Some vague MOT information

[lfrog]

Where the info comes from: A good friend of mine that is also a coworker happens to be good friends with a MOT person in the PPC division (how about vaque). Also my friend knows little about Macs and even less about hardware than I do (we both do software). I gave my friend questions on the G4/G5 to get answers to in December (still not answered). However he did recently get this information for me.

Information: MOT is concentrating resources on the performance of their existing lines (that would be the G4). Although this does not discount a G5 any time soon, this does reduce the probability.

Take this however you want, but those holding their breath for a G5 to appear soon may want to take another breath.

Also on the MOT rumor front, a rumor was going around MOT about becoming sole-source for AAPL (I do not think this is very likely).

[BRussell]

What you're saying is consistent with what Motorola people have said publicly in recent months.

About Mot being Apple's sole supplier - that's almost true now. The exception is the iBook. So is that a hint about the iBook going G4?

[Amorph]

Probably.

It also means that the "G5" is probably just(!) an upcoming version of the G4 on a vastly improved motherboard.

If two G4s can do as well as they do on Apple's current board, imagine what they could do with 333MHz DDR attached to onboard memory controllers and RapidIO-connected ASICs.

[bradbower]

No qualms with what you guys are saying here. Just as long as it's fast. Faster than Pentium 4s would be nice. :)

[Lemon Bon Bon]

Hmmm.

A warmed over G4 set up.

Still got the same single floating point unit.

Like a Ford engine in a Porsche body?

So. Whooppee. Come Macworld New York, after two years or so, we finally catch up with a PC Motherboard.

Nice. Great work, boys...

:(

Lemon Bon Bon

Currently preparing to take in another gulp of oxygen for another two year wait at this rate of going...yeesh.

<img src="graemlins/oyvey.gif" border="0" alt="[No]" />

[Amorph]

[quote]Originally posted by Lemon Bon Bon:
<strong>Hmmm.

A warmed over G4 set up.

Still got the same single floating point unit.</strong><hr></blockquote>

Not necessarily.

Past revisions to the G4 have introduced additional execution units; there's no reason why future revisions won't.

And don't forget AltiVec, which a) has remarkably capable FP abilities, and b) will no longer be totally starved for bandwidth on the new board, if it's anything like what I described.

A higher-bandwidth board will also make multiple processor setups more sensible, and then you have lots of FP units available. :)

[clonenode]

Has anyone considered that Apple might break away from the marketing of it's computers with the G4 or G5 name? This strategy has been around a while and seems like it's beginning to backfire. Look at allof us waiting for G5s, when one really doesn't exist.

I'm going to take this to a new Topic Thread.

[Lemon Bon Bon]

"Has anyone considered that Apple might break away from the marketing of it's computers with the G4 or G5 name? This strategy has been around a while and seems like it's beginning to backfire. Look at allof us waiting for G5s, when one really doesn't exist."

Backfire. Yeah. I'd say. If your processor supplier, IBM or Moto' can't pull their finger out and deliver a better than '500mhz' processor for two years...then...yeah. 'Backfire'.

Simple. They've been playing catch up and the mhz gap has gone from .5 gig to a who gig plus! And Intel merrily announce a 500 plus Motherboard.

G'yeesh.

There's no doubting that a 'G4' plonked into a new Motherboard (rapid io if all we've heard about it...) will do much better than the current ones. Maybe surprisingly so. But until I see what processor they plonk in it...I guess we won't know whether it's 'oh. another G4.' or 'Yay, twinkie cakes, G5.'

...but bar the extra integer head or altivec unit tweak...and...erm...the longer pipelines...this is still, basically the same G4 launched years ago at heart. Yeah...and I guess the 2meg level 3 cache aint bad.

...and if a 'G5' doesn't 'exist' you have to question what the hell Apple and Moto' are playing at. 'Hey guys...we're playing Russian Roulette with our Powermac sales...'

I suspect it does 'exist' but Apple and Moto' for whatever reason are holding back on releasing it...er...because, yay...they have found re: Apollo they can prolong the agony of another year wait for a G5...so...we pootle along in 2nd gear 'catch up' (still behind...) for another year.

I dunno. Maybe Apple likes taking it up the ass with Moto'...and in turn...we take it up the ass... Well, I don't. I don't buy a Powermac until I see the mythical G5 whether it's called that or 8500.

It doesn't matter if they call it a G5 or not. Either way, we'll know if they are palming off a G4 under an undeserving 'G5' monika.

Macworld New York. It's going to be interesting.

Somethings gonna come. But what.

My deep suspicion is that, with Apollo, somehow...with die shrinks they'll take the damn thing all the way to 2 gig. So G4++++++ (plus and another plus...and another 'plus'....) So, yeah, whooppedoo, we get a 2 gig G4 at Macworld Sanfran' 2003. That's my worry.

Big deal, if the fundamental chip doesn't change.

I hear what you're saying Amorph.

But the G4 has to take on board another two floating point units just to pull even with the current Athlon. And...pull another .5 gig gap out the fire just to be competitive.

More and more...it's looking likely that the 'true' G5 won't turn up until 'Itanium' or 'Sledgehammer' are already mainstream.

Shakes head.

To me. And I'm sure alot of analysts/apple watchers see the G5 as 'sticking it to the wintel'. Next July. 'So what'.

This July..?

Fighting talk. That's what I want to hear.

That's what I and alot of Powermac users want to buy.

...and looking at the Powermac sales...maybe they agree.

Lemon Bon Bon.

Yeah. I guess I'm whinging.

:(

[Masker]

Here is conlusive evidence that proves the Moto single source theory.

<a href="http://www.macmart.com/rammart/memory/memoryspecs/sheets/ibookg4.html" target="_blank">G4 iBook Titanium mention</a>

or it could be a typo. ;)

MSKR

[Bozo the Clown]

I see it requires "thine" DIMMs. :)

[THT]

<strong>Originally posted by Lemon Bon Bon:
...but bar the extra integer head or altivec unit tweak...and...erm...the longer pipelines...this is still, basically the same G4 launched years ago at heart. Yeah...and I guess the 2meg level 3 cache aint bad.</strong>

The difference between the 7450 based G4s and 7400 based G4s is greater than the difference between the 7400 based G4s and the 750 based G3s.

<strong>My deep suspicion is that, with Apollo, somehow...with die shrinks they'll take the damn thing all the way to 2 gig. So G4++++++ (plus and another plus...and another 'plus'....) So, yeah, whooppedoo, we get a 2 gig G4 at Macworld Sanfran' 2003.</strong>

This is fine as long as the appropriate tweaks are added. All that is required are steady improvements (increased memory bandwidth, more execution units, increased pipeline depth, multithreading, etc).

<strong>But the G4 has to take on board another two floating point units just to pull even with the current Athlon. And...pull another .5 gig gap out the fire just to be competitive.</strong>

The 7450 based G4 just needs another FPU unit, better memory performance, and better compilers to compete with the Athlon architecture. That's not that much compared to creating a brand new microarchitecture, which isn't really needed that much with the G4's architecture.

<strong>More and more...it's looking likely that the 'true' G5 won't turn up until 'Itanium' or 'Sledgehammer' are already mainstream.</strong>

99.9% of the information on the PPC "G5" is based on the fantasies of rumormongers. There is zero reliable information on the G5. None.

<strong>...and looking at the Powermac sales...maybe they agree.</strong>

A lot of this has to do with the prices charges for a PowerMac, but I would agree lessor performance is part of it.

[Junkyard Dawg]

Seems odd, that for so long Moto has been floundering and all this time there has been no work on a successor to the G4. I have a hard time swallowing such a scenario. What exactly has Moto been doing all this time, then? Sure there are a few G4 tweaks and revisions, but with all the power and might of Moto's Somerset design team, is that ALL that is produced? Only a few tweaks to the G4? Odd.

And what about the G5 currently being offered as an embedded chip? I forget the name and source, but isn't this chip currently on the market? If so, then it means Moto has a new architecture completed, but they need to add altivec units and such to prep it for desktops (Macs). How long would this take?

Furthermore, if Moto becomes Apple's sole supplier, and it seems that this will happen soon as the G3 is phased out, then how will Apple differentiate between consumer and pro macs? Will all Macs have G4s, and the Pro lines will simply have faster G4s? Perhaps this would work, but it seems like a better plan to have both G4s and G5s in Macs, to create a compelling reason to spend a tremendous amount of money on a "power"mac. Otherwise pro-hardware sales will suffer, as we see happening now with the iMac so close in performance to the Powermacs, and since the Powermacs are high margin sales, Apple sustain such a lineup for long. But a G5 would rectify the current lack of distinction in performance between pro and consumer macs and then some.

All the current info and state of affairs with Mac hardware suggests a "G5" PPC soon, as in, a new architecture.

1.8 GHz G4s would be fine, but let's be realistic, they are YEARS away. If there is no G5, then Mac hardware performance will continue to improve incrementally, but the problem is that we need a great leap in performance, not baby steps. Only a new PPC architecture can provide the sort of performance improvement that Apple needs to remain competitive, and to capture some of the new markets Apple has been targeting, such as 3d animation and special effects, and real time DV editing. Newsflash Apple: you cannot take these markets with 1.2 GHz G4s, when the competition wipes the floor with Macs at 3d animation!

Thus I believe a G5 will be forthcoming. If not, then Apple is in a precarious position that cannot be maintained indefinitely. Something will give, and it's going to be ugly for Apple.

[ 04-18-2002: Message edited by: Junkyard Dawg ]</p>

[admactanium]

motorola is not working on the g5. they had a project but it was killed by apple. the g5 should be sampling right about now if my sources are correct.

[Sybaritic]

[quote] motorola is not working on the g5. they had a project but it was killed by apple. the g5 should be sampling right about now if my sources are correct. <hr></blockquote>

Can you elaborate on the sampling remark?

Merci.

[admactanium]

i was told a couple months ago that the g5 would be going into production in the next 2 to 3 months. i guessed from that that it seems a intro date at macworld ny made sense.

i'm not trying to sound cloak-and-dagger. i am just concerned about the person who told me the information. you don't have to trust me, i'm just telling you what i know. but my friend knows these things without doubt.

[ 04-19-2002: Message edited by: admactanium ]</p>

[JBL]

Are you saying that the G5 was going to be going into production but Apple killed it, or are you saying that Motorola had a G5 program which Apple killed because Apple's own G5 is going into production?

[admactanium]

motorola had a project planned to be the next generation, it was killed before production. a chip that could potentially be the g5 is going in to production probably now or very soon, but it isn't being made by motorola but by ibm.

[ 04-19-2002: Message edited by: admactanium ]</p>

[phishy]

Maybe with the rise in competition (AMD) for building faster, cheaper, more powerful chips has made it too competitive, and/or too unprofitable for apple and moto to compete in the "Prosumer" market. Therefore apple will implement the following strategies:

1. Sell G4 imacs and ibooks for one more year or until unprofitable.
2. Phase out "prosumer" hardware
3. Pump out many must have moto embedded chip "digital hub" peripherals
4. Focus primary research, development and marketing efforts on the new, now completely processor independent hub of the "digital hub" version OSX 3 (or OSX 2???)

This strategy will finally take a major stab at the windows OS market share. Why not work with only Motorola when all you'll need are current potency g4's and great embedded chips?

[ 04-19-2002: Message edited by: phishy ]</p>

[Amorph]

Everyone read THT's post until you can recite it backwards from memory.

Whether a processor is called a "G5" or not doesn't matter. This is mostly because there are only murmurings of something that might be a "G5" that Apple could use (the 8500 ain't it, folks), and also because a name like "G5" is cooked up by marketing, not engineering. It has only nominal relevance to anything of significance.

What matters is how the computer performs. As THT points out, the G4 is not far from being a real contender. The better compiler is coming. The other two tweaks - (support for) faster memory and additional execution units - are widely rumored, and simply logical additions to the line.

Somerset has been hard at work on the G4. They obviously weren't kidding when they said it had a lot of life left in it. Is there anything wrong with this? No. What matters is performance. If Mot and Apple deliver that, it doesn't matter what Apple marketing decides to call the chip at the heart of the board.

[ 04-19-2002: Message edited by: Amorph ]</p>

[GardenOfEarthlyDelights]

[quote]Originally posted by Junkyard Dawg:
Seems odd, that for so long Moto has been floundering and all this time there has been no work on a successor to the G4. I have a hard time swallowing such a scenario. What exactly has Moto been doing all this time, then? <hr></blockquote>

That, unfortunately, is what happens in the real world. If a company is in as big of trouble as Motorola, you go to "survival" mode. First step is to cut extraneous activities. Then you put all your eggs in one basket. Eventually, you look for a way out, the light at the end of the tunnel.

In other words, you circle your wagons before trying to fight your way out.

I'm not saying that's the right way to do it-- this is just what companies (or people who lead companies) do.

Therefore, I could see Motorola's G4 efforts increasing as a percentage of resources within the company. (Think about that for a minute.) Motorola has to stop the bleeding. How much have they lost in the last few quarters? Something like GNP of Canada + their brother?

It's not a good scenario for us. If a "G5" is imminent (MWNY), it will be due to Apple's efforts, and not so much Mot. I'd love to see a "G5" this summer (who wouldn't?), but my I'm saving my bottle deposits for after San Francisco next year.

[SYN]

[quote] motorola had a project planned to be the next generation, it was killed before production. a chip that could potentially be the g5 is going in to production probably now or very soon, but it isn't being made by motorola but by ibm.<hr></blockquote>

this is definitely interesting, supposing it has any truth to it. Leveraging IBM's tech to manufacture the CPU would definitely be an advantage over relying on Moto, who's manufaturing process leaves much to be desired.

what was wrong with the next generation design moto was working with compared to the one that is supposedly going into production soon?

[Programmer]

[quote]Originally posted by SYN:
<strong>
this is definitely interesting, supposing it has any truth to it. Leveraging IBM's tech to manufacture the CPU would definitely be an advantage over relying on Moto, who's manufaturing process leaves much to be desired.

what was wrong with the next generation design moto was working with compared to the one that is supposedly going into production soon?</strong><hr></blockquote>

Assuming any of this is remotely true (and I'm not placing bets) then I see two possibilities:

1) IBM has a competing design that is superiour.
2) Apple bought the design work from Motorola and is having IBM complete it (if necessary) and produce it.

[stimuli]

This ties in with the posted rumour that the raycer cats are making a dedicated co-processor that does Altivec.

Of course rumour + rumour != truth, but it makes sense.

I for one would feel better if IBM was Apple's future Gx supplier. However, this puts 'G5' computers at a much more distant release date.

If Apple are going high end (Maya, Shake, etc) then clearly G4s are going to cut it except on hella fast busses in 4 way+ configs.

And given Apple's prices the price/performance ratio would put Intel/AMD's offering even further ahead.

[rickag]

[quote]Junkyard Dawg
"And what about the G5 currently being offered as an embedded chip? I forget the name and source, but isn't this chip currently on the market?"<hr></blockquote>

Unless something has changed since last October, the MPC8540 is not shipping and will not begin sampling until the 2nd half of 2002.

<a href="http://www.motorola.com/mediacenter/news/detail/0,1958,568_322_23,00.html" target="_blank">http://www.motorola.com/mediacenter/news/detail/0,1958,568_322_23,00.html</a>
[quote]"Samples of the MPC8540 are expected to be available in the second half of 2002."<hr></blockquote>

Me personally, I say so what. The MPC8540 is for embedded use.

I believe the G4 does have much room to grow.

The upcoming die shrink will allow a faster clock speed. Surely, Apple/Motorola will implement DDR capability. Surely, Apple/Motorola will beef up the fp performance(re: maybe an additional fp unit?).

Extending the pipelines in the core would allow additional clock increases. The current G4 has what 7 stages? Increasing to 9-10 should allow significant increases, no.

I just don't expect the more advanced tech to appear in a chip for Apple for a while(re: rapid I/O, switched fabric interconnects, etc.). And I still say so what.

A G4 clocked to 1.8GHz.,with DDR would absolutely rock. The question is will it happen.

[Ketracel]

I speculate :

G4 speed bump in July and no mobo revision (hmm maybe ATA100). Mac OS X 10.2

Jan 2003-IBM PPC, Multicore, Rapid I/O, CoreConnect, SIMD Engine. (Perhaps not named G5). Completely new mobo, Rapid I/O, PCI-X, DDR.

Motorola may only supply G4's for the ibook, imac products.

Ketracel

[mattyj]

Ok, what we have to think is what's happening to apple's pro market. At the moment machines aren't selling as much as they did a couple of quaters ago, because of the speculation that G5s are just around the corner. So apple should (hopefully are) be working its ass off to bring the "G5" out.

If they name a G4 the G5 as a marketing ploy, it still won't get them out of the deep, as pro users will know the difference. Unless new chips bring a massive performance boost, then apple's pro line is doomed. <img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" />

[phishy]

Apple may be finding a way out or being pushed out of the prosumer hardware market and shifting that focus and energy to what it is good at (and can do 100% in house):
Develop a great OS.

Hasn't Jobs been trying to do this all along? <img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" />

[Amorph]

The purpose of an OS, at the end of the day, is to control the hardware.

If Apple doesn't control the hardware, it'll be hard for them to write a killer OS.

[Outsider]

There could be a reason Apple went with the moniker "Velocity Engine" Instead of using Motorolas name "Altivec". If Apple decided to go with a vector unit designed by IBM that could make use of normal Altivec optimised code, then they could just append the name "Velocity Engine II" and the average user would be none the wiser.

[phishy]

[quote]Originally posted by Amorph:
<strong>The purpose of an OS, at the end of the day, is to control the hardware.

If Apple doesn't control the hardware, it'll be hard for them to write a killer OS.</strong><hr></blockquote>

Then how can an OS like Linix, FreeBSD, etc. be so awsome and work on most wintel machines. Isn't quality control more a questions of good drivers? ...Plus currently Apple doesn't control all of it's hardware (CD Burners etc.), and it can't seem to control the timely release of a really fast setup.

Sorry about being a little off topic Mr. Moderator Amorph, but what the hey.

:)

[ 04-19-2002: Message edited by: phishy ]</p>

[Lemon Bon Bon]

"Somerset has been hard at work on the G4. They obviously weren't kidding when they said it had a lot of life left in it. Is there anything wrong with this? No. What matters is performance. If Mot and Apple deliver that, it doesn't matter what Apple marketing decides to call the chip at the heart of the board."

Performance matters. Yes.

And I think a 1.2 gig G4 with DDR ram will be a lame attempt at £2,500 quid. Throw in a mobo revision and an extra floating point unit...and you'll be almost even for about half a year...and miles behind in price...before the Hammer lands in January. Apple to offer 1.6 gig G4 next January at this rate of going? I can't believe Apple won't have an architectural PPC revision answer to counter the Hammer.

It's clear that putting that right up against the 64 bit 'Sledgehammer' in workstation markets is suicide. Even a tweaked G4 in a revised mobo is going to get the kicking of its life against the 'Hammer'.

That's the disaster scenario for me.

I'm excited by the prospect of:

1. An IBM chip.
2. Mobo revision.
3. Raycer chipset.
4. It being the 'G5.

All the 'net voodoo whispers of things. What's that about no smoke without fire?

JD. Your post was bang on.

Apple have got to do something to really shake up their pro' hardware. They've got to take more responsibility for their 'guts'.

I find it hard to believe that Steve Jobs can 'take it' when he finds the Mac crawling to 1 gig. I'd imagine him, behind the scenes, pulling heads outta asses to get something done about it.

The idea of Moto doing the design and IBM fabbing sounds interesting. It would explain the confused smoke signals...re: 'G5'. The whole debate is tinged with 'weirdness.'

The next year is going to be very engaging.

Like the last. Who'd a thought we'd have a
£1,495 copy of Maya on the Mac...

Maybe 'G5' dreams can come true...

<img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" />

Lemon Bon Bon

[mattyj]

Did Maya come out on the macintosh platform just because OS X was made? I don't think that is the whole story. I think that Maya came out because the G5 will be targetted at the 3d market, I mean don't you think Mr Jobs would like to see his company Pixar to use 2Ghz macs instead of SGI stations? ome on people. :D

Also doesn't motorola's contract end with apple by MWNY? If so, we could see the "G5" produced by IBM, and a co-processor developed by Raycer which would yield the second "velocity Engine"

[xype]

[quote]Originally posted by phishy:
<strong>

Then how can an OS like Linix, FreeBSD, etc. be so awsome and work on most wintel machines. Isn't quality control more a questions of good drivers? ...Plus currently Apple doesn't control all of it's hardware (CD Burners etc.), and it can't seem to control the timely release of a really fast setup.
</strong><hr></blockquote>

a] open source OS's are not that awesome, they just fit the needs for a cheap unix-alike server/workstation. many "commercial" unix versions are much better engineered and also work better. linux is so popular because it is free. Apple can't abandon it's hardware business to give away MacOS X for free - they need to make a profit, you know.

b] the market of linux is not what Apple is targeting (apart from a few Linux geeks that can afford Apple machines). Apple targets consumers and the graphics/DTP/multimedia market - and can live as long as it keeps exactly like it is. making any change like allowing clones, starting to use x86 can kill it faster than you say "microsoft". linux, *BSD and the like are far away from competing on the desktop OS market.

c] quality control a question of good drivers? there is more than drivers to quality control. stabiliy, user experience, etc.

d] of course apple controls it's hardware. who do you think is deciding which CD-ROM drives get into a PowerMac? IBM?

[Brendon]

[quote]Originally posted by mattyj:
<strong> Also doesn't motorola's contract end with apple by MWNY? If so, we could see the "G5" produced by IBM, and a co-processor developed by Raycer which would yield the second "velocity Engine"</strong><hr></blockquote>

Actually you could see the Raycer guys doing some or all design work of Altivec 2. With Book-e out anyone could add a "unit" to an existing PPC. What makes more sense is to have the Raycer guys add instructions to Velocity Engine 2, or to add their own execution unit to the PPC G5 chip. Why have a dedicated co-processor that requires more wiring something else to cause supply constraint if not now maybe sometime on the future, when they could just add new capabilities to the PPC itself in the form of a new instruction unit or add new capabilities to the Velocity Engine.

[Brendon]

Double Post, Sorry 'bout that <img src="graemlins/embarrassed.gif" border="0" alt="[Embarrassed]" />

[ 04-21-2002: Message edited by: Brendon ]</p>

[Amorph]

[quote]Originally posted by Lemon Bon Bon:
<strong>

Performance matters. Yes.

And I think a 1.2 gig G4 with DDR ram will be a lame attempt at £2,500 quid. Throw in a mobo revision and an extra floating point unit...and you'll be almost even for about half a year...and miles behind in price...before the Hammer lands in January.</strong><hr></blockquote>

This is completely free speculation. For one thing, the rumors have tended toward 1.4 or 1.6 GHz this summer. For another, we'll see what the performance is when the board rolls out. Apple seems to be working on an unconventional board design, and there is a distinct possibility that it will more than catch Apple up. Or it might not. We don't know. Any guesses are not "way off" or "spot on." They're guesses.

[quote]<strong>Apple to offer 1.6 gig G4 next January at this rate of going? I can't believe Apple won't have an architectural PPC revision answer to counter the Hammer.</strong><hr></blockquote>

The PPC 7500, according to an admittedly less than reliable source, lands this winter, about in time to compete with the Hammer. To the extent that Apple will release anything that competes with Hammer, actually - they didn't sweat Intel's 64 bit architecture either. That's a different market than Apple's in.

[quote]<strong>
1. An IBM chip.
2. Mobo revision.
3. Raycer chipset.
4. It being the 'G5.

All the 'net voodoo whispers of things. What's that about no smoke without fire?</strong><hr></blockquote>

There's plenty of smoke without fire as far as rumors are concerned.

As for IBM, they seem to have gotten really good at making incredibly small, efficient G3s and gigantic CPUs destined for gigantic mainframes. They're also really good at announcing products many long months in advance of delivery. What, exactly, are you hoping for from them? They reached 1GHz with their G3 after Mot did with the G4, after all. Raycer chipset? Not gonna happen. The Raycer engineers still at Apple are apparently working on OpenGL. G5? Why? Nobody knows what it is! For all we know, it'll be a 32 bit G4 (upward compatible to 64 bits) reengineered to meet the Book E spec!

[quote]<strong>I find it hard to believe that Steve Jobs can 'take it' when he finds the Mac crawling to 1 gig. I'd imagine him, behind the scenes, pulling heads outta asses to get something done about it.</strong><hr></blockquote>

Probably. There's no evidence that Apple is sitting on its hands, though (nor is Mot - look at the rapid series of improvements to the G4 - or IBM), so he might not be freaking out (now, you can bet he was a couple of years ago...).

Besides, we're at dual-GHz. Multiple processors matter. And unlike Windows, the default consumer OS supports MP. And unlike Linux, the MP support doesn't suck eggs.

[ 04-21-2002: Message edited by: Amorph ]</p>

[Lemon Bon Bon]

"This is completely free speculation. For one thing, the rumors have tended toward 1.4 or 1.6 GHz this summer."

Well, let's hope so. But that's still catch up. With a revised mobo. Still catch up. The G4 was obviously meant to compete with a Pentium III. Where's the 'G5' to compete in general Apps and 3D rendering? I haven't seen a bench for Lightwave that justifies the £1000 plus premium for a dual gig Mac over a 2.1 XP athlon?

"For another, we'll see what the performance is when the board rolls out. Apple seems to be working on an unconventional board design, and there is a distinct possibility that it will more than catch Apple up."

I hope so...or the performance gap seen with the G4 at 500mhz for two years will seem like a picnic.

"Or it might not. We don't know. Any guesses are not "way off" or "spot on." They're guesses."

Guesses. Yes. As we all are. However, I guess you know I was pulling on my 'constipated conservative Kermit' hat for this guess...which was following Apple/Moto's track record these last two years. ie They're going to have to buck previous trends to counter the increasing performance 'gap'. A leap from 1 gig to 1.6 gig this Summer? That's 600mhz. More than they've achieved in the last 3 years?


"The PPC 7500, according to an admittedly less than reliable source, lands this winter, about in time to compete with the Hammer. To the extent that Apple will release anything that competes with Hammer, actually - they didn't sweat Intel's 64 bit architecture either. That's a different market than Apple's in."

The Itanium. Different market/price. Fair enough. But the 'Hammer'? AMD have a track record of compelling price/performance. The 7500 (which I always thought was the G5 before this 8500 thing got bandied about...) will have to be delivered to compete. Even a bumped 1.6 gig G4 is NOT going to compete...at least on price. Apple will charge £2,100 ish for it exc VAT. Athlon PCs will be sub £900 for that kind of performance come the summer.

I guess I'm hoping that Apple and Moto can take 'a hit' and drive the Apollo down into the imac range as fast as possible, leaving the 7500 for the Powermacs.

This, I feel, will leave us competitive again. For a while...

"There's plenty of smoke without fire as far as rumors are concerned.

As for IBM, they seem to have gotten really good at making incredibly small, efficient G3s and gigantic CPUs destined for gigantic mainframes. They're also really good at announcing products many long months in advance of delivery. What, exactly, are you hoping for from them? They reached 1GHz with their G3 after Mot did with the G4, after all. Raycer chipset? Not gonna happen. The Raycer engineers still at Apple are apparently working on OpenGL. G5? Why? Nobody knows what it is! For all we know, it'll be a 32 bit G4 (upward compatible to 64 bits) reengineered to meet the Book E spec!"

Fair point about IBM. But if Apple have 'dropped' moto from the Powermac range then I'm sure IBM's chip making capabilities should not be underestimated. Raycer working on Open GL? Makes sense, I guess. But I don't think we can rule out a 'special' chipset/ co processor or something. I think it's about time Apple had something extra. Y'know...to cover the premium we pay? The premium...which appears ridiculous at times. What we want, we don't always get. But...I remember the Amiga offering extra value Pcs couldn't many moons ago...and Commodore was never the size of IBM...or Apple.

A G4 engineered 'Book E'? Maybe... 'Half-Son of Yikes'? Correct me if I'm wrong. But all I see is a G3 boosted chip. Sure...it's had tweaks. But take away Altivec and what have you got?

To me...the mythical 'G5' is, what? A 'core' upgrade I'm waiting for. I'm not paying for a G3 mark 4. It appears lots of Pro users are waiting for this 'core' upgrade to.

"Probably. There's no evidence that Apple is sitting on its hands, though (nor is Mot - look at the rapid series of improvements to the G4 - or IBM), so he might not be freaking out (now, you can bet he was a couple of years ago...)"

Improvements, granted. But, if we're honest. We're still very much behind and need a 'new chip'. What ever it's called. In general apps, we're clearly not competing with sub £1000 PCs.

So. Where's the 'Power' in 'Powermac'?

Apollo's catch up. Force it into the consumer macs as quick as possible.


"Besides, we're at dual-GHz. Multiple processors matter. And unlike Windows, the default consumer OS supports MP. And unlike Linux, the MP support doesn't suck eggs."

It looks as if the Mac has got the better dual processor support. And, if you've had a Powermac 200mhz clone for four years, the dual 1 Gig mac seems fast. Very. But if you've got an Athlon XP 1.6 gig...with 1 gig of DDR ram and an ATI 8500...then you go...

Err. So what?

Perceptionally, performance doesn't seem different between the two...yet the 1.6 Xp can be had for silly money.

If Apple is charging these kinds of premiums then they should be offering much better performance for the money.

As always, Amorph, nice Jousting with you.

Lemon Bon Bon

;)

[timortis]

[quote]Originally posted by Amorph:
<strong>

To the extent that Apple will release anything that competes with Hammer, actually - they didn't sweat Intel's 64 bit architecture either. That's a different market than Apple's in </strong><hr></blockquote>

I'm not sure I agree with that. The first of the Hammer processors, Clawhammer, will be targeted directly at the high-performance desktop/workstation market. AMD will be phasing out Duron, and Athlon will be demoted to being the budget processor.

In other words Hammer will be directly competing with the PowerMac.

Today an Athlon XP 2100+ (1.73 Ghz) pretty much runs circles around a G4 at most tasks. Now consider the fact that the first Clawhammer to come out will be given a performance rating of 3400+ by AMD. That is a pretty big jump, expected around the same time as MW San Fransisco next year.

[Lemon Bon Bon]

"I'm not sure I agree with that. The first of the Hammer processors, Clawhammer, will be targeted directly at the high-performance desktop/workstation market. AMD will be phasing out Duron, and Athlon will be demoted to being the budget processor.

In other words Hammer will be directly competing with the PowerMac.

Today an Athlon XP 2100+ (1.73 Ghz) pretty much runs circles around a G4 at most tasks. Now consider the fact that the first Clawhammer to come out will be given a performance rating of 3400+ by AMD. That is a pretty big jump, expected around the same time as MW San Fransisco next year. "

Ouch!

'G5'...where are you?

Lemon Bon Bon :mad:

[Amorph]

Well, we'll see.

Apple doesn't need a new model of CPU so much as it needs a new board. The fact that the dual GHz machine is as fast as it is when both processors are sipping data from the same 133MHz SDR bus, running code compiled with the legendarily unoptimized gcc, is frankly incredible. A faster mobo means the same processors go much faster, and it means Apple can use more processors per machine. And, of course, newer chips and compiler optimizations are coming. It's not so hard to add another FPU.

If the 7500 rumor is right, and if it's what the Register says it will be, Apple have an answer to Hammer by sometime around Christmas. Close enough.

Macs do not have to beat x86 machines in every imaginable area in order to be competitive. They use a different architecture, and different approaches, so they'll always be better for some things and worse for others. If all you need is a number cruncher, you'll always be able to build a stripped-down PC for much less than a Mac, because Apple has always emphasized the "whole widget" and high-level utility. So your budget-conscious render farm might be all Athlon even if Apple rolls out a G5 PowerMac tomorrow, but your primary workstation might just be a Mac.

[Eskimo]

[quote]Originally posted by Amorph:
<strong>

If the 7500 rumor is right, and if it's what the Register says it will be, Apple have an answer to Hammer by sometime around Christmas. Close enough.

</strong><hr></blockquote>


The fact that enthusiasts are worried whether their products will compete with a future AMD processor and not Intel gives me a warm fuzzy feeling inside. Thanks for making my day James. :) I'll have to put this on my cube wall.

[Brendon]

[quote]Originally posted by Amorph:
<strong>If the 7500 rumor is right, and if it's what the Register says it will be, Apple have an answer to Hammer by sometime around Christmas. Close enough.
</strong><hr></blockquote>

Here is a link to that article if anyone wants to read it. <a href="http://www.theregister.co.uk/content/archive/24018.html" target="_blank">http://www.theregister.co.uk/content/archive/24018.html</a>

[Amorph]

[quote]Originally posted by Eskimo:
<strong>


The fact that enthusiasts are worried whether their products will compete with a future AMD processor and not Intel gives me a warm fuzzy feeling inside. Thanks for making my day James. :) I'll have to put this on my cube wall.</strong><hr></blockquote>

No prob, Eskimo. :)

Intel might have stolen the Alpha, but AMD got its engineers.

I'm betting on the engineers.

[ 04-21-2002: Message edited by: Amorph ]</p>

[sc_markt]

[quote]Originally posted by Amorph:
<strong>
If the 7500 rumor is right, and if it's what the Register says it will be, Apple have an answer to Hammer by sometime around Christmas. Close enough.
</strong><hr></blockquote>

Amorph,


This quote from the Register article
"the third addition to the G4 family, the 7500 - slated for volume production by this time next year"
which was posted by the Register on 11/02/2002 tells me we'll have to wait at least another year for it.

And that could mean that the powermacs to follow once the chip starts shipping could be announced at the next WWDC and then start to ship within a month from the announcement.

And I'm sure before that time comes, we'll start hearing about the next generation Hammer chip.

[Amorph]

[quote]Originally posted by sc_markt:
<strong>

Amorph,


This quote from the Register article
"the third addition to the G4 family, the 7500 - slated for volume production by this time next year"
which was posted by the Register on 11/02/2002 tells me we'll have to wait at least another year for it.</strong><hr></blockquote>

Thanks for that catch. I was going from my memory of the article, and I really should know better. ;)

All I can say is that: Given that PPC is already designed to scale easily to 64-bit, nothing anywhere says that the 7500 will be the first G4 to go there. Nothing says that Mot won't produce a 32-bit chip that can keep up in all ways except the few that require 64 bit pointers and integers. Certainly, nothing is preventing Apple from going MP across its professional line. PPC chips are inexpensive (relatively speaking) and they do MP well. And the OS they'll be running does MP better than any of its competitors (although it admittedly hasn't been tested for large numbers of processors).

The Hammer looks to be a real threat, but I think Intel's sweating it a lot more than Mot or Apple is, considering Intel's storied, years-long failure to produce a viable 64 bit platform. If Apple produced a screaming 32 bit workstation I don't think too many people would mind - especially if it had good FP performance (including 64 bit FP).

This should be an interesting year. But I can't imagine that Apple isn't confident, given the acquisitions they've been making. They must know they've got something hot in the pipeline.

[Jonathan]

[quote]Originally posted by Amorph:
<strong> But I can't imagine that Apple isn't confident, given the acquisitions they've been making. They must know they've got something hot in the pipeline.</strong><hr></blockquote>


Exactly. I can't seem to understand how or why (barring OBSCENE arrogance from SJ- could he believe the marketing BS?) Apple would be acquiring firms that would lead one to believe they're REALLY going after the low-end-workstation market- it just wouldn't make sense, if there isn't some good piece of hardware coming that will run these apps at a competitive speed.

a mystery indeed. I can only hope that they deliver this time. I don't care if its a G4, G5, GYourMomma. If it's got good SPEC and hammers the competition (har har har) then, well, wow.

[Lemon Bon Bon]

"Well, we'll see.
Apple doesn't need a new model of CPU so much as it needs a new board. The fact that the dual GHz machine is as fast as it is when both processors are sipping data from the same 133MHz SDR bus, running code compiled with the legendarily unoptimized gcc, is frankly incredible."

Yes, I guess 'we'll see'. I think I'm one of those 'Mac watchers' who are dying to see the G4/G5 'debacle' resolved one way or another. It's like when we were 'waiting' for the 'flat panel' iMac. It came almost year after we expected it.

The 'G5' wait is becoming more infamous. It seems that the G4, in its numerous guises is going to take us until well into next year...

The irony will be that the G4 will almost be a G5 by the time it ends its 'reign' as the 'Powermac'..?

"A faster mobo means the same processors go much faster, and it means Apple can use more processors per machine. And, of course, newer chips and compiler optimizations are coming. It's not so hard to add another FPU."

Well, an exta FPU would be nice! But I'd like to see a couple in there. I think the changes you, the Register etc are all talking about will do...for New York Macworld. Well, it'll be interesting to see the 'actual' performance when the DDR mobo arrives. But, having seen the PC DDR benchmarks over PC133...we could be in for a disappointment...

"If the 7500 rumor is right, and if it's what the Register says it will be, Apple have an answer to Hammer by sometime around Christmas. Close enough."

Well, it's not essential the 'G5'/G4 be 64 bit. It will have a 32 bit version by all accounts.

The 32 bit Playstation 'hammered' the '64 bit' Jaguar console. Performance, like you say is more important than 'showy' marketing numbers...

"Macs do not have to beat x86 machines in every imaginable area in order to be competitive."

Of course not. But hammer them in most for the kind of money we pay? I think so :)

"They use a different architecture, and different approaches, so they'll always be better for some things and worse for others. If all you need is a number cruncher, you'll always be able to build a stripped-down PC for much less than a Mac, because Apple has always emphasized the "whole widget" and high-level utility."

I can see what you're getting at. But I think the reason most of all...that Pro sales are suffering is that most people resent the idea of paying almost £3000 for kit that is clearly behind in performance for MOST tasks.


"So your budget-conscious render farm might be all Athlon even if Apple rolls out a G5 PowerMac tomorrow, but your primary workstation might just be a Mac."

Yeah. I could take paying the moolah if the G5 was Apple's middle top end now.


;)

Lemon Bon Bon

[Amorph]

[quote]Originally posted by Lemon Bon Bon:
<strong>
Yes, I guess 'we'll see'. I think I'm one of those 'Mac watchers' who are dying to see the G4/G5 'debacle' resolved one way or another. It's like when we were 'waiting' for the 'flat panel' iMac. It came almost year after we expected it.</strong><hr></blockquote>

Who's we? Nobody with an eye on flat panel prices expected one last year. I don't think anyone predicted a flat panel and a G4 and a SuperDrive, either.

[quote]<strong>The 'G5' wait is becoming more infamous. It seems that the G4, in its numerous guises is going to take us until well into next year...</strong><hr></blockquote>

Yeah, well, strange how long the wait can be for something that doesn't exist in any (publicly) documented form...

[quote]<strong>Well, an exta FPU would be nice! But I'd like to see a couple in there. I think the changes you, the Register etc are all talking about will do...for New York Macworld. Well, it'll be interesting to see the 'actual' performance when the DDR mobo arrives. But, having seen the PC DDR benchmarks over PC133...we could be in for a disappointment...</strong><hr></blockquote>

Apple appears to be using an unconventional motherboard architecture (for a personal computer), so I wouldn't expect a direct comparison to PC implemetations - especially not the early ones.

[quote]"Macs do not have to beat x86 machines in every imaginable area in order to be competitive."

<strong>Of course not. But hammer them in most for the kind of money we pay? I think so :) </strong><hr></blockquote>

Price a PC with all the features a PM comes with, including internal bandwidth and free PCI slots, and see how much you end up saving (granting the fact that you will almost certainly be saving something). Sure, you can get a hot CPU on a stripped-down board in a generic case for a song, but that's not a fair comparison.

[quote]<strong>I can see what you're getting at. But I think the reason most of all...that Pro sales are suffering is that most people resent the idea of paying almost £3000 for kit that is clearly behind in performance for MOST tasks.</strong><hr></blockquote>

Most people seem to assume that the current PowerMacs are stopgap models. For all we know, they are. As for "behind in most tasks," depends on the task, doesn't it? Again, if the system and application integration is better, and the system and application design is easier to use, that addresses the biggest performance bottleneck in most workstation applications: The user. Not RAM. Not the CPU. The user. There are a couple of well-documented exceptions (3D apps, games, etc) but by and large, and particularly in Apple's traditional markets, this remains true.

[Capt. Obvious]

James, gotta tell you: you single-handedly contribute enough sense to make this forum readable. Thank you!!!

(Yes, others make sense, too....)

[HOS]

Some good speculation going on here...

Anecdote-

A friend of mine bought a dual GHz G4 with ACD (just before the $500 off mail-in rebate expired). He's claiming he's getting about a factor of 10 increase in speed over his Pismo G3 laptop (400). This is primarily with apps like Lightwave.

So how much faster does Apple need to go to remain "competitive?"

What's funny is to look at the other major RISC architectures- there's Sun's UltraSparc, currently at 900 MHz, and MIPS is about 600 MHz. Power4 is above 1.2 GHz, but that implements the PowerPC ISA.

So, benched against non-Intel and non-AMD CPUs strictly on a MHz basis, Moto isn't doing that badly. Where they fall down is in CPU to memory bandwidth, narrow busses, fewer functional units and the like.

Don't forget that further, the 7455 is built on a .18 micron process- the jump to .13 is just around the corner, and should bring along a fair (prob. 30% or so- but this is just a guess!) increase in speed. Since it's the same HIP-7 process that AMD is using, the next 745x should show up about the same time as AMD's .13 micron (plus SOI) products.

So consider an incremental improvement to the current design, since 7455's are available at 1.1 GHz today (yes- check out Moto's ordering page!), a 30% improvement gets us to almost 1.5 GHz.

If they can somehow improve bandwidth to prevent starving a 1.5 GHz relatively short-pipelined CPU (shouldn't be impossible) which should be here relatively soon (next 6-12 months at the latest, following AMD's HIP-7 rollout), then I'd say that Moto's still very much in the hunt, if not in the lead.

It really does make for some interesting speculation, no?

HOS

[Programmer]

[quote]Originally posted by HOS:
<strong>So, benched against non-Intel and non-AMD CPUs strictly on a MHz basis, Moto isn't doing that badly. Where they fall down is in CPU to memory bandwidth, narrow busses, fewer functional units and the like.</strong><hr></blockquote>

But Intel and AMD is where Apple/Motorola compete. They are behind in clock rate, memory bandwidth, and functional units. I agree that if they can address 2 out of three, and make sure that the third isn't too far off the mark, then they are back in the game. There isn't holding anything them back except their own ability and willingness to invest the necessary resources. The problem is that Motorola is in fairly dire straights, and has been for a while, which impacts when anything new and competitive can be delivered. The rumours are generally positive about what is coming, and the sooner it arrives the better. Fingers crossed for MWNY.

[rickag]

[quote]Programmer
"The problem is that Motorola is in fairly dire straights, and has been for a while, which impacts when anything new and competitive can be delivered."<hr></blockquote>

Exactly.

For a long time, basically since the G4 500MHz fiasco, I have wondered why Motorola didn't take a chance and really push the envelope on the G4 for the desktop market. After all the G4 is designed for the embedded market.

Granted, Apple is not a large percentage of the desktop market, but they had the ball rolling when the G4 first came out, gigaflops and all that.

I believe that Motorola,with a concerted effort in R&D and Production,they would now be close to reaping the benefits of a stronger Apple, with OS X maturing, inovative design and products,etc.

Where would Apple, and subsequently Motorola, be today if they had a 2.0 - 2.4 GHz G4 on a motherboard with DDRsdram and OS X?

[ 04-23-2002: Message edited by: rickag ]</p>

[gafferino]

[quote]Originally posted by rickag:
<strong>


Where would Apple, and subsequently Motorola, be today if they had a 2.0 - 2.4 GHz G4 on a motherboard with DDRsdram and OS X?

[ 04-23-2002: Message edited by: rickag ]</strong><hr></blockquote>

Hmm... they would be exactly where they are today plus a mild increase in towers. The vast majority of Apple users are not preformance driven. Apple gets most of its users for its UI and styling. Apple has done everything right in terms of getting apps ported to OSX and putting the nicest UI on a Unix core that anyone ever has. The hardare is stable and Apple boasts selling more Superdrives than the rest of the market combined.

Do you really think that any large percentage of the population would switch to Apple if it were a bit faster. Gamers would not switch (there will be a lag between PC and Apple software releases unless Apple gets 20% plus of the hardware market). Corporate users will not move quickly and don't buy cutting edge anyways. Value sensitive consumers would not switch. So yes a handful of scientfic converts, some more Unix/Linux people and maybe a few more loyal Apple upgraders.

Most people buy a new system every 3-5 years. Todays systems are pretty nice compared to what you got 3-5 years ago. The G3/G4 gave Apple lovers to chance to jump on the bleeding edge of speed for a short time and we loved it but Apple's market share did not skyrocket last time they were ahead of intel/AMD and it won't the next time either (presuming this happens).

Apple is in a comfortable niche market and if it plays all its cards right may grow a few more points of market share but really that is all we can expect in a 5-10 year window.

[rickag]

[quote]"Hmm... they would be exactly where they are today plus a mild increase in towers."<hr></blockquote>

I beg to differ, maybe not huge overall increase in sales, but certaintly significant to Apple. If there was just a 2% increase in sales market share that would almost double the # of computers Apples sells.

Lamentably, we will never know the answer.

[quote]"The vast majority of Apple users are not preformance driven"<hr></blockquote>

My first reaction to this statement was, HUH? Then it came to me, most Apple users aren't performance driven because in most apps Apple computers lag behind, so logically most Apple users can not be performance driven. It's not the user's fault they aren't performance driven it's Apple's and Motorola's.

Also, tell that to the graphics artists whose livelihood depends on speed.
(I know, I know, these users aren't the vast majority, but in Apple's market they are most probably a lot higher a percentage of Apple's market that say a Dell.)

And another thing, I was primarily refering to NOW, with the maturity of Mac OS X, innovative designs and products, in my humble opinion, Apple would be very well positioned to have very significant increases in market share, except, for the performance gap.

[ 04-24-2002: Message edited by: rickag ]

[ 04-24-2002: Message edited by: rickag ]</p>

[P]

On the subject of functional units: You do realize that the P4 has one FPU as well, don't you? It's also worth noting that the x87 is just about the worst FPU on the market. <a href="http://arstechnica.com/cpu/01q4/p4andg4e2/p4andg4e2-1.html" target="_blank">This article</a> on Ars Technica explains this pretty well.

I still maintain that the main problem about current Apple offerings is the lack of DDR, and if Apple releases Powermacs with new G4 7470s (with DDR) this summer, like the Reg suggests in the article linked earlier in this thread, I'm almost satisfied. I agree that the pipe coudl be longer to increase the clock speed, and that hyperthreading is the way to go, but I would stop feeling behind the pack if they do that.

[xype]

[quote]Originally posted by gafferino:
<strong>The vast majority of Apple users are not preformance driven.</strong><hr></blockquote>

I wonder to whom Apple is trying to sell "PowerMacs" then.