27-Apr-2024 23:26 GMT.		 UNDER CONSTRUCTION
Anonymous, there are 69 items in your selection [1 - 50] [51 - 69]
[News] Pictures of the MicroA1 "Industrial" motherboardANN.lu
Posted on 17-Oct-2004 12:25 GMT by Not ArticiaP69 comments
View flat
View list
On this link you will see three pictures of the MicroA1 "Industrial" motherboard: http://lain.ziaspace.com/~ryu/a1. On the pictures you can see that it still uses the ArticiaS northbridge, as well as the VIA 686B southbridge, but a standalone Sil0680 IDE controller is there as well.
Pictures of the MicroA1 "Industrial" motherboard : Comment 1 of 69ANN.lu
Posted by Neil Thomas on 17-Oct-2004 11:25 GMT
Interesting, looks like a PC/104 style PCI connector next to the more familiar slot type - can Eyetech confirm this? Hell, no reason why one couldn't have a PMC site in there as a build option.

I would say though if trying to break into the industrial market, then would it not be better to be a 6U or 3U format for CompactPCI or VME? Though VME is long in the tooth (try to catch up with PCI though) it is a widely accepted standard never the less.

Is this for factor perhaps better for fitting multiple boards into a 19" rack?? If so.... how would the two communicate? I guess it could be over a PCI bridge, fully asymetric system, but held up by the 33MHz 32 Bit PCI bus. Now if that were PCI-X...... (ala, the next Artica.. P is it?)

Neil Thomas, AKA MiniBobF
Pictures of the MicroA1 "Industrial" motherboard : Comment 2 of 69ANN.lu
Posted by Adam Kowalczyk on 17-Oct-2004 12:15 GMT
In reply to Comment 1 (Neil Thomas):
The slot is a PCI-104 connector. PMC was looked at as a CPU module option but took up too much board space. The issue was that the location of the through hole connectors made it impossible to do the PCB layout routing. One of the design goals of the "I" version was to include a method of expansion. Initially, it was going to be a custom connector, but PCI-104 turned out to be a better solution as it opened up the possibility of using many cards that are already available.
Pictures of the MicroA1 "Industrial" motherboard : Comment 3 of 69ANN.lu
Posted by Joe on 17-Oct-2004 12:28 GMT
Looks like a nice board but is ruined by the crazy price.
Pictures of the MicroA1 "Industrial" motherboard : Comment 4 of 69ANN.lu
Posted by Anonymous on 17-Oct-2004 13:22 GMT
In reply to Comment 2 (Adam Kowalczyk):
>but PCI-104 turned out to be a better solution as it opened up the possibility of using many cards that are already available.

Do you have any examples of PCI-104 cards? I am not familiar with those but it sounds interesting.
Pictures of the MicroA1 "Industrial" motherboard : Comment 5 of 69ANN.lu
Posted by ikir on 17-Oct-2004 13:23 GMT
Very nice. Thanks for the informations Adam. I already have a A1XE G4 1000Mhz, but maybe i'll buy a Micro in the next 6 months. Keep up the good work, and thanks for your effort as well.
Pictures of the MicroA1 "Industrial" motherboard : Comment 6 of 69ANN.lu
Posted by Anonymous on 17-Oct-2004 13:54 GMT
looks like a proto :
at bottom of fist pic one pin is twister and the stickers look so
artisanal
Pictures of the MicroA1 "Industrial" motherboard : Comment 7 of 69ANN.lu
Posted by Zylesea on 17-Oct-2004 14:55 GMT
A lot of questions.

Why does it still use a via 686b? It's outdated and buggy.
What about usb2.0?
Why a Radeon 7000?
And no dvi out?
Why no G4?
Why not for an attractive price?
Pictures of the MicroA1 "Industrial" motherboard : Comment 8 of 69ANN.lu
Posted by Neil Thomas on 17-Oct-2004 15:58 GMT
In reply to Comment 7 (Zylesea):
Becasue in the real world designs take time to configure and get working correctly, it takes a great deal of cash and time inuptus to get upto, let alone beyond the cuttinr edge curve. It's easy for you to suggest specs of components that you want, but it is not so easy to implement (not just from a technical point of view). Budgets, supply of components, CAD time & software development time, these are all valid answers to your questions. Other products start the developement lifecycle from different positions so are closer or further away from the development curve. That's life and the balance of economics. Engineering isn't about getting the best, it's a compromise between what is possible, and what is necessary. And it all has a price.

Neil Thomas, AKA MiniBobF
Pictures of the MicroA1 "Industrial" motherboard : Comment 9 of 69ANN.lu
Posted by Anonymous on 17-Oct-2004 16:06 GMT
In reply to Comment 7 (Zylesea):
Look, I'm no fan of the design decisions or price point of the board but I think you could answer these questions easily yourself:

>Why does it still use a via 686b? It's outdated and buggy.

Like most of VIAs chips - your point exactly? At least they 'fixed' the IDE issue - albeit at the premium cost of board space for another chip (and connector?)

>What about usb2.0?

Not supported by the chips used - is 2.0 hi-speed mode really a requirement in Industrial Application (I honestly don't know, so I won't comment.)

>Why a Radeon 7000?

This is for industrial use....that's a good enough chip IMO

>And no dvi out?

See above

>Why no G4?

Heat etc.- intended application - see above. Of course I stil think they made a mistake by not having a slot/socket.

>Why not for an attractive price?

Only Eyetech can answer this question - please note we have not seen volume production prices yet though
Pictures of the MicroA1 "Industrial" motherboard : Comment 10 of 69ANN.lu
Posted by Undercover on 17-Oct-2004 17:25 GMT
http://www.vesalia.de/?V02b0f1557504f1570575856520b514e01101f0954414752050140090a1e31115779633a223c6373744e03263f2a2a76610539646d6d2260796a7179257f647f220316424842096945594a0144555c1c515c6d46712#m
according to that info, it will be possible in the future to upgrade the µ-A1-C to a G4 cpu.
Pictures of the MicroA1 "Industrial" motherboard : Comment 11 of 69ANN.lu
Posted by Alkis Tsapanidis on 17-Oct-2004 17:35 GMT
In reply to Comment 10 (Undercover):
The normal version is upgradable, the industrial is not.
Pictures of the MicroA1 "Industrial" motherboard : Comment 12 of 69ANN.lu
Posted by Adam Kowalczyk on 17-Oct-2004 20:18 GMT
In reply to Comment 4 (Anonymous):
Just go to www.pc104.org or do a google on them.....there are too many to enumerate.

Adam
Pictures of the MicroA1 "Industrial" motherboard : Comment 13 of 69ANN.lu
Posted by JoannaK on 17-Oct-2004 20:55 GMT
In reply to Comment 12 (Adam Kowalczyk):
It's nice, but IIRC uA1-I does not seem to be fully 104 compliant.. I'd need to check my 104 docs to be sure.
Pictures of the MicroA1 "Industrial" motherboard : Comment 14 of 69ANN.lu
Posted by Neko on 17-Oct-2004 21:51 GMT
Does the Radeon really need that huge heatsink?

Neko
Pictures of the MicroA1 "Industrial" motherboard : Comment 15 of 69ANN.lu
Posted by Anonymous on 17-Oct-2004 23:28 GMT
In reply to Comment 2 (Adam Kowalczyk):
I guess that you cannot use PCI-104 slot in uA1-I if you have the PCI card installed in the PCI slot.

Let's calculate.
Articia S (northbridge) has two PCI buses:

A) BUS-0. Here up to 5 PCI bus masters can be supported. Let say that to the BUS-0 they connected:

1. Southbridge (VIA 686B with ATA-100 (not used), USB 1.1, AC'97 (not used), PS/2)
2. PCI slot
3. ATA-133 controller (Sil O680)
4. FireWire (VT6307)
5. Sound (CMI7838)

B) BUS-1. This bus can be used as AGP x2 bus with sideband address bus protocol supported. In this mode any other PCI devices cannot be used on this bus.

It means that BUS-1 is configured as 66MHz PCI without the AGP features. In this mode BUS-1 supports up to 2 PCI bus masters.

They probably connected here:
1. Radeon 7000 in AGP x0 mode
2. Network (RTL8110)

PCI-104 slot cannot be connected to the BUS-1 because PCI-104 devices do not support 66MHz capabilities.

So where PCI-104 slot is connected?

I guess that PCI-104 slot is connected in parallel with the PCI slot on BUS-0. It means that it shares all signals of the PCI slot and you must select if you want to use PCI slot or PCI-104 slot. You cannot use both.

BTW. The PCI-104 slot is nothing more than downgraded PCI slot with untypical connector. The PCI-104 spec is here: http://www.pc104.org/technology/PDF/PCI-104%20v1_0.pdf
Pictures of the MicroA1 "Industrial" motherboard : Comment 16 of 69ANN.lu
Posted by Adam Kowalczyk on 18-Oct-2004 00:57 GMT
In reply to Comment 15 (Anonymous):
WRT PCI BUS 0, a PLD was added to allow 6 bus masters. The PCI 104 slot is on BUS 0.

Adam
Pictures of the MicroA1 "Industrial" motherboard : Comment 17 of 69ANN.lu
Posted by Adam Kowalczyk on 18-Oct-2004 00:58 GMT
In reply to Comment 13 (JoannaK):
The slot on the "I" is PCI-104 compliant. Check your definitions carefully.
Pictures of the MicroA1 "Industrial" motherboard : Comment 18 of 69ANN.lu
Posted by Anonymous on 18-Oct-2004 05:15 GMT
In reply to Comment 15 (Anonymous):
Hi!

By the waya, it's a bit weird to put a VIA686B (which include IDE/UDMA controller and sound), an external sound chip (CM873) and an external IDE/UDMA controller (Sil...).
It must raise the cost for almost nothing...

Bye
Pictures of the MicroA1 "Industrial" motherboard : Comment 19 of 69ANN.lu
Posted by Anonymous on 18-Oct-2004 06:16 GMT
In reply to Comment 18 (Anonymous):
Yes, they have spent the last year or so complaining about the 686B, and yet they choose to use this very same southbridge on this completely new board, with additional chips instead, adding cost, occupying space. Why not another one? Nah, pointless to argue about that at this point, its of course their call, and this is how they wanted it.

BTW: Will these boards be sold as fully working, completely tested products, with a real warranty, or is it still "it may turn out not to be fully working in the end, but hey, blame yourself for buying our products"? These boards ARE sold without any fine print on the Eyetech website now, are they?
Pictures of the MicroA1 "Industrial" motherboard : Comment 20 of 69ANN.lu
Posted by Anonymous on 18-Oct-2004 06:29 GMT
In reply to Comment 6 (Anonymous):
You are pathetic. Don't you have anything else to say than whine about stickers ?
Pictures of the MicroA1 "Industrial" motherboard : Comment 21 of 69ANN.lu
Posted by Anonymous on 18-Oct-2004 06:31 GMT
In reply to Comment 11 (Alkis Tsapanidis):
Wrong, MicroA1-C has CPU on board MicroA1-I can be upgraded to have G4.
Pictures of the MicroA1 "Industrial" motherboard : Comment 22 of 69ANN.lu
Posted by Anonymous on 18-Oct-2004 06:35 GMT
In reply to Comment 19 (Anonymous):
As much as the Pegasos boards.
Pictures of the MicroA1 "Industrial" motherboard : Comment 23 of 69ANN.lu
Posted by kalmar on 18-Oct-2004 06:38 GMT
Neat board. Does anyone know who has designed it and where it is going to be manufactured?
Pictures of the MicroA1 "Industrial" motherboard : Comment 24 of 69ANN.lu
Posted by JoannaK on 18-Oct-2004 06:44 GMT
In reply to Comment 17 (Adam Kowalczyk):
Well. Those pictures aren't perfect for analysis but PC/104-Plus (PCI-104) Compliance reguires quite a strict specs to be followed. Both electrical and mechanical. And based on those pictures, closest I could it be reaching would be 'Bus-compatible', cause on uA1-I board has plenty variations from PCI-104 diemsions, component clearing requirements, mounting support etc.

And considering board allready having multitude of PCI connected chips (6 + normal extansion slot?) in the same PCI-bus as PCI-104 I seriously doubt it being Bus-compatible either. Afterall PCI-104 master is supposed to be able to handle Four add-on cards on a stack with slightly twisted timings.

I hope you'll test this before selling it to clients.. Industrial people aren't as forgiving as Amiga users.
Pictures of the MicroA1 "Industrial" motherboard : Comment 25 of 69ANN.lu
Posted by Anonymous on 18-Oct-2004 06:45 GMT
In reply to Comment 21 (Anonymous):
"Wrong, MicroA1-C has CPU on board MicroA1-I can be upgraded to have G4."

Wrong, µ-A1-C has a CPU card like the A1 XE and could, at least in theory, be upgraded to a G4. The "industrial" version of the micro A1 has the CPU soldered directly onto the PCB and can not be replaced.
Pictures of the MicroA1 "Industrial" motherboard : Comment 26 of 69ANN.lu
Posted by zylesea on 18-Oct-2004 06:49 GMT
In reply to Comment 23 (kalmar):
Atum Computers aka MAI Logic.
Pictures of the MicroA1 "Industrial" motherboard : Comment 27 of 69ANN.lu
Posted by Anonymous on 18-Oct-2004 07:01 GMT
In reply to Comment 22 (Anonymous):
"As much as the Pegasos boards"

If that is the case, then it's great news. Good to see that Eyetech finally is ready to shoulder some responsibility towards end users now, because the A1 SE and XE situation was simply not acceptable, at least not in the real world outside the Amiga(tm) fanclub.

But is this really official? Does the Eyetech webpage EXPLICITLY say so? I really think that's required now, as one step in the process of trying to raise the confidence for the Amiga One product outside the tightest little group of supporters.
Pictures of the MicroA1 "Industrial" motherboard : Comment 28 of 69ANN.lu
Posted by Adam Kowalczyk on 18-Oct-2004 11:30 GMT
In reply to Comment 24 (JoannaK):
Of course it has been tested. There has never been any claim that it is PC/104+ compliant. The slot is PCI-104 compliant. PCI/104+ compliance would require the ISA connector as well which was not included. As for stacked modules, the PCI-104 connector shares REQ/GNT lines in order to achieve stackability.
Pictures of the MicroA1 "Industrial" motherboard : Comment 29 of 69ANN.lu
Posted by Johan Rönnblom on 18-Oct-2004 13:31 GMT
In reply to Comment 27 (Anonymous):
I don't think it matters much what they SAY. They didn't say "buyer
beware" about the non-micro AmigaOnes either. What matters is what
they DO. And here they have clearly showed this: They will not repair
or replace products that turn out to have issues on a large scale,
because they don't have the resources to do that. They will take care
of single units that have individual problems, but they if there is a
general problem, people will have to blame themselves.

So, it will of course be very interesting to see independent reviews
and test results with the micro A1. It has been claimed here on ANN
that this board is fixed and that standard Linux drivers will work on
it. If that is the case, it should be fairly easy to confirm whether
they have solved the problems discovered in the non-micro version, or
not.
Pictures of the MicroA1 "Industrial" motherboard : Comment 30 of 69ANN.lu
Posted by JoannaK on 18-Oct-2004 13:46 GMT
In reply to Comment 28 (Adam Kowalczyk):
Hmm.. Sounds familiar. I've heard allmost same words coming from Eyetech couple years ago. For your sake I hope you'll learned frome those years.
Pictures of the MicroA1 "Industrial" motherboard : Comment 31 of 69ANN.lu
Posted by Total Quality Experience on 18-Oct-2004 15:44 GMT
In reply to Comment 28 (Adam Kowalczyk):
> Of course it has been tested.

*giggle*

Is there an echo in here? I've heard this for almost four years now.
Pictures of the MicroA1 "Industrial" motherboard : Comment 32 of 69ANN.lu
Posted by Andreas Wolf on 18-Oct-2004 17:52 GMT
In reply to Comment 26 (zylesea):
They do design but they do not manufacture. A guess would be the manufacturer is FIC (as with the former AmigaOnes).
Pictures of the MicroA1 "Industrial" motherboard : Comment 33 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 00:19 GMT
In reply to Comment 16 (Adam Kowalczyk):
> WRT PCI BUS 0, a PLD was added to allow 6 bus masters. The PCI 104 slot is on BUS 0.

What was the main idea of including this PLD chip?

A) To solve problems with ArticiaS and VT82C686B communication? (VIA ATA-100 controller problems in UDMA mode)

If this PLD fixes this problem why additional ATA controller (Sil 0680) has been added on the uA1-I board?

If this PLD does not solve this problem why Sil 0680 are not present on the uA1-C board? Isn't it because C version was designed earlier, when VIA ATA-100 controller problem wasn't known?

B) Only for adding 6 (instead of 5 supported by Articia S) bus masters devices to BUS-0?

If yes, why support for only 6 bus master has been added? Why not 7?

Support for 6 devices only means that one PCI device (RTL8110 NIC controller) had to be placed on BUS-1. It means that BUS-1 have to be configured as 66MHz PCI without the AGP features. This downgrades Radeon 7000 capabilities from AGP x2 to AGP x0 (= PCI).
Why they did it? Because Articia S do not support AGP x2 features in spite of MAI claims?
Pictures of the MicroA1 "Industrial" motherboard : Comment 34 of 69ANN.lu
Posted by priest on 19-Oct-2004 04:09 GMT
In reply to Comment 33 (Anonymous):
counter speculation:

"What was the main idea of including this PLD chip?"

To allow 6 bus masters. (if you could read, you would have figured it out)

"A) To solve problems with ArticiaS and VT82C686B communication? (VIA ATA-100 controller problems in UDMA mode)"

No. Because I model does not use 686B for IDE.

"If this PLD fixes this problem why additional ATA controller (Sil 0680) has been added on the uA1-I board?"

It is not said to fix anything. 0680 gives UDMA133 (not sure if it somehow also enables compact flash...).

"If this PLD does not solve this problem why Sil 0680 are not present on the uA1-C board?"

If the PLD is to enable 6 bus masters, perhaps 6 bus masters are not needed without PC104.

"Isn't it because C version was designed earlier, when VIA ATA-100 controller problem wasn't known?"

More likely it's because C version is more of a MAI desing rather than Ayetech design.

"B) Only for adding 6 (instead of 5 supported by Articia S) bus masters devices to BUS-0?
If yes, why support for only 6 bus master has been added? Why not 7?"

Only one more was needed?

"Support for 6 devices only means that one PCI device (RTL8110 NIC controller) had to be placed on BUS-1."

Via, Nic, 3*PCI/4*PC104, Sil0680. hmmm... it seems like that.... unless the 4*PC104 limits the use of something else...

"It means that BUS-1 have to be configured as 66MHz PCI without the AGP features. This downgrades Radeon 7000 capabilities from AGP x2 to AGP x0 (= PCI). Why they did it? Because Articia S do not support AGP x2 features in spite of MAI claims?"

Perhaps they did not do that? (there's something we do not know)
Perhaps Radeon7000 does not gain anything from AGP2x ?


btw. what's "PLD" in the first place ? ProgrammableLatchDevice ??? ;)
Pictures of the MicroA1 "Industrial" motherboard : Comment 35 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 04:39 GMT
In reply to Comment 34 (priest):
>Via, Nic, 3*PCI/4*PC104, Sil0680.

Do not forget the sound chip and IEEE1394.
Pictures of the MicroA1 "Industrial" motherboard : Comment 36 of 69ANN.lu
Posted by priest on 19-Oct-2004 05:15 GMT
In reply to Comment 35 (Anonymous):
Then... do every PCI device need "bus mastering"?
Otherwise 9 bus masters would be needed...

btw. is PCI bus mastering directly related to the PCI device's capability to do DMA transfers?
Pictures of the MicroA1 "Industrial" motherboard : Comment 37 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 06:21 GMT
Only 36 comments you haven´t reached 200 yet? Seems like things are slowing down
Pictures of the MicroA1 "Industrial" motherboard : Comment 38 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 06:29 GMT
In reply to Comment 34 (priest):
Programmable Logic Device...They come in different flavours....
CPLD EPLD etc...
Pictures of the MicroA1 "Industrial" motherboard : Comment 39 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 07:03 GMT
Wow, this PC-104 slot is actually very cool. Did a little search on on pc104.org and found a dual PCMCIA (TYPE I/II/III) card, firewire, USB 2.0. SCSI controllers, HD controllers, ethernet controllers, video capture and hardware compresion. These cards are verry small and some are even stackable.

It looks a lot like the original Amiga Walker concept. Someone should make this case in miniITX form factor (if it's possible because of IP etc...).
Pictures of the MicroA1 "Industrial" motherboard : Comment 40 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 07:09 GMT
In reply to Comment 39 (Anonymous):
Hmm.. after reading the PC-104 specification, it's just an 16bit ISA bus....So don't expect any useful bandwidth..... :-(

Ah well. I must wait for the PCI riser card then.
Pictures of the MicroA1 "Industrial" motherboard : Comment 41 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 07:12 GMT
In reply to Comment 40 (Anonymous):
Am I blind or something....it's PCI-104 not PC-104.....

So I'm excited once more :-)
Pictures of the MicroA1 "Industrial" motherboard : Comment 42 of 69ANN.lu
Posted by priest on 19-Oct-2004 07:19 GMT
In reply to Comment 40 (Anonymous):
"I must wait for the PCI riser card then"

I thought any ITX PCI ricer card is usable straight out of the box? Is it not?
Pictures of the MicroA1 "Industrial" motherboard : Comment 43 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 07:33 GMT
In reply to Comment 42 (priest):
I don't know...I never came across any riser card...But then again...I never searched for it either....

But in the meanwhile I found an utra wide 2 SCSI PCI-104 card and two PCMCIA type II (or one type III) with three Firewire ports PCI-104 card.... ;-)

All stackable ofcourse. So now I hope that there are enough interrupt lines available.
Pictures of the MicroA1 "Industrial" motherboard : Comment 44 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 07:34 GMT
In reply to Comment 42 (priest):
I think you are right. Normal ITX riser should work, I don't see any reason why it would not.
Pictures of the MicroA1 "Industrial" motherboard : Comment 45 of 69ANN.lu
Posted by priest on 19-Oct-2004 07:34 GMT
btw. ... it does not seem to have the "composite video out" as it was supposed to have ...

Also, it seems to have only one IDE connector... so only two IDE devices are possible? (I do now want to buy pricey CD+DVD+RW combo drive...)
Pictures of the MicroA1 "Industrial" motherboard : Comment 46 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 07:51 GMT
In reply to Comment 44 (Anonymous):
I did a little reseach and the single slot riser cards seem to work on all boards but then again, it's just a right angle pass-thru. But the dual PCI riser cards seem to be VIA EPIA specific.

And there are a couple of three slot riser cards, but I couldn't find any info on them.
Pictures of the MicroA1 "Industrial" motherboard : Comment 47 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 07:53 GMT
In reply to Comment 45 (priest):
No, it has second 44 pins one ( 2.5" connector. The same as in an a600 and a1200).
I hope it's connected to the second IDE channel.
Pictures of the MicroA1 "Industrial" motherboard : Comment 48 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 08:07 GMT
In reply to Comment 46 (Anonymous):
http://www.adexelec.com/pci32.htm has one called the P4TX21.

It's a Two PCI slot riser card, but it is not guaranteed to work on all boards.
Pictures of the MicroA1 "Industrial" motherboard : Comment 49 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 08:42 GMT
In reply to Comment 48 (Anonymous):
OK and it seems that Vesalia is going to sell a 3 slot right turn version (http://www.vesalia.de/?V02b0f1557514f1572575d50520b514e01101f095441475205014d1718457212087f766c34323e756e0c0a2c657f2e392c1c293b6b62296c6c7b3a362d07191936474452121250341d11424415614f5e5742337)
Pictures of the MicroA1 "Industrial" motherboard : Comment 50 of 69ANN.lu
Posted by Anonymous on 19-Oct-2004 10:20 GMT
In reply to Comment 34 (priest):
>> It means that BUS-1 have to be configured as 66MHz PCI without
>> the AGP features. This downgrades Radeon 7000 capabilities from
>> AGP x2 to AGP x0 (= PCI). Why they did it? Because Articia S do not
>> support AGP x2 features in spite of MAI claims?

> Perhaps they did not do that? (there's something we do not know)

They did it. This is what we just know from the comment #16 (from Adam Kowalczyk).
According to him a PLD was added to allow *6* bus masters on BUS-0.

To BUS-0 six bus masters are connected:
1. Southbridge VIA 82VT686B,
2. one PCI slot (for ONE PCI card only),
3. ATA-133 controller (Sil O680),
4. FireWire (VT6307),
5. Sound (CMI7838) and
6. one PCI-104 slot.

The 7th bus master included on the uA1-I board (RTL8110 NIC controller)
had to be connected to BUS-1 together with the ATI Radeon 7000 chip.

This downgrades ATI Radeon 7000 capabilities from AGP x2 to AGP x0 (= PCI).
It means Radeon on the uA1-I board can work in PCI mode only.

But why they did it? No need for AGP in Amiga?

> Perhaps Radeon7000 does not gain anything from AGP2x ?

Maybe you should send your comment to ATI? ;-)
Anonymous, there are 69 items in your selection [1 - 50] [51 - 69]
Back to Top