[News] Yet another PCI-board for Amiga | ANN.lu |
Posted on 24-Sep-2000 14:33 GMT by Teemu I. Yliselä | 14 comments View flat View list |
As rumoured, DCE are also working on a PCI implementation for the Amiga. The board will connect to a CyberstormPPC/mk III card in big box Amigas and to a BlizzardPPC card in an A1200. A Voodoo3 graphics card will also be available. Here's the press release from amiga-news.de (in german).
|
|
Yet another PCI-board for Amiga : Comment 1 of 14 | ANN.lu |
Posted by Anonymous on 23-Sep-2000 22:00 GMT | Yahoo! It's about time someone had the sense to make such a
card for Big Box Amigas including the A3000(t) and A4000(t).
There are as many if not more big box Amigas in use as there
are A1200 whether you want to believe it or not!
Like, I said it's about time. They are going to fare much better
than the two card manufacturers that have cut half if not more
of their own market completly out by not making such a card
for big box Amigas. |
|
Yet another PCI-board for Amiga : Comment 2 of 14 | ANN.lu |
Posted by Daniel Allsopp on 23-Sep-2000 22:00 GMT | Hmm, buyt it's still limited to the PPC owners, and according to
Mediator press realse from Elbox, it's a complete waste of time. I
have an excellent Amiga, the only thing missing is a PPC, but I will
not pay 500 pounds for a half manufacured (no backside cache), out of
date, slow product.
The A4000 mediator should include a new zorro busboard (as
supplied with A4000 Power/Winner Tower) with 3 ZorroIII slots and 4
PCI slots. Problem solved for A4000 owners. |
|
Yet another PCI-board for Amiga : Comment 3 of 14 | ANN.lu |
Posted by Someone on 23-Sep-2000 22:00 GMT | What I think is that we should wait and see what will come up from
that all Elbox, DCE, Eyetech "products". I wouldn`t like to buy
sth, which on another day, can appear as too slow becouse
another better PCI board will be released. |
|
Yet another PCI-board for Amiga : Comment 4 of 14 | ANN.lu |
Posted by Sam Smith on 23-Sep-2000 22:00 GMT | This is all very well, but DCE have yet to re-release the Cyberstorm
PPC cards yet. Without them there will be no additional users to the
existing PPC user base. |
|
Yet another PCI-board for Amiga : Comment 5 of 14 | ANN.lu |
Posted by Sharwin & Rakesh Raghoebardayal on 23-Sep-2000 22:00 GMT | In reply to Comment 2 (Daniel Allsopp): It is very sad to hear such a comment: It seems that you simply don't know what you are talking about. The PPC603/604 cards are by no means dated: The most commonly used Apple Macintoshes use G3 processors at around 300MHz. G3 processors do have a backside cache but are only about 25-30% faster at 300MHz than a 604 at 233MHz. This is just a little bit more than the difference between a 68000 and a 68010! It is almost foolish to keep yourself from buying a phase5 PPC board: A new world opens with those cards... If you are waiting for a PPC-only acceleratorcard without a G4, keep in mind that it is nearly useless for 68k software as it runs MUCH slower than on an 060. Nearly the entire PPC is used for emulating a 68k program, which means that PPC programs suffer from this as well. Just take a look at Frogger for MorphOS or a PowerMac running 68k software... |
|
Yet another PCI-board for Amiga : Comment 6 of 14 | ANN.lu |
|
Yet another PCI-board for Amiga : Comment 7 of 14 | ANN.lu |
Posted by Amifan on 23-Sep-2000 22:00 GMT | In reply to Comment 5 (Sharwin & Rakesh Raghoebardayal): This would only appear when you're using a CPU demanding 68k program and a cpu demanding PPC program at the same time...
and there's always a priority meganism.
a G3 300Mhz "only" 25% faster then a 233Mhz 604e???
Are your calculations based on clockspeeds??
Ever used sspeed to calculate your mips and flops, and compared it to the performance of the same PPC CPU WITH L2 cache??
Another thing: The perfomance of the 604e on the IBM PPC site are based on 604e's with a 83Mhz FSB.
Some cyberstormPPC's have only a 50Mhz FSB....
Only 25 % doesn't sound much on below 100MIPS (060/50) systems, but remember that a G3 500Mhz does more then 1.2 GIPS (giga instuctions per second) so 25%300MIPS! that's almost 4 times the MIPS of a 060 (80 MIPS).
Take a G4.....it can do both GIPS and GFLOPS.....
You can find on IBM's PPC perfomance page that a 740 (G3 without L2 controller) 300Mhz without L2 cache can do 642MIPS.
Can someone run Sysspeed on a 604e 233Mhz and place a comment with the MIPS and FLOPS??
I need a fast G3/G4 card WITH L2 cache. 400Mhz and up. |
|
Yet another PCI-board for Amiga : Comment 8 of 14 | ANN.lu |
Posted by Sharwin & Rakesh Raghoebardayal on 23-Sep-2000 22:00 GMT | In reply to Comment 7 (Amifan): When talking about high speed processors it is quite logical that we are especially emphasizing the CPU-demanding programs. Indeed there is a priority mechanism, although this will only make the 68k programs slower than they already are under emulation...
Our statement about the G3 versus PPC604 are based upon benchmarks from Motorola. Comparing the 740 with a 604 is simply nonsense: The 604 has a full 64-bits bus interface, 3 integer units instead of 2 and a much faster FPU. Just because of this the 604 nearly as fast as the 740 with L2 cache also known as the 750 (G3). Motorola's rating: 418 MIPS@233MHz for the 604 and 488 MIPS@266MHz for the G3. This is 27% speed increase at +33MHz: However, good 604 optimizing using all three IUs makes this even less.
It sounds like a whole lot when you write down the percentual part in 060s. However, CPU intensive utilities and games will have only a 30% speed increase. Non-CPU intensive programs are usually very fast and the 30% is not even useful in that case.
There's no reason to compare a G3 with a G4. The G4 is miles ahead of everything else and is therefore simply a different story.
Oh yes, and it is naive to compare a SysSpeed rating with benchmarks taken from Motorola/IBM. It is a completely different benchmark. How about comparing Sysinfo with SysSpeed? "Gosh, my Amiga is much slower with Sysinfo!" DOH! |
|
Yet another PCI-board for Amiga : Comment 9 of 14 | ANN.lu |
Posted by Sharwin & Rakesh on 24-Sep-2000 22:00 GMT | Sorry for the little mistake: 488 versus 418 MIPS is not a 27% speed increase but 17%! |
|
Yet another PCI-board for Amiga : Comment 10 of 14 | ANN.lu |
Posted by Amifan on 24-Sep-2000 22:00 GMT | In reply to Comment 8 (Sharwin & Rakesh Raghoebardayal): doh....sysspeed calculates drystone2.1 MIPS and flops...so you can compare them with the ones on the IBM site...... |
|
Yet another PCI-board for Amiga : Comment 11 of 14 | ANN.lu |
Posted by Anonymous on 24-Sep-2000 22:00 GMT | I can tell you one thing for sure in comparison of the 604e 233mnz
and an 060 in the real world. I have rendered the same scene several
times using the 060, then using the 604e. I have found that the 604e
will render the same scene at least 16 times faster than the 060.
Not to mention that it will encode an mp3 much faster than real time.
The 604e is going to be fast enough for many people. It's a whole
different world, and I love it. :) Of course many people just want
to go faster and faster, but hey a 604e isn't a bad place to be. |
|
Yet another PCI-board for Amiga : Comment 12 of 14 | ANN.lu |
Posted by Electric Head on 24-Sep-2000 22:00 GMT | Okay, so now we have three different versions of a PCI interface
from different companys. This would be great in a bigger market
maybe, but in the amiga world wouldnt a single combined effort
be more effective? Theoretically each company here will have a
smaller slice of sales.. effectively splitting the market up into
bits.
... or am I looking at this the wrong way?
And btw, I'm not 'moaning', I think its great news the classic
finally gets pci. I plan on buying one of the busboards myself :o) |
|
Yet another PCI-board for Amiga : Comment 13 of 14 | ANN.lu |
Posted by Kresimir Rogic on 26-Sep-2000 22:00 GMT | About time this card becaming available to the public. If I remember correctly, didn't Phase 5 use the G-Rex to demo a PCI Permedia 2 Card at WOA97 or WOA98, because they didn't have any Cybervision PPC's ready, even at prototype stage? |
|
Yet another PCI-board for Amiga : Comment 14 of 14 | ANN.lu |
Posted by Sjoerd de Vries on 01-Oct-2000 22:00 GMT | In reply to Comment 13 (Kresimir Rogic): The GRex you mentioned was/is the PCI Permedie2 card of phase5 for the MAC. |
|
Anonymous, there are 14 items in your selection |
|
- User Menu
-
- About ANN archives
- The ANN archives is powered by #AmigaZeux. It was updated daily (news last: 22-Oct-2004; comments last: 18-May-2005).
ANN.lu was created, previously owned and maintained by Christian Kemp, www.ckemp.com.
- Contribute
- Not possible at this time!
- Search ANN archives
- Advanced search
- Hosting
- ANN.lu was hosted by Dreamhost. Sign up through this link, mention "ckemp" as referrer and he will get a 10% commission on any account you purchase.
Please show your appreciation for any past, present and future work on ANN.lu by making a contribution via PayPal.
|