Amiga.org
Amiga computer related discussion => Amiga Hardware Issues and discussion => Topic started by: mingle on January 24, 2011, 11:27:13 AM
-
Hmm...
With these new accelerators appearing, I wonder how feasible it would be to make one with built-in USB?
A nice '030 card with 4 USB 2.0 ports (via simple header-pins) would be lovely... :-)
Any takers?
Mike.
-
Ofc it's feasible. The question is WHY would someone do it :)
Since there is a production of "ACA1230" and a stock of already tested and working "Subway" adapters I don't see why someone would invest into something like that.
-
I'm all for this. I love my Subway, but its very limited in speed off the clockport with its theoretical max of 12mbit.
Seeing as Hyperion are releasing Amiga OS 4.1 classic next month, I've mentioned a few time over on Amibay that I would go one or THREE steps further (If I was say, A-EON) and make a full blown PPC Accelerator (with that nice processor in the X1000 for example running at 1.6GHz) with no 68K processor (make a shit hot 68K emulation that runs on the PPC) and instead onboard Radeon mobile chipset of some sort, Laptop DDR memory slot for 1 or 2GB (2 slots if they'll fit) and for fits and giggles add a USB header (or two) and a feature connector to add such things as standard laptop wifi modules internally. Whop it all on an A1200 accelerator and sell for around 4 to 500 Euro. They'll sell more of these than the X1000 thats for sure.
If the release of os 4.1 classic has seriously good compatibility with the AGA chipset and older software, then all that would be needed is software emulation for the 68K (which on a 1.6Ghz processor may well give better speeds than a full blown 060 at 50MHz) and we'd be looking at a seriously awesome classic amiga able to use all the new os 4 software and all the old 3.x software in one. No more need for dual partitions for os 3.x and os 4.x.
-
A single 030 expansion with a proper USB2.0 implementation would be far better than having to by a slow subway card.
I know quite a few people who'd buy one (or more) in an instant.
I guess we can only dream! :-)
Mike.
-
@Phipscube
And if I had a dozen hering, I could make a bowl of fish-soup ....
a) there no way to get all that stuff functional on a PCB sized for the A1200-trapdoor slot
b) there is no way you can keep it cool enough or supply enough power to it without serious modifications to the A1200
c) since you've replaced allmost everything why bother with having an A1200-mobo dangling on it's side ?
d) there might be some level of AGA-compability in OS4.x, but it's surely not enough to run those games and apps that make still having classic worthwhile.
e) given it's small size and special need to connect to the A1200 this board would create greater costs both in development and production as the X1000
d) that price would decrease the market to a 2 figure number (at best)
f) such a small production-run will blow pricing completly out of the water
Just a reminder, there is a reason why the new ACA-board are 030-only and that price vs. marketsize. Making it an 060-board would have increased pricing so much that it would be out of the impulse-buy range and as such it wouldn't be cost-effective.
-
@Phipscube
And if I had a dozen hering, I could make a bowl of fish-soup ....
a) there no way to get all that stuff functional on a PCB sized for the A1200-trapdoor slot
b) there is no way you can keep it cool enough or supply enough power to it without serious modifications to the A1200
c) since you've replaced allmost everything why bother with having an A1200-mobo dangling on it's side ?
d) there might be some level of AGA-compability in OS4.x, but it's surely not enough to run those games and apps that make still having classic worthwhile.
e) given it's small size and special need to connect to the A1200 this board would create greater costs both in development and production as the X1000
d) that price would decrease the market to a 2 figure number (at best)
f) such a small production-run will blow pricing completly out of the water
Just a reminder, there is a reason why the new ACA-board are 030-only and that price vs. marketsize. Making it an 060-board would have increased pricing so much that it would be out of the impulse-buy range and as such it wouldn't be cost-effective.
a)The ACA cards are tiny, there's plenty of room for other stuff on them.
b) USB2 does not draw much power, if it did, every phone and it's dad would not have it.
c) why bother upgrading these fiddly old computers at all?
Also, there is spare blanking slot on the back of the A1200 where the USB ports could go.
d-f) Not a OS4 guy so I can't comment.
f) All current accelerators are fairly small production runs. USB 2 chips cost peanuts.
The ACA cards are so well priced, i'm sure many people would pay a premium for extra functionality, as Amigans already pay stupid money for a lot of their kit.
-
Sorry for being off topic, but one of the posts here made me think of it.
Does os4.x for classic even support the custom chipset in a way that its possible to use custom chipset based software ? I know AGA is "supported", but in what manner ? Im under the impression its only usable as a basic gfx display (ie. akin to a weak gfx card)
-
@kephren
I was not replying about the orginal proposal (USB-chip on simple 030-card) but to that everything-AND-the-kitchensink crapshot.
-
My mistake Kronos. yeah, that post wants a lot of stuff in the trapdoor slot.
@Phipscube:
With the total OS4 market being small, the classic OS4 market would be smaller still, amd the A1200 market even smaller (i'm guessing most OS4 classic users will be using bigbox & RTG?)
doubtfull this would be cost effective.
Now a cost reduced wedge X1000 a couple of years down the line, that i'd buy into ;) (still like the wedge home computers, call me old fashioned!).
-
Ah but who hasn't towered their 1200s? There's plenty of room in most 1200 towers...
-
@persia
The right question isn't about who hasn't... but who has :) (j/k)
Seriously though... Desktop case is teh sexy shiz
-
@Phipscube
And if I had a dozen hering, I could make a bowl of fish-soup ....
a) there no way to get all that stuff functional on a PCB sized for the A1200-trapdoor slot
b) there is no way you can keep it cool enough or supply enough power to it without serious modifications to the A1200
c) since you've replaced allmost everything why bother with having an A1200-mobo dangling on it's side ?
d) there might be some level of AGA-compability in OS4.x, but it's surely not enough to run those games and apps that make still having classic worthwhile.
e) given it's small size and special need to connect to the A1200 this board would create greater costs both in development and production as the X1000
d) that price would decrease the market to a 2 figure number (at best)
f) such a small production-run will blow pricing completly out of the water
Just a reminder, there is a reason why the new ACA-board are 030-only and that price vs. marketsize. Making it an 060-board would have increased pricing so much that it would be out of the impulse-buy range and as such it wouldn't be cost-effective.
a) It could be possible without the 68K in place and a header for Wifi
b) Would run MUCH cooler than the BPPC and BVPPC solution we have now and would need less power.
c) To allow integrated backwards compatibility for OS 4.1 and make it seamless without the need of E-UAE running (badly) in a window at the speed of a 020. And to a point where one could eliminate the need for OS 3.9.
d) You have a point here, it would need developing further to integrate AGA.
e) I don't believe that. How? they would in effect be using parts they already have, or have access to. The board layout would cost some dolla, but do you really think it would be MORE than X1000 mainboard?
d?) I think this would sell more than the X1000, but thats just my opinion. so more 3 figure. it could be a more simple PCC only accelerator for even cheaper. It was only an idea.
f) The current PPC cards (if you can find one) cost more than what I mentioned, and it would effectively be the "cheapest" OS 4 solution on the market.
I understand the ACA market and why Jens made them, but the target is also of pure classic users that in most cases simply want to use WHDLoad. This is a proposal for a card that could allow Hyperion to sell more than a small handful of OS 4.1 Classic CD's as the market is now and allow clasic users the option if they wanted it, becasue the current PPC accelerators are too slow and run too hot.
If you can be arsed, look at my A1200D project (link in my sig) And you'll see how much fun can be had fitting crazy stuff in a standard Wedge A1200. I'm maybe a minority of guys that are willing to go nuts with the idea of a high specced classic Amiga hardware.
@kephren
I was not replying about the orginal proposal (USB-chip on simple 030-card) but to that everything-AND-the-kitchensink crapshot.
Errm.. no need to be offensive! :furious:
My mistake Kronos. yeah, that post wants a lot of stuff in the trapdoor slot.
@Phipscube:
With the total OS4 market being small, the classic OS4 market would be smaller still, amd the A1200 market even smaller (i'm guessing most OS4 classic users will be using bigbox & RTG?)
doubtfull this would be cost effective.
Now a cost reduced wedge X1000 a couple of years down the line, that i'd buy into ;) (still like the wedge home computers, call me old fashioned!).
I would imagine IF this ever did happen, there would be an A1200/A4000 variant of each. And I do still like the wedge machines... and trying to pimp them up :D Just for kicks really. And I think a more realistic option is for A-Eon to make a wedge X500 shall we say further down the line..... but seieng as OS 4.1 classic is coming out (which to be honest surpriesd me) then why not release an uber cool PPC accelerator to compliment it?
Anyway, sorry for slanting this thread off the OP comments, it was just a simple airing of an idea.
I'll answer the question as I SHOULD have :)
YES I think an ACA 1230 with a USB option is a good idea.
-
@persia
The right question isn't about who hasn't... but who has :) (j/k)
Seriously though... Desktop case is teh sexy shiz
I agree ;)
-
I have no interest in OS4.x for classic unless I can use software reliant on custom chipset. If I want an amiga type system where I have to emulate to use my old favorites I'll go AROS, which emulates 68k/custom chipset software an order of magniturde times faster, and cheaper. I got my classic amiga to use its software base. FOr system friendly stuff I have amithlon, which runs 68k software much, much faster and again for cheaper.
To each his own, but I really cant see a new ppc based accelerator with everything + the kitchen sink for classics selling enough to warrant being developed.
-
b) Would run MUCH cooler than the BPPC and BVPPC solution we have now and would need less power.
PWRficient PA6T produces more heat than the 603 CPUs used in BPPC. You also need to consider the support chipset power requirements.
e) I don't believe that. How? they would in effect be using parts they already have, or have access to.
Parts are only part of the problem. There are much larger problems with trying to convert a full blown mobo to an accelerator. Also, the company producing the card has no amiga experience at all, which would make it much harder and more expensive.
d?) I think this would sell more than the X1000, but thats just my opinion. so more 3 figure. it could be a more simple PCC only accelerator for even cheaper. It was only an idea.
As far as I can tell BPPC already was a borderline at being profitable at all. These days the potential market is only a fraction of the old one. No way would such an expensive and complex board be profitable now.
f) The current PPC cards (if you can find one) cost more than what I mentioned, and it would effectively be the "cheapest" OS 4 solution on the market.
I very much doubt your price quotes.
-
a) look at the space required by the CPU alone on the X1000
b) you may want to back up that one
c) it's still pointless for most NG-uses and won't be compatible enough for most classic-uses
d) see above
e) there is no way one could just shrink the current X1000 layout to that size, not without replacing most of the parts with varieties used in laptops and going seriously multilayered on the board (which adds even more design problems). So yes it would be a seperate design and production is gonna be costly
f) we would atleast be talking about inflation corrected prices for the orginal PPC-cards back in 1997 .... read >3000Euro IF the boards would be produced in a similar amount. But given that the market shrank by atleast 90% in the past 13 years we could easily be talking about 10000 or more.
And yes a crapshot remains a crapshot and deserves to be called a crapshot.
-
I very much doubt your price quotes.
Yeah allmost missed that one were he wants to tell us that a miniturized version of the 1500Euro Xena-board could be sold for 500 Euro or less.....
-
No offense meant but in classic amiga-land what does USB2 get you?
Most amiga apps clock in at a few megs, at most. Moving those off of or to a memory stick at USB1.1 speeds (even through the clock port) isn't a big deal. A stock Amiga really can't do H.264 video, or MP4, or AVI, so you're not moving multi-hundred megabyte or gigabyte files around. CD burning or burning DVD images needs to be done at lower speed on m68k Amigas anyway, so you don't need it for external optical drives. You certainly can't type or move the mouse fast enough to oversaturate the 1.1 USB bus! :) I don't think classic Amiga systems can boot off of USB; even if they can there's no swapfile or any such thing being used so it's not like the transfer speed of 1.1 will affect you there.
Would it "be nice"? Of course it would! About the only real "need" would be if you wanted to do USB ethernet.
Now, on that x1000 thing, assuming it ever comes out? For it to even be considered a starter USB2 has to be standard. With an eye to USB3.
-
Ofc it's feasible. The question is WHY would someone do it :)
Why? Because it would be easy to throw in, not as easy as the IDE port on the X-Surf, but similar idea. With a full speed USB 2.0 port built in, you could boot from a small CF, then run your hard disk, DVDROM, even network, off USB. Speaking of which, how are you going to interface your CDROM when using the new accels coming out?
-
Now, on that x1000 thing, assuming it ever comes out? For it to even be considered a starter USB2 has to be standard. With an eye to USB3.
I don't care much about usb3 even on PC. esata is a lot better when it comes to external hdds.
-
Rather build in ethernet than USB.
-
Rather build in ethernet than USB.
Why, when a USB ethernet adapter is so cheap?
-
No offense meant but in classic amiga-land what does USB2 get you?
Most amiga apps clock in at a few megs, at most.
Would it "be nice"? Of course it would! About the only real "need" would be if you wanted to do USB ethernet.
USB cup warmer and christmas tree? ;)
Seriously though, I would just like to see ANY extras on future accelerators (USB1 would do me), we know the ACA's are tiny, there's space there for extra abilities, rather than squeezing umpteen different expansion boards at extra cost into various Amiga orifices.
As for USB2, maybe your right. I still render large anims and bulk copy MODS -but I have the PCMCIA for this, and it's fast enough.
-
No offence meant but in classic Amiga-land what does USB2 get you?
Nothing. You can do everything that USB can do with alternatives. Most alternatives actually work better because the way that USB is shoe-horned into the OS.
-
I don't care much about usb3 even on PC. esata is a lot better when it comes to external hdds.
Nah, eSATA is too slow for RAID and cannot provide power so you need yet another power supply or a USB vampire cable. Had the eSATA connector had power things may have been different.
-
What about the SCSI connector on the Apollo 1240 boards? Could somebody use those signal lines to make a USB daughterboard rather than a SCSI one, and just use that??
Would be a typical Amiga-style solution.
Look at the clock port ffs.
-
PWRficient PA6T produces more heat than the 603 CPUs used in BPPC. You also need to consider the support chipset power requirements.
Parts are only part of the problem. There are much larger problems with trying to convert a full blown mobo to an accelerator. Also, the company producing the card has no amiga experience at all, which would make it much harder and more expensive.
As far as I can tell BPPC already was a borderline at being profitable at all. These days the potential market is only a fraction of the old one. No way would such an expensive and complex board be profitable now.
I very much doubt your price quotes.
Thanks for providing a more mature response Piru.
a) look at the space required by the CPU alone on the X1000
b) you may want to back up that one
c) it's still pointless for most NG-uses and won't be compatible enough for most classic-uses
d) see above
e) there is no way one could just shrink the current X1000 layout to that size, not without replacing most of the parts with varieties used in laptops and going seriously multilayered on the board (which adds even more design problems). So yes it would be a seperate design and production is gonna be costly
f) we would atleast be talking about inflation corrected prices for the orginal PPC-cards back in 1997 .... read >3000Euro IF the boards would be produced in a similar amount. But given that the market shrank by atleast 90% in the past 13 years we could easily be talking about 10000 or more.
And yes a crapshot remains a crapshot and deserves to be called a crapshot.
a) (http://www.a-eon.com/pics/PA6T.png)
Its not big is it? True, I was off on the fact that it runs "MUCH Cooler" what I should have said is Relative to speed. The 603e runs at approx 6W @ 300MHz the PA6T runs at 7W @ 2GHz, so 1.6 GHz could well be quite cool and at least feasable with a similar heatsink and fan placement as the Phase 5 boards.
b) I partly explained above but read here for the ratings:-
http://en.wikipedia.org/wiki/PowerPC_600
http://en.wikipedia.org/wiki/PWRficient
And such a chipset as the Radeon Vision for example could be used. I didn't mean the R700, just some form of Radeon MOBILE solution.
c) Why is it pointless if its the direction FORWARD for the NG Amigas according to Hyperion and A-Eon? and if it runs (for example) an FPGA implementation of various 68K cores that is transferred to the PA6T it could potentially boot classic floppy games etc.
d) I actually agreed with you as it stands now... so?
e) Errrm... Where did I mention the WHOLE of the X1000? I said The PA6T and a Radeon Mobility chip, and then a laptop sized DDR slot (or two) SATA port and USB header and POSSIBLE Wifi (if it wouldn't fit, then a connector to add as an option). Didn't expect them to squeeze the whole X1000 mainboard in the A1200! :D Think about the real estate a 68060 takes up, negate that and you have a heck of a chunk of space. And like I said about my A1200D, if you take the Subway, IDE Express, Delfina etc etc, even though they are seperate and added individually, they all still add up to a virtual system replacement, My idea of the accelerator having these extra features would provide someone like myself something with most if not all those things in a much neater package. And with the advancements in technology over the last 12 years I think what I said is feasable on the amount of space available on an A1200/4000 Accelerator slot. My opinoin, not fact. I also stated it would cost a bit, but I don't believe MORE than the X1000. My opinion again, not fact.
f) And how does that argument hold against the current potential market for the X1000? A-Eon must be CRAAAAZY! ;) :D
Oh yeah, and a troll is still a troll. Snap! :laughing:
Yeah allmost missed that one were he wants to tell us that a miniturized version of the 1500Euro Xena-board could be sold for 500 Euro or less.....
Ok, now your starting to amuse the hell out of me. 1500 Euro gets you a rather hefty rebranded PC case with an ATX PSU, HDD, DVD Drive, Mainboard, GFX card Sound card keyboard etc... And once again I didn't say "Scuweez Teh X1000 inz da A twelve innit!" For F sake man :D
I can't be arsed anymore, i'm quite sure you'll respond and have an answer to all my "facts" but I wont be reading.. ta taa for now! ;)
-
Ok, now your starting to amuse the hell out of me. 1500 Euro gets you a rather hefty rebranded PC case with an ATX PSU, HDD, DVD Drive, Mainboard, GFX card Sound card keyboard etc...
Nope that would be more like 2000+ for the whole shebang...
1500 will just buy you a motherboard (which won't be available seperatly unless your part of that betatester-prepay-scheme).
-
1500Euro Xena-board
Wasn't it "north of 1500 UKP" ? Currently something like > 1760 eur.
-
PA6T uses more than 7W at 2ghz, heck it uses more at lower frequencies even. According to pretty much every source online at 2ghz it has a TDP of 25W. I was quite surprised by this actually, as this power consumption/heat disappation has now been somewhat surpassed by often criticised x86 cpus (bulldozer variant Llano, AMD current generation cpu for example is significantly faster and proportionally significantly lower consumption while including an HD 6xxx class gpu ondie).
Include a gpu (even a mobile variant), ram, etc. onboard, and suddenly there's little hope of an accelerator card for a1200 even being possible unless it's towered.
-
And then what's the point of keeping the 1200 MB since most components would be replaced... ?
Original chipset ? Well first it's likely you'll need a second monitor (or at least a new port in your exisiting monitor) to handle original resolutions.. It's not very easy to use. UAE would be much better, and then there's no need for the slow, big 1200 MB.
-
Sorry for being off topic, but one of the posts here made me think of it.
Does os4.x for classic even support the custom chipset in a way that its possible to use custom chipset based software ? I know AGA is "supported", but in what manner ? Im under the impression its only usable as a basic gfx display (ie. akin to a weak gfx card)
OS4 does not support the custom chipset at all as the 68k chip is switched off. Whole OS run on the PPC.
You have to run classic games and hardware banging apps in EUAE the same as on MorphOS.
If you want to run classic games there is no point in running os4
-
I'm still waiting for a 060 accelerator. :angry:
-
Originally Posted by fishy_fiz :
Sorry for being off topic, but one of the posts here made me think of it.
Does os4.x for classic even support the custom chipset in a way that its possible to use custom chipset based software ? I know AGA is "supported", but in what manner ? Im under the impression its only usable as a basic gfx display (ie. akin to a weak gfx card)
yes, this is how i remember it to be.
-
@JJ and wawrzon
I thought so, but had never seen it written in a clear cut manner before.
Each to thier owm and whatnot, but this is part of why os3.x still appeals to me. While I can appreciate a more "advanced" system, there's appeal in seeing what you can do with the custom chipset hardware for me. Theyre like 2 different sides of the same hobby. Between Amithlon and my classic amiga I'm pretty happy :)