modern memory). This would need to be 100% compatiable with OS3.9 and the older software.
For PPC users, something like the proposed Shark (http://www.buy.elbox.com/cgibin/shop?info=350S1G) but with a modern PCI-E interface. Instead of mating it with a busboard there would be a new motherboard and case (based on PC tech). This board would be for running OS4.0 with 68k emulation for older tiles.
:-P
-
LOL.
This is great stuff, very entertaining.
I'll keep ya posted ;)
-
Hi bd1308,
This is what I would buy. I have an Amiga 2000.
(Note: I'm going to buy a NatAmi60. I would buy this as well, though.)
So, I would want an accelerator of these ideal specifications.
It would be:
1 Gig of DDR2 ram [1]
68060 @100 MHz [2]
Have a 68000 @20 Mhz on there for 100% compatibility [3]
Hard drive controller, if it's there or not, I'm not concerned.
However, if you could make some SCSI 1 to compact flash card readers so I could boot off of a CF from my controller on my Zorro II bus, that would be very handy..... Are those available new anywhere? I've looked on the internet and couldn't find any at all.
What would be cool is if you could have a "development board".
Put on circuitry and 2 empty sockets, one for a second 68000 and a second 68060 of the same speeds as the first 2. (you could probably put the 68000 onto the board, as it isn't too expensive). Also a 1 Gig DDR slot for them. Then, people can try to make AOS SMP capable. Interesting experiment!!!!
I have no idea how much it would cost, but even $600 US is not too much, I think, for taking AOS's prowess out for a test spin. :-D
Now we get crazy, but if it ALSO had a socket for an FPGA, then we could really experiment. People could try things similar to DSP, Altivec etc. designs....
Amiga's potentials are unlimited!
[1] sell board unpopulated
[2] that's the fastest available ever, from what I understand
[3] what's the fastest a 68000 can go? Ever had a 33MHz model?
-
Oh, and, an ethernet and USB 2.0 port would be VERY useful, but I'd still get it if it didn't have one.
Would it be difficult to include a telephone jack and 56K modem on there? "Plug and play" baby!!!!
-
Bandis wrote:
I applaud your initiative but I would recommend that you rather looked into the 68020/30 project over at the EAB website.
68020/30 - RAM - Project (http://eab.abime.net/showthread.php?t=36596)
The reason 68060 accelerators go for 300 on ebay is because they are not taken off an assembly line. Flood the market and you will hardly make 300/board. Unless you plan to sell one every month for years to come.
The only "market" in my view is for the casual gamers who just want to play some games and tinker around a bit. the 32bit extra ram and a faster 020 or 030 is more than enough to do this while keeping it all at a reasonable price level. Perhaps around 100 euro.
I am one of the 3 responsibles for the project.(http://www.amigabr.org/uploads/smil47e876b1cbff5.gif)
For information to everyone:
- It is designed to simpler Amigas (those with 68000 CPU), not to interface with more complex ones (020 and up)!
- The most economical version will cost less than 50 Pounds (main goal), using static RAM (no CPLD to interface memory);
- There is a SDRAM controller ready for PPC CPUs, it uses a simple CPLD for this;
- The initial version is for A600, second for A500/2000/similars;
I don't announce it here because I want to do it only after the first working prototype.
BTW:
[color=ff0000]I don't accept preorders![/color][/b] And no one of my companions will.
(http://www.amigabr.org/uploads/smil47d44be81f1dd.gif)
-
If its SO impossible and out of reach......how did PHASE5 accomplish this impossible task?
Phase 5 created a new engineering feat. This guy wants to reinvent someone else's photograph. Duplicating someone else's design, particularly the FPGA/CPLD magic, is a huge effort. Going by a photograph is laughable. Ha ha.
My main point is not that this is impossible. It's doable, but very difficult. My point is that, considering the time and effort required to actually duplicate the Phase5 PPC boards, that this time and effort would be better spent on giving us something more modern, better, faster, more modern features, etc. than in recreating something now so "old". It's easier to decide how to do something new and do it that way than it is to figure out how someone else did something and then redo it that way. The DCE boards aren't known for their reliability, design us something newer, simpler (no 040/060 chip), more powerful and more reliable instead.
-
Given the high cost of developing such a board, and the need for specific skills, I think it would make more sense to develop a product at first that's specific to the amiga market but which give you a lower cost of entry than an accelerator card. There seems to be a large enough market for the inexpensive gadgets out there [perhaps start with an A1200 scsi controller like the data flyer 1200, or something else that is not easy to come by but increases the flexibility of the system?]. If you started small you could not only offset the need for pre-orders by actually having a cash flow first, but you would also improve your credibility with potential customers and firm up your ability to estimate just how much of a market there really is, before you go and drop 20 grand on development.
-
This is actually something I have been expecting somehow, a replacement motherboard that fits the A1200 case - I suspect we all have them lying around? :-)
Yes, I have a few :roll:
-
Here's an easy project.... And one that will be sellable.
Make a PCI card for the NatAmi60 that can hold 16 Gigs of ram.
A person could put in 1 gig at a time, upto 16.
This is a ram: "hard drive". It would be accessed only through device calls, i.e.
copy dh0:Bladerunner.mpg2 RHD0: clone
run dvplayer rhd0:Bladerunner
When you turn off the computer, it erases everything from RHD0:, because it's only usable when the computer is on.
That way, MAYBE a mere 100 MHz Amiga could play MPEG2 DVDs????
Let's show 'em what we're made of! I'm your customer for that one!
-
Personaly i'd like to see new accelerators have a mini PCI slot.
-
Welcome to this weeks wishful thinking thread.
-
countzero wrote:
A soft 68k core would be dead cheap compared to real 68ks
Yeah right... NOT. An FPGA large enough and fast enough to have a 50MHz 060 would cost $$$
-
If the goal is to run os/4 then the far easier solution is to get os/4 to run on old ppc Macs. Not easy, but far easier than trying to make an eight layer board from a grainiy image for the top...
-
Make a PCI card for the NatAmi60 that can hold 16 Gigs of ram.
Is the PCI address space even that big?
-
No, the Natami is 68060 based and therefore only addresses 2 gigs of addressible fast ram.
That's why Atheist wants it as a Ram disk.
-
A 16gb ram disk would be _sweet_. But as long as we're dreaming, why not make it accept 4gb modules, and take it all the way up to 64gb total? :)
-
SamuraiCrow is correct. AOS would see is as a hard drive in a PCI slot, but it's made of RAM!
This could be usable on ANY computer even Macs and pcs and even AmigaOnes, big market potential. Since you CAN'T turn off the swap file on an XP (it won't work right anymore, it is pure garbage), then people would want to use it as swap space!!!
Since it's really a hard drive controller, you could put 2 compact flash card and 2 SD ram card sockets on there too.
IN FACT, if you make it with 4 compact flash card sockets, and create a RAID circuit, so that every byte read and written is spread across the 4 CF cards (nothing to do with the OS, the OS thinks it's writing to a standard IDE device), then you could get a throughput of 4 cards * 266x transfer * 150,000 bytes (base flash card speed) a second = 159,600,000 Bytes per second!!!!!!!!
159 Megs a second transfer rate, sustained, can hard drives go that fast?
State that 4 CF cards must be on it for the controller to work and they all have to be the same ram capacity.
pkillo wrote:
A 16gb ram disk would be _sweet_. But as long as we're dreaming, why not make it accept 4gb modules, and take it all the way up to 64gb total? :)
Hi pkillo,
I'll tell you why not, because I didn't know they make 4 Gig ram modules. ;-) :-)
I heard that there are 2 Gig ram modules, though. Never seen them listed.
-
The whole idea is that he's dreaming--IE like "i am"
He's dreaming of 4GB modules, like I'm 'dreaming' of new accelerator cards
-
@Atheist
Before posting I would urge you to read up on the 21st Century... despite nearly 10 years, it seems to have totally passed you by...
First off, I never have to swap file switched on on my XP machines (they all have over 1gig ram...).
Secondly... wander over to a local Computer store and check out the latest USB Flash drives... I have two 16gig OCZ Rally2s sitting on my desk right now... they only cost me £35 each...
Finally... a 2Gig DIMM can be bought from any good computer store, I bought a 2Gig Corsair DD2 SODIMM last week for my MacBook Pro for £29 from http://www.dabs.com
-
bloodline wrote:
@Atheist
First off, I never have to swap file switched on on my XP machines (they all have over 1gig ram...).
Hi bloodline,
I'm not sure if I have service pack 1 or not (and I don't have 2), but, when I turn off the swap file, it is IMPOSSIBLE to install EverQuest.... It just would not install and there was no g DAMN explanation given to the USER (that's me) as to WHY it COULD NOT install!!!!!! (It probably would have been IMPOSSIBLE for a MS tech, through e-mail, to figure it out, too!)
bloodline wrote:
Secondly... wander over to a local Computer store and check out the latest USB Flash drives... I have two 16gig OCZ Rally2s sitting on my desk right now... they only cost me £35 each...
Well, if you're putting 4 in, and remember, the highest rated costliest types, it gets rather pricey. I was thinking, you put the most essential things in there, like the OS and EverQuest.
bloodline wrote:
Finally... a 2Gig DIMM can be bought from any good computer store, I bought a 2Gig Corsair DD2 SODIMM last week for my MacBook Pro for £29 from http://www.dabs.com
Isn't it much cheaper to buy 3 1 Gigs?
Now that I think about it, I think I saw it in a computer store once, but thought the clerk made a mistake describing it.
Also, you STILL need a 1 Gig to get to xps limit of 3 Gigs.
Funny that, AOS came out in 1985, with a 2 Gig limit and billy gates, in 2008 has a mere 1 Gig ram space edge over us!!!!!!!!
AND, their OS uses up almost all of the extra 1 gig, so we're STILL ahead!!!!
NatAmi60, 512K of OS, 270.5 Megs free!!!!!!!!!! Woohoooo!! :-D :-D :-D :-D
-
AOS would see is as a hard drive in a PCI slot, but it's made of RAM!
It would not. No drivers.
big market potential.
lol
Since you CAN'T turn off the swap file on an XP (it won't work right anymore, it is pure garbage), then people would want to use it as swap space!!!
That's huge load of nonsense. First of all you can turn the swap off and it works just fine. Second you would obviously put the memory to the mobo directly, rather than going thru dead slow PCI bus.
But hey, tens of GB/s isn't as impressive as 159,600,000 Bytes per second!!!!!!!!
-
Isn't it much cheaper to buy 3 1 Gigs?
You lose dual channel for 1GB of memory then. Bad. 2 x 2GB is much better idea, even if you lose the 1GB due to 32bit OS.
Funny that, AOS came out in 1985, with a 2 Gig limit and billy gates, in 2008 has a mere 1 Gig ram space edge over us!
You're living is some parallel universe. 64-bit vista can handle 8GB (home), 16GB (home premium) and 128+ GB (the rest).
In 32-bit XP, the 3GB limit is actually hardware one. Windows XP itself could easily use more.
-
Piru wrote:
AOS would see is as a hard drive in a PCI slot, but it's made of RAM!
It would not. No drivers.
Hi Piru,
Where? On an XP computer? The PCI CARD would have the logic that separates the data to the separate CF cards, and rebuilds them before sending it back to the computer in a read.
On the NatAmi60, they're making PCI drivers, and to the computer, it would look like a NORMAL IDE controller. The OS does not need to know the data is being written across four CF cards.
Piru wrote:
big market potential.
lol
You're so negative! It's easy to say that when they're NOT available.
EVERYONE would want one!
Piru wrote:
Since you CAN'T turn off the swap file on an XP (it won't work right anymore, it is pure garbage), then people would want to use it as swap space!!!
That's huge load of nonsense. First of all you can turn the swap off and it works just fine.
As described above, MY xp computer DOESN'T work with the swap file off.
Do you have EverQuest? Try what I did.
[/quote]
Piru wrote:
Second you would obviously put the memory to the mobo directly, rather than going thru dead slow PCI bus.
But hey, tens of GB/s isn't as impressive as 159,600,000 Bytes per second!!!!!!!![/quote]
Don't you understand what I wrote? It's a HARD DRIVE made of 4 CF cards..... Not main system ram.
And the NatAmi60, unfortunately won't have more memory slots,.... wish it would. :-((((
-
Piru wrote:
In 32-bit XP, the 3GB limit is actually hardware one. Windows XP itself could easily use more.
Hi Piru,
On the Amiga 1000 in 1985, it was a hardware one too!
-
On the NatAmi60, they're making PCI drivers, and to the computer, it would look like a NORMAL IDE controller.
NORMAL IDE controller doesn't work either. No drivers.
It's a HARD DRIVE made of 4 CF cards..... Not main system ram.
Hard drive keeps its contents over power off. What you're proposing is a not a hard disk, and it has absolutely zero benefit over system main memory.
On the Amiga 1000 in 1985, it was a hardware one too!
But AmigaOS cannot use more than 2GB, even though the hardware would allow it.
-
Atheist wrote:
Don't you understand what I wrote? It's a HARD DRIVE made of 4 CF cards..... Not main system ram.
You can already buy these, Essentially SATA caddies that take CF cards... the performance is less than stella... Just buy an SSD...
Honestly Atheist, why do you bother posting on technology issues, with no technological knowledge?
-
Piru wrote:
It's a HARD DRIVE made of 4 CF cards..... Not main system ram.
Hard drive keeps its contents over power off. What you're proposing is a not a hard disk, and it has absolutely zero benefit over system main memory.
Hi Piru,
Huh? It's using CF, compact flash cards, all the contents would still be there the next day, you can boot up off of it!
Piru wrote:
On the Amiga 1000 in 1985, it was a hardware one too!
But AmigaOS cannot use more than 2GB, even though the hardware would allow it.
I think sane people would ONLY choose an Amiga capped at 2 gigs of ram over xp capped at 3!
-
Atheist wrote:
Isn't it much cheaper to buy 3 1 Gigs?
Also, you STILL need a 1 Gig to get to xps limit of 3 Gigs.
Funny that, AOS came out in 1985, with a 2 Gig limit and billy gates, in 2008 has a mere 1 Gig ram space edge over us!!!!!!!!
AND, their OS uses up almost all of the extra 1 gig, so we're STILL ahead!!!!
No... since my MacBoookPro only has two slots and due to a design limitation in the PreSantaRosa chipset can only address 3Gigs, it makes sense for me to have a 1gig + a 2Gig SODIMMs.
I don't know why you are going on about XP anyway... it's an 8 year old OS... Both Vista (the expensive versions) and MacOSX 10.5.3 are 64bit systems that can address well over 4gig...
-
Huh? It's using CF, compact flash cards, all the contents would still be there the next day, you can boot up off of it!
Ah, I failed to see your change of mind. You first spoke of ram disk that would lose its contents at power off.
I think sane people would ONLY choose an Amiga capped at 2 gigs of ram over xp capped at 3!
If you say so. However, outside of the fanboy fantasy world...
-
then you could get a throughput of 4 cards * 266x transfer * 150,000 bytes (base flash card speed) a second = 159,600,000 Bytes per second!!!!!!!!
159 Megs a second transfer rate, sustained, can hard drives go that fast?
The real bottleneck would be the 14mhz 32 bit path to main memory, which would be 56mb/s at best, the other bottleneck would be your IDE drive, even fastATA can only manage 5.5mb/s.
-
newegg sells 4gb modules. granted, they're expensive, but they'll get cheaper.
the pci bus would of course be a performance limiter, but I expect you would see lower latency on access to large data sets than you would off of anything but a raid array with a fair number of drives and a large cache.
-
HAI GAIS I 1NCE SAW AN AMEGA 40000 BOX ON TV IM PRITTY SHURE I CAN HAVE MY M8 BUILD 1 COZ. NE1 WANT TO BUY ONE WHEN THEY R REDY???//
IT WILL HAVE A BILLION FAST MEGAPLEXER + LOADS OF CHIPRAM FOR TRACKERZZZZZZ AN IT WILL BEAT WINDOZ OR WINBLOWZ AMIRITE?!?!??!!!ONEONE AND BILL GAETS WILL SE IT AND CRY LIEK "BOOK HOO I WISH I HAD AN AMEGA!!11ONE"
:crazy: :roll:
-
This is one of the funniest threads. Who can up with a wilder imagination next?
-
@ B00tDisk
:lol: *dies*
-
Atheist wrote:
159 Megs a second transfer rate, sustained, can hard drives go that fast?
The SAS / SATA II 3G drives I design do 180Mbyte/sec average 270Mbyte/sec peak.
Never buy a PCI SATA card unless you want PCI to be the bottleneck.
-
@Atheist
First of all, PCI RAM disks did exist before Vista came out and, by using them as a swap partion, they did make Windows XP boot at Amiga-like speeds. They used the always-live power rail from the power supply to hold their contents.
Second of all, flash drives wear out faster than hard drives and cost more.
Thirdly, hybrid flash/hard drive combos are available for Vista and allow Vista to almost perform as well as XP or better if you have more RAM than the 32-bit version of XP could address.
These technologies are real and are almost no selling point to the Amiga since they only work on PCs.
-
IT WILL HAVE A BILLION FAST MEGAPLEXER + LOADS OF CHIPRAM FOR TRACKERZZZZZZ
:D
-
Yeah that would be nice...
-
@Aethist
Are you about 8 years old or something ?
-
Tupac is Caput :)
-
BTW= the whole XP memory limit isn't because of the OS itself.
Its because 32-bit systems can only address 4GB of RAM total, including video memory. Has nothing to do with Bill Gates or his antics.
3.75GB RAM+256M video card = 4GB
etc etc.
as laughable as the above comments are, this one really stuck out. Why would MS of all people put some arbitrary limit on memory?
-
Piru wrote:
Since you CAN'T turn off the swap file on an XP (it won't work right anymore, it is pure garbage), then people would want to use it as swap space!!!
EDITS........
That's huge load of nonsense. First of all you can turn the swap off and it works just fine.
EDITS............
Actually Piru, I really hate to agree with anything in the other statements but there are applications that will not function well with Windows XP, if you turn off the swap file. So while yes you can turn it off, Even with 3.5GB of RAM installed it is not normally recommended to permanently turn off the page file.
So while it can be "turned off" it's better to only do that temporarily to work on problems like fragmentation or to run a specific application like a game for better FPS, than to leave it off. Windows normally runs optimally with the page file on. (There are a few exceptions, mainly in high FPS gaming that I've seen.)
There's a good discussion here (http://www.petri.co.il/pagefile_optimization.htm)
and
Microsoft's recommendation here. (http://support.microsoft.com/kb/314482/en-us)
I feel so dirty, I've been told it's never a good idea to disagree with Piru. :lol:
-
bd1308 wrote:
as laughable as the above comments are, this one really stuck out. Why would MS of all people put some arbitrary limit on memory?
:lol:
To quote Bill Gates, "Nobody one will ever need more than 640K of ram." Ever, no how, no way - 640K of ram is just WAY too much ram for the average person. We'll be running 640K of ram for the next 50 years.
Well I might have added the last part. :lol:
Sorry just kidding around because actually your statement is 100%, even back with 640K+384K it was a limit of the 20-bit address of the processor.
"I laid out memory so the bottom 640K was general purpose RAM and the upper 384 I reserved for video and ROM, and things like that. That is why they talk about the 640K limit. It is actually a limit, not of the software, in any way, shape, or form, it is the limit of the microprocessor. That thing generates addresses, 20-bits addresses, that only can address a megabyte of memory. And, therefore, all the applications are tied to that limit. It was ten times what we had before. But to my surprise, we ran out of that address base for applications within—oh five or six years people were complaining."
* Smithsonian Institution interview (1993)
Bill was still surprised that anyone would need more so quickly. Techno-genius that he is. :roll:
-
Nlandas wrote:
To quote Bill Gates, "Nobody one will ever need more than 640K of ram."
You do know that's not true, right?
-
Atheist wrote:
Make a PCI card for the NatAmi60 that can hold 16 Gigs of ram.
http://techreport.com/articles.x/9312/1
Something like that? I am not sure if there was enough of a market to sustain that product in the PC world. Yeah, only 4 GB max, but what can ya do.
http://www.gigabyte.com.tw/Products/Storage/Products_Overview.aspx?ProductID=2179
Oh hey there's a second gen version:
http://www.gigabyte.com.tw/Products/Storage/Products_Overview.aspx?ProductID=2678
-
Reminds me, I was talking to a PC salesman the other day, we had to buy equipment to use software that only runs on 32 bit XP (Zeiss LMS software). And he said you really have to load up the ram, so without thinking I just said would 16GB be enough (that's typically what we order on Mac Pros) and he was flabbergasted, he meant 4 GB, the max for 32 bit XP. To him THAT was a lot of memory....
Anyway, thankfully I don't deal with MS Windows machines more than once a month...
(http://www.lilytherese.com/Snow_Leopard_face_shot_Photo.jpg)
-
"640K ought to be enough for anybody"
Is the correct Bill Gates quote...
-
according to wikiquote:
640K ought to be enough for anybody.
Often attributed to Gates in 1981. Gates considered the IBM PC's 640kB program memory a significant breakthrough over 8-bit systems that were typically limited to 64kB, but he has denied making this remark.[3] Also see the 1989 and 1993 remarks above.
I've said some stupid things and some wrong things, but not that. No one involved in computers would ever say that a certain amount of memory is enough for all time... I keep bumping into that silly quotation attributed to me that says 640K of memory is enough. There's never a citation; the quotation just floats like a rumor, repeated again and again.
-
Thought about making an Amiga fast card from a piece of toast>> this is the max of my technical knowledge!
hmmmm smells great!
(http://C:\Documents and Settings\All Users\Documents\My Pictures\amitoast.gif)
-
Worst thread ever.
--
moto
-
skurk wrote:
Nlandas wrote:
To quote Bill Gates, "Nobody one will ever need more than 640K of ram."
You do know that's not true, right?
Not true according to who? Bill Gates?
-
not true that that is a quote of bill gates, see my post above from wikiquote. he said that he never said that, noone has any reference or citation of him ever saying that. it is probably from someone who doesnt like him or microsoft.
-
KThunder wrote:
according to wikiquote:
640K ought to be enough for anybody.
Often attributed to Gates in 1981. Gates considered the IBM PC's 640kB program memory a significant breakthrough over 8-bit systems that were typically limited to 64kB, but he has denied making this remark.[3] Also see the 1989 and 1993 remarks above.
I've said some stupid things and some wrong things, but not that. No one involved in computers would ever say that a certain amount of memory is enough for all time... I keep bumping into that silly quotation attributed to me that says 640K of memory is enough. There's never a citation; the quotation just floats like a rumor, repeated again and again.
Yeah, I wouldn't agree that I said it either. It was a quote in a magazine interview that I read back in the 80s some time with Bill Gates. He didn't say forever but he did make it seem like a really long time.
He then was later interviewed for the Smithsonian and admitted to being surprised that 640K was only good for a few years. - Wiki isn't always right.
"I laid out memory so the bottom 640K was general purpose RAM and the upper 384 I reserved for video and ROM, and things like that. That is why they talk about the 640K limit. It is actually a limit, not of the software, in any way, shape, or form, it is the limit of the microprocessor. That thing generates addresses, 20-bits addresses, that only can address a megabyte of memory. And, therefore, all the applications are tied to that limit. It was ten times what we had before. But to my surprise, we ran out of that address base for applications within—oh five or six years people were complaining. "
* Smithsonian Institution interview (1993)
Edit:
from the 1989 interview linked to by Dan Oblak:
BILL GATES:
" I have to say in 1981 making those decisions I felt like I was providing enough freedom for ten years, that is the move from 64k to 640k felt like something that would last a great deal of time."
[source] (http://csclub.uwaterloo.ca/media/1989%20Bill%20Gates%20Talk%20on%20Microsoft.html)
-
in 81 he said it should be good for 10 years or so. it wasnt till 6yrs later people really wanted more but not until win95 came out that it was really put to good use.
he was comparing the 640k to the 64k 8bit cpus could use
i think everything about how pcs took off suprised everyone back then and then the price of ram dropped
-
KThunder wrote:
in 81 he said it should be good for 10 years or so. it wasnt till 6yrs later people really wanted more but not until win95 came out that it was really put to good use.
he was comparing the 640k to the 64k 8bit cpus could use
i think everything about how pcs took off suprised everyone back then and then the price of ram dropped
Yes, that's how he more clearly put it 8 years later. I remember reading him quoted in a major magazine. I didn't have the Internet back in 1986 when I read that quote. I also know that I didn't hear it from anyone as a verbal tradition, I read it in print.
I guess they could have mis-quoted him but I doubt it considering he's put in context in several other interviews later that he did feel that 640K was going to be adequate for longer than it was. Anyway, I'll stop here as this is a totally different thread at this pojnt that "New Amiga Accelerators." Appologees.
I'd love to see new Amiga accelerators. I'd like even better a new Amiga Motherboard that ran OS 4 and moved us into the late 90s and hopefully generated enough $$$ to make it possible to continue AmigaOS development. Ah, to dream.
-
Please somebody run by me again why the development (apart from obvious reasons - cash, time, people) of an Intel / AMD based accelerator is conceived as NOT a possibility?
Agreed that OS4 and other recent apps are 68k or PPC based but why not a solution where in the subroutines of a ppc or a 68k chip / processor can be mapped to a more modern instruction set of an Intel or an AMD and then let the processing be done?
-Cynk
-
the amount of glue logic to adapt signals from a x86 to a 68k socket would be very expensive, inefficient, and require a significant amount of engineering.
imagine adapting a core2 duo to a 286 motherboard add emulation in (not a big hit but still) and expect how much in perforamnce.
two words [color=ff0000]Amiga Forever[/color]
emulation gives you all the benefits of the x86 accelerator plus it has a new (no caps leaking) motherboard and practically free piccasso video.
-
bloodline, your ignorance is amazing, and yet you reply to every thread as though you have all the answers. If you spent half as much time learning half of the stuff you think you know, instead of spewing drivel on amiga.org forums, you'd be a rocket scientist by now. You have no qualifications to be making assertions such as the ones you routinely make. Naysaying isn't exactly a difficult task. Someone who knows what they are doing can hire someone to design an 8-layer PCB, so please, think before you allow your drivel to hit the keyboard, with all due respect.
-
@aperez
Which parts exactly were ignorant or wrong? How?
-
@Piru
Yes, a curious first posting.
Plaz
-
I am just having fun reading the angry posts.
bloodline: you sure did pi$$ off that first-time poster though. :)))))))
What is all this nonsense about 2GB/3GB of RAM? We all know Macs are cool. I have a few too. Both Windows XP and Macs are great machines, if you ask me. I am honestly curious how all hard-core Maccers are pushing their Macs to the limit. Do you photoshop/animate/calculate PI to the umpteen zillion decimal places? What kind of software uses 16GB of RAM? I really want to know if you are doing that on a daily basis.
And what is all this trashing of a guy who wants to build the PPC board? All you have to do is not respond if you don't have one (PPC board) to sell. We all know that PPC boards have the worst bang for the buck associated with them. This is why they sucked in the marketplace: too expensive for what they offered.
Remember that this is a forum for enthusiasts. If you are feeling unenthusiastic about Amiga (or anything else), log out for a day. Wait until enthusiasm creeps in again.
Bokay?
-
codenetfx wrote:
What is all this nonsense about 2GB/3GB of RAM? We all know Macs are cool. I have a few too. Both Windows XP and Macs are great machines, if you ask me. I am honestly curious how all hard-core Maccers are pushing their Macs to the limit. Do you photoshop/animate/calculate PI to the umpteen zillion decimal places? What kind of software uses 16GB of RAM? I really want to know if you are doing that on a daily basis.
Photo and video editing will chew through ram like it's going out of fashion, I suspect handling any large dataset whilst requiring more or less realtime input would be the same, so I would add to that music. Compiling large programs also tends to like RAM a fair bit as well.
The point in the thread discussing 16Gb was someone wanting to use ram dimms as a replacement for a hard drive. It was later pointed out that there are better alternatives to this out there already - hybrid drives or even solid state disks. Hell, even USB flash drives.
codenetfx wrote:
And what is all this trashing of a guy who wants to build the PPC board?
I'm guessing that you didn't read the bit where the initial poster stated that he would be doing it not from a tech drawing, or a schematic they'd aquired or made themselves, but from a photograph.
codenetfx wrote:
Remember that this is a forum for enthusiasts. If you are feeling unenthusiastic about Amiga (or anything else), log out for a day. Wait until enthusiasm creeps in again.
Yes, lets let people come on here, with ludicrous plans about this that or the other, and not take the time to explain to them why it won't work, and suggest better alternatives (both of which has been done in this thread if you'd bothered to read it).
-ninjar edits :lol:
-
i dont know where this thread is going,but some people are capable of amazing things.
-
@aperez
Can you elaborate that? Personally, I fail to see how Bloodlines comments have been irrelevant or ignorant, and frankly I don't consider myself a total moron in this context :-)
-
Sure, I'd be happy to. Let me start by just saying that while, for many, computing may be simply a hobby; Retrocomputing certainly is for me, and I never owned an Amiga when they were new. For me it's not about reminiscing and longing for the good/bad old days, and wishing for an Amiga revolution, which is highly unlikely, at best, to ever occur.
I suspect strongly that my perspective on the matter of the "future" of the Amiga/Amiga-workalikes is perhaps quite divergent from the views of a lot of old-timers. Computing is my profession. As a systems administrator, I'm currently responsible for administering over 150 physical machines of varying age, from bleeding edge (8-way quad-core AMD Opteron machines with 64GB of RAM, and Sun UltraSparc T2-based machines). I'm right in the middle of building out a new facility in a data center that is serious top-shelf stuff. The website I administer the machines for receives millions of hits per week. There are a *lot* of moving parts, figuratively speaking.
Why explain the above? I believe it lays the framework for a very different perspective on computing, both past and present, that's all. Secondly, I also feel it does provide my opinions with a certain amount of reputableness. If I come across as elitist or simply out to toot my own horn, I offer my sincerest apologies. That's not the intention. Yes, I can sometimes be arrogant. Yes, I call it like I see it, sometimes even when it's not genuinely necessary. Fundamentally, I'm a very reasonable person, and I'm sure I could have an entirely reasonable and civil conversation with anyone here on the forums.
Anyways, bloodlines' comments are often factually false. If you really need me to, I will go off and cite specific examples, but some of them are so far off the mark that they should be self-evident to anyone with some semblance of clue. I'd urge you to simply re-read some of his comments (both in this thread and others) and come to your own conclusions. My opinion is in no way canon :)
-
aperez wrote:
Anyways, bloodlines' comments are often factually false.
If you can refute what I say, with concrete examples then go ahead. I will stand by my comments until proven otherwise.
If you really need me to, I will go off and cite specific examples, but some of them are so far off the mark that they should be self-evident to anyone with some semblance of clue.
If you are not prepared to back up your libellous comments with evidence no matter how stupid it may seem to you, then don't bother saying anything at all!
If you are privy to information I am not, then please share... that the point of conversations.
I'd urge you to simply re-read some of his comments (both in this thread and others) and come to your own conclusions. My opinion is in no way canon :)
With only 2 posts to your name, I suspect you are probably just someone I pissed off who is now using a separate account to try and get back at me. Well done aren't you clever.
-
@aparez
So you're a system admin. That alone doesn't qualify you for much, however (and dropping all those hw specs... oh please).
I also want specific examples, thank you.
-
codenetfx wrote:
I am just having fun reading the angry posts.
bloodline: you sure did pi$$ off that first-time poster though. :)))))))
An artist is nothing without an audience... ;-)
-
Piru wrote:
@aparez
So you're a system admin. That alone doesn't qualify you for much, however (and dropping all those hw specs... oh please).
I also want specific examples, thank you.
@Piru
if he work's for a big company he may have others to help him.
@bloodline
you posted a comment in one of my photo album i don't think it's funny but you must think a little bit more before posting :-D
-
@aperez
System administrator. No offence aperez, but in this context it's like saying "Hey! I used MS Paint once". It's not applicable when it comes to hardware design fundamentals imo.
-
@aperez
If you really need me to, I will go off and cite specific examples,
Yes, you'll need to. You are among many other IT, programming and hardware professionals here. I could completely bore you and every one else on the list with an increadibly long resume of the hundreds of systems and apps I manage and have managed in the past 24 years. So could many others here. As a common courtesy, we don't. We try to stick to specific topics and play nice wether we disagree or not. When we discuss pros and cons, yes we want to see the data that goes with it.
Plaz
-
delshay wrote:
@bloodline
you posted a comment in one of my photo album i don't think it's funny but you must think a little bit more before posting :-D
Hahah yeah :-)
-
I hardly find the comparison fitting, however IMHO any genuinely talented systems administrator has a genuinely rich knowledge of underlying hardware design (CPU architecture, memory subsystem limitations, etc) and in my case, I work with engineering staff on a daily basis to aid them in addressing these sorts of issues.
@Bloodline: You have my sincerest apologies. I believe I misattributed several posts from another user who shall remain nameless as your own, much to my dismay. Additionally, in my late-night reading of several different threads, I believe I mistook some comments of yours from other threads and unwittingly confused them as being replies to this thread. I re-read the entire thread, and besides your first reply, I don't really see anything in this particular thread that is at all objectionable.
-
aperez wrote:
I hardly find the comparison fitting, however IMHO any genuinely talented systems administrator has a genuinely rich knowledge of underlying hardware design (CPU architecture, memory subsystem limitations, etc) and in my case, I work with engineering staff on a daily basis to aid them in addressing these sorts of issues.
Dude. These are your first posts in this forum. Think again. People here are generally not retards.
You could either gain respect by stating your case using hard facts *or* give people the impression that you're an ignorant dick by using the "I'm right because I'm a systems administrator with rich knowledge"-approach.
(and frankly, "systems administrator with a genuinely rich knowledge" != "hardware engineer").
-
"any genuinely talented systems administrator"
Met loads of those......
"has a genuinely rich knowledge"
Yep and met one or two of those
with "underlying hardware design (CPU architecture, memory subsystem limitations"
Hardly any of those....!
IMHO....it takes a ikle bit more than being Sys Adm to understand hardware to that level.
Sys Adm gives you good overall but not the fine detail to understand at the level we are talking here.
I've been a Sys Adm for 9 years, a developer for 10 years before that and I began as an electronics tech in the 6502/z80 era....the stuff they do today is way beyond my patience and understanding.
I certainly can hold my own at the x86 architecture level but I would never claim I could component level design/build or repair one.
To much for my brain to hold onto without leaking. :)
-
@shoggoth: "(and frankly, "systems administrator with a genuinely rich knowledge" != "hardware engineer")."
I never claimed it did...
-
I notice you've not offered a single shred of evidence to your assertion of Bloodlines alledged innacuracies.
You're repeated "I'm a systems administrator" cuts about as much mustard here as it would if I were to start spouting off about building linux based pbx's for {bleep}s and giggles (which I did, a couple of times, which was fun).
IE, none. It's not relevant to the conversation.
So. Your money where your mouth is, show us where bloodline has messed up.