29-Mar-2024 01:40 GMT.
UNDER CONSTRUCTION
Anonymous, there are 73 items in your selection [1 - 50] [51 - 73]
[News] Articia IANN.lu
Posted on 01-Jan-2004 20:24 GMT by ID4 - tHe SuRvIvOr -73 comments
View flat
View list
AGPx8, DDR2, MPEG4 ..... Are some of the new Articia I chipset features!!! Link here AGPx8, DDR2, MPEG4 ..... Are some of the new Articia I chipset features!!! Link here
Articia I : Comment 1 of 73ANN.lu
Posted by Burnt Muller on 01-Jan-2004 19:33 GMT
Wow, they don't even sample the chipsets yet that were annouced 2 years ago and already have a new chipset sampled in paper! MAI is unbelieveable. They should join forces with AInc.
Articia I : Comment 2 of 73ANN.lu
Posted by Anonymous on 01-Jan-2004 19:40 GMT
"Articia I is also equipped with self-repairing and self-testing functions to
ensure performance stability."

I think these will be the most used new features of the chip.
Articia I : Comment 3 of 73ANN.lu
Posted by marcik on 01-Jan-2004 20:27 GMT
In reply to Comment 2 (Anonymous):
> "Articia I is also equipped with self-repairing and self-testing functions
> to ensure performance stability."
>
> I think these will be the most used new features of the chip.

And remeber. It wouldn't be repairing bugs, but a features :)
Articia I : Comment 4 of 73ANN.lu
Posted by Eva on 01-Jan-2004 21:39 GMT
In reply to Comment 1 (Burnt Muller):
LOL :D
Articia I : Comment 5 of 73ANN.lu
Posted by Anonymous on 01-Jan-2004 21:48 GMT
very interesting - so long as they can be harnessed correctly. I wouldnt mind seeing this in either a PowerPC system, or the next batch of MiniITX x86 systems
but recall that they do have non-standard features compared to the "x86-norm" eg being able to access memory from 2 devices etc very nice if its supported, a pain on DMA and data corruption if not
Articia I : Comment 6 of 73ANN.lu
Posted by James Carroll on 01-Jan-2004 22:44 GMT
Just ignoring the trolls for a minute.. :) ..

If they can deliver the goods on time, then its looking good for a G5 Amiga in the not too distant future. IIRC Alan Redhouse was saying they were going to be thinking about doing a G5 Amiga when PPC970 prices were low enough...

Okay just found an exact quote from an amigaworld.net interview... "The 970 chip will not be available for general sampling until well into next year and realistically not generally available at consumer-compatible prices until the end of 2004."

It's looking good.

Anyone know what the projected speeds are for the 970? Around about the 4th quarter of 2004?
Articia I : Comment 7 of 73ANN.lu
Posted by Anon on 01-Jan-2004 23:10 GMT
It's been 4+ years and counting, and the Articia-S isn't even supported correctly yet... so damn boy do NOT use a dirty word like "new features" around here!
Articia I : Comment 8 of 73ANN.lu
Posted by BrianK on 01-Jan-2004 23:30 GMT
In reply to Comment 6 (James Carroll):
James -- In Apple's June 03 announcements Steve Jobs stated the G5 usto hit 3Ghz within 12 months. Well, 6 months have past and we've seen faster G5s @ 2.4Ghz. It appears from the surface 3Ghz by June 04 are close to schedule. Thus, I'd at least expect a 3Ghz G5 by Dec 04 and possibly something a bit faster.
Articia I : Comment 9 of 73ANN.lu
Posted by James Carroll on 02-Jan-2004 00:05 GMT
In reply to Comment 8 (BrianK):
Jeez thank god for that. Its a bit worrying seeing the x86 doing so well, and the PPC lagging behind. Now if they can just keep the power consumption down too we'll haev a pretty cool platform.

Thanks for the info Brian.
Articia I : Comment 10 of 73ANN.lu
Posted by hammer on 02-Jan-2004 01:16 GMT
In reply to Comment 5 (Anonymous):
Refer to NVIDIA's nForce2 chipset for similar functions....
Articia I : Comment 11 of 73ANN.lu
Posted by hammer on 02-Jan-2004 01:25 GMT
In reply to Comment 8 (BrianK):
Where? Apple is still selling them at 2.0Ghz......
Articia I : Comment 12 of 73ANN.lu
Posted by JoannaK on 02-Jan-2004 02:34 GMT
All I can say at this point that specs look interesting.. And if they manage to make it happen in a year (or so) they do have nice potential on market.

Even though I do worry about a bit on realism in this. Afterall..Mai has no expericence on faster busses, they'll need to get a lot new tech to make this happen (AGP8, Elastibus, PCI-X, PCI-express, DDR2)... That's a lot harder project than Articia-S ever was.
Articia I : Comment 13 of 73ANN.lu
Posted by BrianK on 02-Jan-2004 03:24 GMT
In reply to Comment 11 (hammer):
I'll (or alternatively you) can search the web. I remember a 2.4Ghz announcement but the quantities haven't been reached to a good enough level to bring it to market.
Articia I : Comment 14 of 73ANN.lu
Posted by BrianK on 02-Jan-2004 03:53 GMT
In reply to Comment 11 (hammer):
@Hammer,

Here's the article from Forbes I was remembering. http://www.forbes.com/2003/12/22/cx_ah_1222aapl.html

'IBM is said to be ready to deliver a new version of its PowerPC processor to Apple by the end of this year, in time for a revision to the flagship PowerMac G5 line. The main improvement to the PowerPC chip, known internally at IBM as the PowerPC 970fx, will be a reduction of the features on the chip itself, from sizes of 130 nanometers to 90 nanometers...

Beyond 90 nanometers lie transitions to even more advanced manufacturing technologies. After the 90-nanometer process comes the 65-nanometer process in 2005 and 45-nanometers in 2007 or so. IBM is collaborating with several other chipmakers on that transition....

A mid-step between the current systems, which top out with two chips running at 2 Ghz, and systems with chips as fast as 2.6 GHz would be a logical move come January, says Peter Glaskowsky, analyst with Instat/MDR, San Jose, Calif. and editor-in-chief of the influential newsletter Microprocessor Report.

"Speeds of 2.4 to 2.6 GHz would be consistent with exactly where I would expect them to be right now," Glaskowsky says...."

Other sites are putting rumors of 2.4Ghz and 2.6Ghz G5's as the processor for the Apple XServe G5. Apple did update Desktops but not their XServe product line.

We'll all need to stay tuned to the MacIT conference on Jan 7th... If anything new is coming down the pipes it's sure to be announced there.
Articia I : Comment 15 of 73ANN.lu
Posted by smp266 on 02-Jan-2004 04:04 GMT
Definitely need high power/low heat. My 80GB hard drive gets hot enough to cook on, that must add a heavy increase to the case temperature. On the plus side refrigerator manufacturers may see a new market appear.

I've had a go of OpenOffice and I like it. It's almost a complete rip-off of MS Office (shh!)

Those specs are great, but let's see the software.
Articia I : Comment 16 of 73ANN.lu
Posted by BrianK on 02-Jan-2004 04:05 GMT
In reply to Comment 11 (hammer):
@ Hammer...

Use your favorite search engine for PowerPC970fx. They are a bit further out but the you could also search for PowerPC980. If the rumors are close to true it's 980 (G6?) is a 40% faster chip at the same Ghz as the 970 (G5). But, 3.2Ghz version is expected to ship end of 2004.
Articia I : Comment 17 of 73ANN.lu
Posted by James Carroll on 02-Jan-2004 04:23 GMT
In reply to Comment 16 (BrianK):
Sounds too good to be true. I dont really believe many rumours anyway.
Articia I : Comment 18 of 73ANN.lu
Posted by Anonymous on 02-Jan-2004 07:39 GMT
In reply to Comment 7 (Anon):
@Anon

Care to explain what you mean by "the Articia-S isn't even supported correctly yet"?
Articia I : Comment 19 of 73ANN.lu
Posted by Fred on 02-Jan-2004 08:05 GMT
In reply to Comment 18 (Anonymous):
It's best to ignore the trolls...
Articia I : Comment 20 of 73ANN.lu
Posted by Eric_Z on 02-Jan-2004 10:11 GMT
In reply to Comment 11 (hammer):
Just because Apple only sells CPUs running at 2.0 Ghz does not mean that IBM can't manufacture faster CPUs, there are other things to consider like heat dissapation and the increased bus speed as the 970 scales.

AFAIK IBM can deliver 970 CPUs @ 130nm clocking up to 2.5 Ghz or atleast that is what they told the Chinese goverment officials last september(2003).

http://www.atumtech.com/testssss/12.pdf (page 23)

The more public information on IBMs home page can be explained by Apple not wanting to get accused for holding back on faster CPUs.
Articia I : Comment 21 of 73ANN.lu
Posted by Alex Klauke on 02-Jan-2004 10:19 GMT
In reply to Comment 7 (Anon):
You must have a different year counting then.. :)

The Articia S was introduced on Aug, 15, 2001, that's 2 yrs, 4 mths, 18 days..
Check it yourself at mai.com.

Ciao, Alex
Articia I : Comment 22 of 73ANN.lu
Posted by Alex Klauke on 02-Jan-2004 10:44 GMT
In reply to Comment 16 (BrianK):
> Use your favorite search engine for PowerPC970fx.
(90nm PPC970)

Which could be used in PowerMacs as early as MWSF starting Jan. 6th. or shortly
thereafter, according to rumors (www.macrumors.com to be exact.. ) This chip will
most probably max out at 2,8 GHz.

> They are a bit further out but the you could also search for PowerPC980. If the > rumors are close to true it's 980 (G6?) is a 40% faster chip at the same Ghz as > the 970 (G5). But, 3.2Ghz version is expected to ship end of 2004.

This is most likely the version that'll reach 3 GHz in June, as Jobs promised.

But then it's questionable if the Articia I will support that proc, most likely
not, as the 980 will be derivative from the POWER5 (vs. POWER4/970) which means
some design changes that could be important for the chipset support. Btw the
POWER5 is said to reach 4 times(400%!!) the performance of its predecessor.

I think we should be happy if we're getting access to 970's in early 2005. It
also depends on the expected sales of AmigaOne/970, I'd guess it would need a
few thousands of montly sales to be viable at a competetive pricing.

Ciao, Alex
Articia I : Comment 23 of 73ANN.lu
Posted by Johan Rönnblom on 02-Jan-2004 12:31 GMT
To the best of my knowledge, MAI have not thus far been able to deliever a chip that can give even AGP 1x speed, so I agree this sounds rather unrealistic.
Articia I : Comment 24 of 73ANN.lu
Posted by ikir on 02-Jan-2004 12:41 GMT
"It's best to ignore the trolls..."
-------------------------------------

Yes don't feed the trolls. This is a very nice news! But i think that G3 and G4 are enough for me for a while :-P
Articia I : Comment 25 of 73ANN.lu
Posted by Kjetil on 02-Jan-2004 13:52 GMT
In reply to Comment 23 (Johan Rönnblom):
What?
Articia I : Comment 26 of 73ANN.lu
Posted by Anonymous on 02-Jan-2004 13:55 GMT
In reply to Comment 23 (Johan Rönnblom):
The best of you knowleedge is not wery good then :)
Articia I : Comment 27 of 73ANN.lu
Posted by Anonymous on 02-Jan-2004 13:55 GMT
In reply to Comment 26 (Anonymous):
you=your
Articia I : Comment 28 of 73ANN.lu
Posted by Kjetil on 02-Jan-2004 14:20 GMT
In reply to Comment 7 (Anon):
I like the new features of Articia-I and AGP 8x, I hope they finish the product soon, this will be gradate for Amiga System if it’s supported,

I fail to Se what Articia-S or Articia-P or have any ting whit the Articia-I,
Clearly this is to be more multimedia tuned chip, then the previous chips,

It strange how an PowerPC chip not supported in I686 Linux world, makes is a bug, I do not se how that can be twisted inn to being a bug, I most be living on other planet.

Anyway, I agree that the chipset manufacture should have done more to fix the drivers them self, after all they are the ones that know there chips, and there sales depend on making there chips work or supporting there products as it’s called, I hope they do not call the Linux market a small unrelieved market as that is maybe the only big markets for a PPC chipset whit the exception of PowerMac/MacOS.
Articia I : Comment 29 of 73ANN.lu
Posted by Kjetil on 02-Jan-2004 14:24 GMT
In reply to Comment 28 (Kjetil):
“Linux market a small unrelieved market” should be “Linux market a small irrelevant market”
Articia I : Comment 30 of 73ANN.lu
Posted by Leif on 02-Jan-2004 14:35 GMT
In reply to Comment 25 (Kjetil):
AGP x8 *compatible*.
If Im not mistaken, so is Pegaosos I :).
Articia I : Comment 31 of 73ANN.lu
Posted by Amon_Re on 02-Jan-2004 15:07 GMT
In reply to Comment 30 (Leif):
No, if i'm not wrong, AGP8 uses different voltages then AGP2

Cheers
Articia I : Comment 32 of 73ANN.lu
Posted by Leif on 02-Jan-2004 15:25 GMT
In reply to Comment 31 (Amon_Re):
Yeah ,true. However some cards may detect it and adapt.
I know x4 works on Peg1 atleast.
Articia I : Comment 33 of 73ANN.lu
Posted by Kronos on 02-Jan-2004 16:17 GMT
In reply to Comment 28 (Kjetil):
>I most be living on other planet.

Yeap, thats for sure ....

I just can't understand how people will fall so easy for cheap PR and start insiting that
a crystal clear bug is a special feature .... all on the word of a belgian wannabe lawyer *g*
Articia I : Comment 34 of 73ANN.lu
Posted by Seehund on 02-Jan-2004 16:39 GMT
In reply to Comment 6 (James Carroll):
If they can deliver the goods on time, then its looking good for a G5 Amiga in the not too distant future.

But there IS a "G5 Amiga" already, since almost half a year. Apple sold 667,000 PowerMacs last year, of which a sizeable chunk are G5 PowerMacs.

Oh... We have to pretend that there still are "Amigas", and someone has to buy a license to sell them under a meaningless "Amiga" trademark, bundled with AmigaOS and a dongle, and then we would only be allowed to buy our hardware from such a dealer? Sorry. How could I ever forget that?! ;)

Anyway, good news if anything will ever materialise from this PR. There can never be too many chipsets.
Articia I : Comment 35 of 73ANN.lu
Posted by Nate Downes on 02-Jan-2004 17:49 GMT
In reply to Comment 31 (Amon_Re):
You are wrong. AGP 8x 2x etc are speed. Volatges are determined by AGP revision level, in the Peg II's case, AGP 2.0 as opposed to the Peg I's AGP 1.0.

However, the AGP 3.0 standard (the first and only with 8x speed at this time) does use a different voltage than AGP 1.0, making 1.0 cards incompatible with 3.0 controllers. However, it still supports 2.0 devices in AGP 3.0, so an 8x card would still function, in theory. I'm not trying it tho, no 8x cards here to try with.
Articia I : Comment 36 of 73ANN.lu
Posted by Neko on 02-Jan-2004 18:38 GMT
In reply to Comment 35 (Nate Downes):
Nate, stop now.

Pegasos I and II both implement the AGP 2.0 standard. That's a 1.5V spec.

AGP 8x is 1.5V (in 4x "compatibility" mode) and 0.8V - all AGP 8x boards
are REQUIRED to at least TOLERATE operating at 1.5V - i.e. not explode, but
not necessarily work either.

The fact of the matter is that all the current "8x" boards can run at all
speeds and 2.0+ voltage modes (nVidia's next GeForce board, ATI Radeon 9800XT,
XGI Volari). This means even the top of the line AGP 8x workstation cards will
fit into a Pegasos I or II, and display a screen, drivers for MorphOS
notwithstanding.

=Neko=
Articia I : Comment 37 of 73ANN.lu
Posted by BrianK on 02-Jan-2004 20:03 GMT
In reply to Comment 22 (Alex Klauke):
PowerPC 980 is supposed to start around 3.2Ghz in speeds but grow to 4.5-5Ghz range.

2006 is target for the Power 6 and PowerPC 990 looking to grow to 8Ghz in speed.


Amiga getting on the 64-bit PPC wagon will be a good thing for the future.
Articia I : Comment 38 of 73ANN.lu
Posted by Amon_Re on 02-Jan-2004 21:12 GMT
In reply to Comment 32 (Leif):
Depends on the card then, no?

Cheers
Articia I : Comment 39 of 73ANN.lu
Posted by Amon_Re on 02-Jan-2004 21:13 GMT
In reply to Comment 33 (Kronos):
I'll take his word over yours anyday tho

Cheers
Articia I : Comment 40 of 73ANN.lu
Posted by Amon_Re on 02-Jan-2004 21:14 GMT
In reply to Comment 36 (Neko):
Thx for the clarification

Cheers
Articia I : Comment 41 of 73ANN.lu
Posted by Matt Parsons on 02-Jan-2004 21:23 GMT
In reply to Comment 39 (Amon_Re):
Given the Evidence, I'm pretty certain it's a bug... If you look too you should come to the same conclusion.
Articia I : Comment 42 of 73ANN.lu
Posted by Amon_Re on 03-Jan-2004 00:04 GMT
In reply to Comment 41 (Matt Parsons):
"Given the evidence" is just the problem, nobody has presented any evidence in eighter direction, there are people saying it's a bug & people saying it ain't a bug.

Personally, i'm the "innocent untill proven guilty" type, without anything solid i won't say it's a bug.
Give me some solid evidence (and not the "i know but ain't allow to say it" type of lines) & i'll believe you & the rest.

Cheers
Articia I : Comment 43 of 73ANN.lu
Posted by Matt Parsons on 03-Jan-2004 00:27 GMT
In reply to Comment 42 (Amon_Re):
Well, at first I didn't care... but then I became interested by how much fuss this was causing.

It you look around you will see that the chip does not do what it is supposed to, MAI don't seem to want to do anything more with the Chip, Many Companies have deserted the Chip, The so called "feature" does not make much sense (to me at any rate)... but what really made me think it was a bug was when certain individuals started blaming other chips/software for the problems before announcing the "feature".

But this is all academic anyway. I seriously doubt an A1 is going to ever be in a situation to use this "feature", and should the time come it becomes a problem I expect MAI to have new chips out.
Articia I : Comment 44 of 73ANN.lu
Posted by Kjetil on 03-Jan-2004 00:31 GMT
In reply to Comment 41 (Matt Parsons):
What evidence? the data corruption can only to be found whit the DMA enabled on the via chipset, the rest of the time my computer is stable,

there for the issue given from Ben/Hyperion for the fault is more then good, to explain the problem, the Articia S chipset will need to cashe the data flow while writing or reading form the other source, an data bus can not be really be accessed at once by tow end points, how ever the delay time can be improved by improving the efficiency in how it's used, this what I think that call "Bus-Centric-Architecture", http://www.mai.com/products/BRA660R2.0.pdf well if you read the few lines about it says it out of spec; "more balanced architecture serving both CPU and peripheral bus requirements then the traditinal "CPU-Centric" design approach"

if that type cashing conflict whit dma, DMA will try to request data by it's own, and there for conflict whit the articia S caches, data output may end up being the wrong bits of data, so the solution to the problem will be that via drivers need to flush/fill cache cache at right time when Articia S is ready to read/write data, or disable the DMA. The Via driver might expect an transparent bus (not waiting for a ready state on the bus), and that is not what Articia S requires. Aritica S requires cooperation to work whit DMA.

the problem will be when the Articia cache is filled whit some data, and the DMA tryes accessing it when it's not ready.

the Pegasus fix might be that the via chipset is chip select or clock signal being disabled while Articia busy, or some thing similar; how ever that approach effects alott of things, one might expect some DMA devices to fail, it the drivers fail to fetch data when thay think the DMA is ready.
Articia I : Comment 45 of 73ANN.lu
Posted by Kronos on 03-Jan-2004 02:44 GMT
In reply to Comment 44 (Kjetil):
The Articia-prob is with DMA, and switching of DMA will therefore cure it, not matter
wether there is another prob in the VIA or not.


"Bus-centric" is a nice buzz-word, and it might even be that MAI were the 1st to implement
it on the PPC as they claim, but that was 3-4 years ago, and others (IBM,Marvell,Apple)
have long catched up and passed by. Don't even think about comparing it with anything
in the x86-world that was produced in the last 3 years....


@Amon_Re
Problem is that Ben has allready several times prooven that he completly lacks technical
understanding, and can only repeat what others have told him (>1GB/sec on SDR-133
anyone ? ). This behaviour (of the Articia) is what has kept it from being commercially used
for ~3 years (and now a few 100 or 1000 boards for an obscure retro-community don't count), there is
still not a single OS actually supporting it, and wasn't even documented officially until recently.
So discussing wether it is a bug or feature is a mere technicallity, commercially it has allready
prooven itself to be a fatal bug.
Articia I : Comment 46 of 73ANN.lu
Posted by Johan Rönnblom on 03-Jan-2004 15:15 GMT
In reply to Comment 26 (Anonymous):
Then, if you know so much better, could you tell me which MAI chip
does deliver AGP1x transfer speed?


Hint: There's a huge difference between writing AGP whatever-x in the
specs, and providing a physical, existing chip that can sustain a data
transfer rate at 256MB/s or more.
Articia I : Comment 47 of 73ANN.lu
Posted by Alkis Tsapanidis on 03-Jan-2004 15:16 GMT
In reply to Comment 44 (Kjetil):
The Pegasos fix is *NOT* VIA specific.
Articia I : Comment 48 of 73ANN.lu
Posted by Matt Parsons on 03-Jan-2004 15:33 GMT
In reply to Comment 45 (Kronos):
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
"Bus-centric" is a nice buzz-word, and it might even be that MAI were the 1st to implement
it on the PPC as they claim, but that was 3-4 years ago, and others (IBM,Marvell,Apple)
have long catched up and passed by. Don't even think about comparing it with anything
in the x86-world that was produced in the last 3 years....
<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<

Hey Kjetil, haven't you heard? Busses are out, Point to Point serial links are the future!!! ;-)
Articia I : Comment 49 of 73ANN.lu
Posted by Amon_Re on 03-Jan-2004 16:12 GMT
In reply to Comment 43 (Matt Parsons):
This feature might or might not been known, i have no access to developer documentation of the ArticiaS, the situation remains fused.

Cheers
Articia I : Comment 50 of 73ANN.lu
Posted by Amon_Re on 03-Jan-2004 16:14 GMT
In reply to Comment 44 (Kjetil):
Don't forget that the ArticiaS also features a "floating buffer" on it's busses.

Cheers
Anonymous, there are 73 items in your selection [1 - 50] [51 - 73]
Back to Top