27-Apr-2024 03:40 GMT.		 UNDER CONSTRUCTION
Anonymous, there are 50 items in your selection
[Web] Forbes article on next gen IBM powerPCANN.lu
Posted on 14-Oct-2002 03:03 GMT by Christophe Decanini50 comments
View flat
View list
Here it is.
Forbes article on next gen IBM powerPC : Comment 1 of 50ANN.lu
Posted by Pav on 14-Oct-2002 01:53 GMT
Nice the hear but as the article states it's still 12+ months away... I wonder if Apple's able to hold off with 1.25GHz G4's till then?
Forbes article on next gen IBM powerPC : Comment 2 of 50ANN.lu
Posted by Bill "tekmage" Borsari on 14-Oct-2002 02:58 GMT
In reply to Comment 1 (Pav):
I guess they will just keep adding more :0
Forbes article on next gen IBM powerPC : Comment 3 of 50ANN.lu
Posted by coldfire on 14-Oct-2002 04:59 GMT
In reply to Comment 2 (Bill "tekmage" Borsari):
hehehehe...I can see it now....Quad G4 Mac!
coldfire
Forbes article on next gen IBM powerPC : Comment 4 of 50ANN.lu
Posted by Per on 14-Oct-2002 05:00 GMT
In reply to Comment 1 (Pav):
Ofcourse they will. They'll just keep on using their "All PPC-processors are twice as fast as the top-of-the-line Intel/AMD processors, no matter what" and that's it. No Apple-sheep will dare refute it.
Forbes article on next gen IBM powerPC : Comment 5 of 50ANN.lu
Posted by anarchic_teapot on 14-Oct-2002 07:16 GMT
In reply to Comment 4 (Per):
"They'll just keep on using their "All PPC-processors are twice as fast as the top-of-the-line Intel/AMD processors, no matter what" and that's it. No Apple-sheep will dare refute it."
Possibly. However, if they had a really fast, non-bloat OS on it, things might be different(apart from the pure number-crunching maybe, but seriously how many ordinary consumer programs need this?)
:-)
Forbes article on next gen IBM powerPC : Comment 6 of 50ANN.lu
Posted by Don Cox on 14-Oct-2002 07:22 GMT
In reply to Comment 5 (anarchic_teapot):
"Possibly. However, if they had a really fast, non-bloat OS on it, things might be different(apart from the
pure number-crunching maybe, but seriously how many ordinary consumer programs need this?)
:-)"
Mainly, anything to do with audio, which Apple are pushing as a major
use for Macs.
Forbes article on next gen IBM powerPC : Comment 7 of 50ANN.lu
Posted by Johan Rönnblom on 14-Oct-2002 08:14 GMT
It's sad to see how the clueless journalist can't fathom that speed !=
MHz, and in an attempt to satisfy both this technical cluelessness and
to distribute the IBM marketing spin, invents a new measurement,
"workload capability", somehow different from "speed".
Forbes article on next gen IBM powerPC : Comment 8 of 50ANN.lu
Posted by Lando on 14-Oct-2002 09:38 GMT
In reply to Comment 6 (Don Cox):
>Mainly, anything to do with audio, which Apple are pushing as a major
>use for Macs.
I think they need to concentrate on speeding up OS X.
It runs like a dog with three legs on my iMac (450Mhz G3, 512MB ram) - barely usable at times. AmigaOS, even on an 060, is a positive speed demon in comparison. It looks very pretty too but you can't change it's appearance (apart from your desktop wallpaper and colour scheme) which I don't like either.
Forbes article on next gen IBM powerPC : Comment 9 of 50ANN.lu
Posted by sutro on 14-Oct-2002 09:40 GMT
In reply to Comment 5 (anarchic_teapot):
So, MacOSX is a bloat and WinXP isn't ? Get a life !
Forbes article on next gen IBM powerPC : Comment 10 of 50ANN.lu
Posted by Marcus Sundman on 14-Oct-2002 09:46 GMT
In reply to Comment 9 (sutro):
> So, MacOSX is a bloat and WinXP isn't ?
Who said WinXP isn't bloated??
Forbes article on next gen IBM powerPC : Comment 11 of 50ANN.lu
Posted by Anonymous on 14-Oct-2002 10:10 GMT
This is a very bad thing for the Amiga community. If Apple switch to the new IBM chip, it's unlikely Moto or IBM will continue producing G3s or G4s for Amigas. The Amiga will be forced to switch to this new chip to remain on the PPC track, which will, although not that difficult (comparable to switching from 486 to Pentium I imagine), will be time consuming. Especially from a hardware point, as the new chip is VERY unlikely to use the same old G4 sockets.
Or maybe it's an excuse to get cracking on the AmigaTwo!
Forbes article on next gen IBM powerPC : Comment 12 of 50ANN.lu
Posted by cheesegrate on 14-Oct-2002 10:15 GMT
In reply to Comment 10 (Marcus Sundman):
winxp runs pretty fast on a 300mhz p2 with 64mb or ram..
osx is a dog in comparison on the the g3 macs..
even a 400mhz g4 is not the best..
Forbes article on next gen IBM powerPC : Comment 13 of 50ANN.lu
Posted by Freddy on 14-Oct-2002 10:28 GMT
In reply to Comment 12 (cheesegrate):
Oh come on... You must have a weird definition of fast if you think XP is quick to respond on a 64MB 300MHz P2. There's times at work where I feel like smashing the 256MB 1GHz P3 running XP I use.
Forbes article on next gen IBM powerPC : Comment 14 of 50ANN.lu
Posted by Freddy on 14-Oct-2002 10:32 GMT
In reply to Comment 9 (sutro):
No one said that XP wasn't bloated... just that OSX isn't far off.
Forbes article on next gen IBM powerPC : Comment 15 of 50ANN.lu
Posted by sutro on 14-Oct-2002 11:08 GMT
In reply to Comment 14 (Freddy):
OK, everything is bloated when compared with AmigaOS. That doesn't make OSX a bad OS. In fact I could use it over WinNT any day. It is clean, open, extensible and ready to be optimised. BeOS would have been an even better option had it stayed alived and QNX could be really good if they also put their attention to UI.
Forbes article on next gen IBM powerPC : Comment 16 of 50ANN.lu
Posted by Anonymous on 14-Oct-2002 11:10 GMT
In reply to Comment 11 (Anonymous):
No it is not. Apple is not only one who uses PPC-chips. Most of the PPC chips are used in embed systems so there sure will be lot's of demad for G3 and G4 too.
Forbes article on next gen IBM powerPC : Comment 17 of 50ANN.lu
Posted by Anonymous on 14-Oct-2002 11:45 GMT
Try checking out this site http://www.yellowtab.com/
They are making BeOS next generation :)
Forbes article on next gen IBM powerPC : Comment 18 of 50ANN.lu
Posted by priest on 14-Oct-2002 11:46 GMT
In reply to Comment 11 (Anonymous):
I do not see any problem, this is very good news to the community.
As far as I know IBM PPC970 will be fully SW compatible with G3 and G4.
IIRC, almost every PPC processor has it's own kind of socket or at least different pin arrangements (even different G4s have different pin arrangements), so you need to re-design the motherboard between G3, G4 and G5 anyway, unless you have the CPU on a slot like in pegasos and AmigaOne-XE.
When PPC970 ships later in 2003, there is still the need for low end CPUs, like 600Mhz G3, but I'm pretty sure the faster G4 (motorola chip) cards for pegasos and/or AmigaOne-XE are available at that time anyway and almost no-one will want 600Mhz G3 any more.
PPC970 can mean that it or G4 (or G5) chips are more easily available because there is more high end PPC ships on the market in general.
Forbes article on next gen IBM powerPC : Comment 19 of 50ANN.lu
Posted by priest on 14-Oct-2002 11:48 GMT
In reply to Comment 18 (priest):
I meant: ...and almost no Amigan will want 600Mhz G3 any more.
...
Forbes article on next gen IBM powerPC : Comment 20 of 50ANN.lu
Posted by coldfire on 14-Oct-2002 12:28 GMT
In reply to Comment 19 (priest):
Maybe not then...but I'd sure as hell like one now. :)
coldfire
Forbes article on next gen IBM powerPC : Comment 21 of 50ANN.lu
Posted by priest on 14-Oct-2002 12:30 GMT
In reply to Comment 20 (coldfire):
;)
Forbes article on next gen IBM powerPC : Comment 22 of 50ANN.lu
Posted by Don Cox on 14-Oct-2002 13:37 GMT
In reply to Comment 11 (Anonymous):
"This is a very bad thing for the Amiga community. If Apple switch to the new IBM chip, it's unlikely Moto or IBM will continue producing G3s or G4s for Amigas. The Amiga will be forced to switch to this new chip "
You mean you actually _want_ the Amigas to be slow? The major problem of the AmigaOne, which will make it unsaleable to any normal computer buyer, is the slow processor. The sooner Amigas use this new chip the better - ideally, a couple of months before the Mac.
Even then, there will probably be faster processors on the market from AMD and Intel by the time IBM actually release these 64-bit PPC chips.
Forbes article on next gen IBM powerPC : Comment 23 of 50ANN.lu
Posted by Alkemyst on 14-Oct-2002 13:48 GMT
In reply to Comment 22 (Don Cox):
"The major problem of the AmigaOne, which will make it unsaleable to any normal computer buyer, is the slow processor."
we are not normal.
normal ppl use PC or Macs & the Aone will be fast to amiga ppl & thats all its aimed at atm. including Aos4.
& that has been stated many times, the first round of AmigaOs PPC & PPC HW if for Amiga PPL.
Forbes article on next gen IBM powerPC : Comment 24 of 50ANN.lu
Posted by Don Cox on 14-Oct-2002 16:17 GMT
In reply to Comment 23 (Alkemyst):
"We are not normal."
Hey, you noticed.
My point is that it will only sell to extreme Amiga enthusiasts with
plenty of money. There are not very many of those - not enough to
create a market big enough for software authors.
Forbes article on next gen IBM powerPC : Comment 25 of 50ANN.lu
Posted by Alkemyst on 14-Oct-2002 17:17 GMT
In reply to Comment 24 (Don Cox):
yes, but we look at this round of PPC OS & HW as the building blocks for the OS HW that will be aimed at more than just the amiga ppl.
Forbes article on next gen IBM powerPC : Comment 26 of 50ANN.lu
Posted by hgm on 14-Oct-2002 18:05 GMT
In reply to Comment 23 (Alkemyst):
<that has been stated many times, the first round of AmigaOs PPC & PPC HW if for Amiga PPL.>
It is hard talking to deaf people.
regards
Forbes article on next gen IBM powerPC : Comment 27 of 50ANN.lu
Posted by strobe on 14-Oct-2002 20:40 GMT
Cripes, maybe I should write a Mac OS X FUD-buster.
There are many reasons OS X feels sluggish but apparently the only thing Amigans can say is "bloat bloat bloat!". I mean crap, if you don't want bloat let's get rid of these damned colors and go back to B+W. It's clear you don't have a clue when all you can do is say "bloat". Next you'll be blaming the kernel which is over 4MB in size.
Of the many reasons OS X feels sluggish to some people I'll list the obvious ones:
1) OS X has a completely new kernel and display server.
The kernel, xnu, may be loosely based on some previous technologies like Mach 3 and BSD but that doesn't tell the whole story. Darwin has a completely new driver architecture called IOKit which has an API which resembles the NeXT version but was rewritten for a C++ language subset. Furthermore Apple is using a lot of new Mach3 technologies like real time thread scheduling and IIRC a new pager. Apple then put ALL these technologies into a single binary so message passing takes a mere function call. Apple wanted a flexible foundation to last the next decade but they didn't have the time to optimize it to a great degree. It isn't as fast as Linux or BSD as of last time I tested it, but it has a much better designed driver architecture so making improvements from here on out will be far easier than other monolithic designs. IOKit is also very handy for power management and is what makes things like instant wake possible. You might call this bloat, but it can potentially be very fast indeed.
The display server is completely new and the first one of its kind which actually WORKS. By that I mean it has real double-buffered 32bit RGBA LAYERS instead of regions. It is like comparing MS Paint with Photoshop. Every pixel in a region-based graphics library belongs to one proces or another, but in Quartz every pixel is the result of blending all the layers. This makes it FAR more computationally difficult to do things like moving a window, and until very recently you had to do this ALL with the CPU. Somebody here complained it ran like crap on his G3, that is likely due to the fact that Quartz is still blending layers using the CPU (and it doesn't even have AltiVec to do so efficiently). When you use Quartz with a cheap AGP card like a GF2MX it *flies* since it uses the GPU to blend layers. The Amiga handled things like scrolling in a similar manner, but those chips were designed for the last generation of graphics libs which are region-based, not layer-based. Your G3 Mac is similarly designed, it isn't designed for Quartz. In Quartz you don't see the rendering process since that is done offscreen. Furthermore the windows are borderless which is really easy on the eyes. These are just two advantages of Quartz and more will become apparent as programmers get used to it. You might call this bloat, but I call it forward-thinking.
2) The ABI isn't standard.
Apple made an error in judgement in my opinion here. Apple changed from CFM/PEF to dyld/Mach-O. The problem here is the Mach-O ABI is ARCANE and didn't allow for the PowerPC-standard ABI to be used. What you end up with is fetching and storing registers for arcane reasons which slows everything up. Somebody estimated this slowed down the system by 30%. Why did they do this? They didn't want to develop a new linker which they estimated saved them 6mo of development time. I don't agree with that but whatever, Apple could port their CFM linker at any time and I suggest you tell Apple to do it.
3) Many Carbon apps are using deprecated APIs.
Well written Carbon apps are actually faster in OS X than Cocoa apps. Cocoa in OS X is actually implemented as a higher-level language to Carbon/Quartz/POSIX. This is how it should be since Carbon does work well as a lower level language for things like widgets and Carbon events. However many developers are still using deprecated Toolbox APIs which are there for compatibility reasons only. The primary problem is apps using the Event Manager which uses polling (a piss-poor decision made by Apple a looong time ago which I don't want to get into right now). Carbon events is the official replacement of the Event manager and is actually easier to develop for. This is a big problem because polling uses the CPU when it doesn't need to, and it doesn't get the CPU when it ought to. Yup, it's that crappy, that's polling for you. However there is no excuse for using this old API since Carbon events are available and easier to use. Running the "top" command will point out which apps are polling, like Explorer and Mozilla. You might call this bloat, but it's actually stupid scheduling being used by lazy 3rd party programmers.
All this being said the #1 problem I have with OS X isn't the speed and responsiveness of the apps, but rather the schizophrenic nature of the interface. The human <-> computer bottleneck is still the most limiting factor. Fortunately I'm still happy since I use nearly 100% Carbon apps which are designed to complement each other. This isn't a bloat issue either.
So stop saying "bloat" you eeediots!
Forbes article on next gen IBM powerPC : Comment 28 of 50ANN.lu
Posted by strobe on 14-Oct-2002 20:50 GMT
One more thing...
You can FORGET about using the PPC 970 with your Pegasos or AmigaOne. Even if the northbridge chip is on a CPU module (like it is on new world macs) you would still be stuck with a slow bus. Cripes even the G4 is currently being held back.
The new macs with DDR might have a different situation since the northbridge chip is on the CPU module. Not that anybody here would care.
Anyway take a few deep breaths and realize these boards are still crappy even in comparison to Apple's portables.
Forbes article on next gen IBM powerPC : Comment 29 of 50ANN.lu
Message removed by CK for violation of ANN's posting rules.
Specific reason from moderator: Trolling
Forbes article on next gen IBM powerPC : Comment 30 of 50ANN.lu
Posted by david lowenstein on 14-Oct-2002 22:29 GMT
Amiga could have been better than any game console or even the Mac, now IBM shut the door from Amiga forever. Now Amiga will never catch up with Apple or the Playstation. Let us bewail the terrible fate. (The only computer that is better than the Mac or the Playstation is only in fantasy.)
Forbes article on next gen IBM powerPC : Comment 31 of 50ANN.lu
Posted by Anonymous on 14-Oct-2002 22:41 GMT
In reply to Comment 28 (strobe):
>You can FORGET about using the PPC 970 with your Pegasos or AmigaOne.
We'll see.
>Cripes even the G4 is currently being held back.
Held back by what?
Stupid in-order execution? The MAI chipset surely doesn't held back G4
since G4 only supports 133MHz SDR bus speed anyway.
>The new macs with DDR might have a different situation since the northbridge
>chip is on the CPU module. Not that anybody here would care.
No one cares because you are spreading misinformation, shame on you! Take a look
at http://www.xlr8yourmac.com/systems/g4_mirrored_drive_doors/G4_MDD_CPU_Module.html
and tell me where to find the northbridge chip, smartass!
>Anyway take a few deep breaths and realize these boards are still crappy
>even in comparison to Apple's portables.
LOL!
What's the name of the company that helds back the "already bandwidth choked"
G4 with a crappy 100MHz bus - just because blind Mac followers buy any crap
and be happy to do so (eMac, iMac)?
Oh yes, even the iBook runs at only 100MHz although the CPU supports 150 or
even up to 200MHz FSB. W00t!
Forbes article on next gen IBM powerPC : Comment 32 of 50ANN.lu
Posted by priest on 15-Oct-2002 02:27 GMT
In reply to Comment 29 (Comment man):
LOL!
Forbes article on next gen IBM powerPC : Comment 33 of 50ANN.lu
Posted by priest on 15-Oct-2002 03:23 GMT
Last friday on a local Mac/PC shop, the dealer tried to sell me a 500Mhz G3 iBook... he said that even video editing works fluently on it ... I had to doubt it a little, even basic OS X behaviour seemed a bit sluggish on it...
No deal for €1400.
Forbes article on next gen IBM powerPC : Comment 34 of 50ANN.lu
Posted by Ole-Egil on 15-Oct-2002 03:46 GMT
In reply to Comment 33 (priest):
Well, MacOS X _DOES_ use neat graphical effects which kind of uses CPU time. Whether this is bloat or not depends on if you want neat graphical effects, but I for one would rather have a few hundred MIPS available to video editing than a pretty looking docking panel with zoom ;-)
I hear X.2 runs better on G3, though...
Forbes article on next gen IBM powerPC : Comment 35 of 50ANN.lu
Posted by cheesegrate on 15-Oct-2002 05:20 GMT
In reply to Comment 29 (Comment man):
oh comment man i just relized you flame mos and amiga one just to cause a reaction..
Forbes article on next gen IBM powerPC : Comment 36 of 50ANN.lu
Message removed by CK for violation of ANN's posting rules.
Specific reason from moderator: Trolling
Forbes article on next gen IBM powerPC : Comment 37 of 50ANN.lu
Posted by Don Cox on 15-Oct-2002 08:56 GMT
In reply to Comment 29 (Comment man):
"When will mickey mouse companys like Eyetech learn to "GET WITH THE
TIMES" "
I think Eyetech is doing the very best they can in the circumstances.
Nobody has the capital to develop special Amiga hardware, and the OS 4
authors are determined to write only for PPC.
Given those constraints, the G3 AmigaOne is the best anyone could do.
That doesn't mean it will sell in large numbers. That would only
happen if PPCs were substantially faster than off-the-shelf
processors.
Forbes article on next gen IBM powerPC : Comment 38 of 50ANN.lu
Posted by Alkis Tsapanidis on 15-Oct-2002 15:34 GMT
In reply to Comment 37 (Don Cox):
Given those constraints, the G3 AmigaOne is the best anyone could do.
--
It's not, and something better has already been done:)
Forbes article on next gen IBM powerPC : Comment 39 of 50ANN.lu
Posted by Leki on 15-Oct-2002 16:05 GMT
In reply to Comment 38 (Alkis Tsapanidis):
must be the AmigaOneXE your talking about
Forbes article on next gen IBM powerPC : Comment 40 of 50ANN.lu
Posted by Alkemyst on 15-Oct-2002 16:58 GMT
In reply to Comment 39 (Leki):
DaMN RiGHT :)
Forbes article on next gen IBM powerPC : Comment 41 of 50ANN.lu
Posted by strobe on 15-Oct-2002 18:18 GMT
http://www.apple.com/powerbook/specs.html
133Mhz FSB
4x AGP
1MB L3
As for the G4 being held back I didn't mean to imply that it was the fault of the Pegasos or AmigaOne, it's an inherent problem with the G4's memory bus. However it IS being held back by that limitation and the inherent limitation of these two boards would severely limit the PPC 970's performance (assuming you could even get that combination to work) which was the whole point.
As for the north bridge chip I assumed, possibly wrongly, that all new world macs have it on the CPU module. It is definitely true with respect to PowerBooks but perhaps not with the new desktop macs. Anyway my original point stands.
Forbes article on next gen IBM powerPC : Comment 42 of 50ANN.lu
Posted by anarchic_teapot on 16-Oct-2002 04:58 GMT
In reply to Comment 30 (david lowenstein):
"Amiga could have been better than any game console or even the Mac, now IBM shut the door from Amiga forever."
You do get them, don't you? Since when is producing a new generation processor, compatible with previous CPUs but running faster and with SIMD (alright, Altivec, but that's not an IBM word) betrayal?
Tsst.If IBM were anti-Amiga they wouldn't be lending one of their employees to the AmigaOne project. Also, it's a top-end processor and is going to be expensive for a year or two, which gives all the hardware designers plenty of time to produce new mobos and upgrade solutions for existing boards.
Forbes article on next gen IBM powerPC : Comment 43 of 50ANN.lu
Posted by anarchic_teapot on 16-Oct-2002 05:09 GMT
In reply to Comment 38 (Alkis Tsapanidis):
>> Given those constraints, the G3 AmigaOne is the best anyone could do.
> It's not, and something better has already been done:)
Alkis, you naughty troll, you! ;-)
Technically you're right, since both the XE and Pegasos can support G4 processors at least (looks like the G5 may be going the way of the 68050, if Apple have decided not to use them). OTOH everything is supposed to be shipping with G3 to begin with.
What looks promising to me is that if Apple is going to be switching to a new CPU, then G4s may well be easier to get hold of next year.
Forbes article on next gen IBM powerPC : Comment 44 of 50ANN.lu
Posted by priest on 16-Oct-2002 06:46 GMT
In reply to Comment 38 (Alkis Tsapanidis):
Still... not even G3 is available for the average Amiga user.
Forbes article on next gen IBM powerPC : Comment 45 of 50ANN.lu
Posted by Anonymous on 16-Oct-2002 14:18 GMT
In reply to Comment 41 (strobe):
>As for the G4 being held back I didn't mean to imply that it was the fault of
>the Pegasos or AmigaOne, it's an inherent problem with the G4's memory bus.
The MPX/Maxbus should have been a 128Bit bus according to the original plans.
>However it IS being held back by that limitation and the inherent limitation
>of these two boards would severely limit the PPC 970's performance (assuming
>you could even get that combination to work) which was the whole point.
Right. Still, the Northbridge on the CPU card would change nothing.
Come, Pegasos2! :-)
Forbes article on next gen IBM powerPC : Comment 46 of 50ANN.lu
Posted by strobe on 16-Oct-2002 17:41 GMT
From MacCenteral:
"Microprocessor Report editor in chief Peter Glaskowsky said that the PowerPC 970's memory architecture is proprietary. One of the hurdles that Apple would have to jump to use IBM's latest processor would be to design a memory controller for the part."
Another interesting tidbit is at 1.2Ghz it uses less power than the G4 at 1Ghz despite having more transistors; 19 vs 21.3 respectively. This might be due to the 0.13 micron process but I wouldn't know.
Forbes article on next gen IBM powerPC : Comment 47 of 50ANN.lu
Posted by david on 16-Oct-2002 19:08 GMT
In reply to Comment 42 (anarchic_teapot):
so is there still a faint glimmer of hope?
Forbes article on next gen IBM powerPC : Comment 48 of 50ANN.lu
Posted by david on 16-Oct-2002 19:11 GMT
I just live th a fantasy world where Sony and Apple get their butts kicked!
Forbes article on next gen IBM powerPC : Comment 49 of 50ANN.lu
Posted by Hammer(Same entity as in amiga.org's forum) on 17-Oct-2002 00:56 GMT
In reply to Comment 4 (Per):
>Ofcourse they will. They'll just keep on using their "All PPC-processors are >twice as fast as the top-of-the-line Intel/AMD processors, no matter what"
That’s a very centric view. Independent test results was not disclosed.
Refer http://www.geocities.com/sw_perf/ for an example programs to test with.
Refer to http://rss.com.com/2100-1001-961862.html?type=pt?=rss&tag=feed&subj=news
=== Quote =====================================================================
The PowerPC 970 chip, due next year, will run at 1.8GHz, nearly twice as fast as Big Blue's quickest existing PowerPC chip, the 1GHz 750FX.
===============================================================================
The comparison was with IBM's own PPC chips. Where did you get your results?
A 2X speed 1GHz 750FX(current) is not a quantum leap in computing speed (i.e. imagine o/c 2Ghz 750FX chip).
PPC 970(future) was only stated to be twice as fast when compared to the existing IBM's PPC flagchip.
In the case of top-of-the-line Intel processors. Did you forgotten Itanium II(Current)(and soon to be released Itanium III(future)?
In the case of top-of-the-line AMD processors. Did you forgotten AMD K8 (and soon to be released @ H2 2003)? Not factoring soon to released AMD Athlon XP (with Barton core)rumored have better IPC compared to existing T-Bred-B cores.
If you refer to http://www.geocities.com/sw_perf/
PPC solution was behind(compared to X86 solution) with the following benckmarks
1. Maya
2. Lightwave 7.5 (it takes a dual G4@1Ghz to beat a middle end Athlon 1.6Ghz)
3. Cinema4D (about equal @ a given Mhz i.e. non-XP enbled K7@1Ghz(100)G4@Ghz(98))
4. Mathematica
5. OGR
6. SETI
(Note that G4 did well in RC5, and the G4 was competitive with X86s in Photonshop 7 test)
Other benckmarking programs are;
1. Quake III
http://www.barefeats.com/graf31.html
2. OpenSSL
http://www.macosxhints.com/article.php?story=20020113045343563
Most real life benchmarks hasn’t been tested with the latest AMD Athlon XP~2.2Ghz(Tbread-B core) or Intel Pentium 4 ~2.8 Ghz (Northwood core).
(Never automatically equate AthlonPentium 4 in terms of application performance, both has different characteristics.)
Factors for evaluating a HW solution is
1. Main application being used
2. Price/Performance
3. Availability
4. Running legacy software.
>and that's it. No Apple-sheep will dare refute it.
Its Apple's call.
Please review the source materials again.
Forbes article on next gen IBM powerPC : Comment 50 of 50ANN.lu
Posted by Anonymous on 17-Oct-2002 09:01 GMT
In reply to Comment 49 (Hammer(Same entity as in amiga.org's forum)):
>A 2X speed 1GHz 750FX(current) is not a quantum leap in computing speed (i.e. imagine o/c 2Ghz 750FX chip).
>PPC 970(future) was only stated to be twice as fast when compared to the existing IBM's PPC flagchip.
Are you aware this was only refering to the clockspeed?
SPEC numbers are in the region of 3GHz+ PPCs.
>In the case of top-of-the-line Intel processors. Did you forgotten Itanium II(Current)(and soon to be released Itanium III(future)?
Not really a desktop chip imho - four times the amount of transistors of PPC970 which is quite comparable to Pentium4 in size (a bit smaller even).
>PPC solution was behind(compared to X86 solution) with the following benckmarks
Its current FPU is not very special, that's well-known.
The PPC970 will be good for 7.2 gigaflops in theory (actually 5.22 dhrystone-gflops).
>(Note that G4 did well in RC5, and the G4 was competitive with X86s in Photonshop 7 test)
PPC really comes into its own territory when Altivec is used - up to 8 flops/MHz :-)
While an optimal use is realized with asm code, I think there will be more support to
Anonymous, there are 50 items in your selection
Back to Top