I thought that the G5 didn't have altivec in it, but the 8540 was a cantidate for altivec II, when it was ready.
In thought of all that is going on, I think the G5s are baited for a MW Tokyo announcement,, then released a month later.
Hopefully we'll see
1.4Ghz - 1.8Ghz G5s
333Mhz DDR
400Mhz effective bus
plus other 'goodies'.
Of course, this would be very expensive, but I think the new G5s will be aimed to combat SGI workstations. Do you think Mr Jobs would love to see animators using G5s than SGI worksations?
As for the optimistic clock speeds, the G5 is a much better chip than the G4, far better. It is supposed to have three times the speeds at the same clock speed in terms of GigaFlops.
Before the brain trust is totally dismantled, and we return to our regular diet of "predictions", counter-"predictions", and character assassination...here's my contribution.
What follows is a speculative scenario; I do not claim it is true; I do claim that it is possible and reasonable, and based in (some kind of) reality. I ask you to hold your fire until you've read the whole thing.
First, a timeline:
SJ returns to Apple, determined to vindicate his vision & hard work by reclaiming Apple & the Mac & returning them to 'respectability'; in doing so, he kills the clones. Motorola, hurt financially by the loss of their clone investment, + sales to other clonemeisters, says, "Apple is just another customer now.
I'll spare us an analysis of the AIM partnership (FTM), except to observe that this was a watershed event: Motorola's turning-away. From this point on MOT has been perceptibly cool on the subject of general-purpose CPUs. Not only did the clone-death cut into their revenue projections, but the endless death-dance of Amiga reduced MOT potential desktop penetration to the point of invisibility (remember, this was before the iMac & the post-Amelio rennaissance). MOT no longer counts on the desktop as a keystone for the future.
Knowing that only the fastest machines - much faster than those then-existing - will be able to support his plans for the future, SJ keeps the G4 under wraps & begins to map out G5 & G6. The G4 is released to much wonder, but trouble surfaces: the design doesn't scale, meaning, the usual tricks for squeezing speed increases out of a CPU design bring no speed increases; not only that, but yield problems lead to a clock downgrade shortly after release. The G4 begins to stagnate, and there are reports that Apple is not merely sinking more money into MOTs processor design unit, but that they're becoming 'more actively involved' in the design process.
As one of the AIM partners, Apple has ALWAYS been heavily involved in PPC design - for damn' good reason: Apple's machines are the only ones that DEPEND on the PPC (of any generation) being a solid, fast, scalable and extensible design.
There are 2 significant pieces in this section of the timeline: the paralysis of the G4, and the news of AAPLs increasing involvement in the design team. In some ways, "G4" is a misnomer: essentially a "G3+AltiVec", the chip might be more accurately called G3.5, rather than a generational leap (as the G3 was over the 6xx, and as G5 is expected to be). I speculate that the failure of G4 is due to fundamental design flaws related to the Altivec bolt-on tactic, and that these proved 'unfixable' in the short term - and this led to the real megaHertz gap. Satisfying as it may be to blame MOT's "stupidity", all it really took was a serious flaw and a lack of motivation...and their motivation was already waning. MOT's eggs were increasingly placed in the telephony/embedded-systems basket, and they lacked enthusiasm for the emergency design surgery AAPL demanded, even with the influx of extra AAPL cash.
In this scenario (here's where my speculation heads for the Outer Limits), faced w/ a public-relations disaster, a seriously black eye among tech types, sales resistance, and an increasingly distracted design partner, AAPL reacts by taking over G5 development. My reasoning is that AAPL must have some contractual leverage to protect them from losing the PPC; despite all the gains expected by each partner, the fact is that AAPL is the most vulnerable of the 3: it needs the PowerPC; not true of the others. So, AAPL being full of pretty smart characters, and some of them being lawyers, it stands to reason that the arrangement between AAPL & MOT would lock in Apple's intellectual property rights, etc., to the PPC, in the event that "changing fortunes" threatened MOT's (or IBM's) abiity or willingness to perform as needed in PPC development...in which case, MOT would be partner no nore, but just a contract chip fabricator (and AAPL "...just another customer".
This being the case, I'd expect that AAPL would have hired certain key people from MOT (perhaps after one of their "layoffs" - you know AAPL loves to travel under the radar) and elsewhere, &push HARD on G5 (w/ an eye toward G6, 'cause that's how these things go), undistracted by cell-phones and satellites. It's been 2 years (?) since then, and what have we heard? Vague rumours about mystery boxes that make Athlons look like dead snails, and pretty much nothing more.
RECAP
AAPL moves from triumph to triumph: G3, SJ's homecoming, iMac, B&W PMG3, PMG4;
Disaster strikes: tho a perfectly acceptable CPU, design flaws cripple its speed and its scalability;
Dismayed by MOT's preoccupation w/ its other products, and faced by destruction in the marketplace, AAPL takes steps: assumes control of PPC development; assigns MOT the task of getting whatever legs they can on the G4; assigns G5 crash priority; plays the situation close to the vest, in order to prevent further humiliation while sorting things out.
With both the iMac & PowerMac lines stale and floundering, a share-price tumble that leaves the entire market gasping, and the Cube fiasco still fresh, light begins to gleam: iMac's re-re-design comes together, OSX matures, and G5 prototypes exceed expectations.
The computer industry desperately needs "a compelling reason" for sales to increase, and AAPL is poised to pull the trigger on the most compelling line of computers ever seen.
Faced w/ an abundance of riches, AAPL chooses to release the iMac first; this will give it an undiluted spotlight, and allow time for orders, production and enthusiasm to gear up; it will also allow time to put the finishing touches on the new PowerMac G5.
After teh G4 fiasco, AAPL is taking no chances: the PowerMac G5 is compelling: state of the art, from mobo to I/O - the latest & greatest buzzword compliance, drool factor 11; to make it so, the extra time will all go into the legendary AAPL fit & finish; but when to announce? After all new iMac models are shipping: once that's done, the decks are cleared for introduction of the PMG5. Beyond that: any major software releases on the horizon? How's 10.2 coming along? Any big events coming up?
The piece of resistance: as soon as possible after G5 introduction, move all non-G5 machines to 1+gHz. The fastest, most capable, most attractive comuters ever made: half a dozen compelling reasons to buy not just a new computer, but a new Macintosh.
Hardware & software developers, desperate for a selling platform to support (& thereby generate sales of their own), flock to the Mac: gadget and peripheral makers fall over each other to be part of the digital hub.
This is me speculating. Not predicting, not claiming, not reporting. Speculating.
<strong>It's been interesting watching those who say "There is no G5" claim some sort of superiority over those who insist the G5 is coming out tomorrow.
...
And the mere statement that the G5 will _not_ be here is based on as much (or little) fact as any other argument.</strong><hr></blockquote>
A basic principle of prediction is that the less evidence, the more conservative your prediction should be. Right now, it IS a better prediction to say that the G5 is not coming out for a long time, simply because there's a lack of evidence that it's coming out soon.
However let me correct you on the G4 being more of a G3+Altivec a bit.
First of all, the G3 and a G4 without altivec would still differ, for example in the FPU, being double precision in the G4, while only single prec in G3.
Also, the G3 wasn't a new "architecture" either, it's based on the 603e and ev chip's design, the G4 being the son of the 604e line of chips.
Of course they were most likely rebuilt up from scratch, but the ideas and tricks of the old G2 chips flowed in, thus it's nothing abnormal, that the G4 also was influenced by the G3, as it was released afterwards.
One thing I like especially about your piece above is the fact that you alone seem to consider/know that development of a chip starts WAY before the public ever hears about it. Or does anyone here think the P4 core was developed during the last 1.5 years? 3 is most likely closer to it.
Chip design is a very complicated thing, you gotta start early, to deliver in time. And in time you want to deliver, because the competition doesn't wait. Now I think even Moto knows that, even if you guys don't think so.
I'm confident we'll see the G5 sooner than many think.
Tell me philbot why are you so confident about these G5s?
<hr></blockquote>
OK I don't want to turn this thread into another G5 existence debate (it is going well so far) but, seeing is believing.
Let's all just have a good time discussing the future shall we?
The iMac is great publicity and will expand market share for Apple, which is top priority.
I think they can afford to wait a while longer to get the G5 just right. By the way, it might not be that expensive to put duals in the lineup. But they could go bananas over the graphics system etc.
Brilliant post, Capt. Obvious! I have said and believed for some time now that Apple is playing a big role in G5 development, moreso than a lot of people here think. I think this is why the 8540 embedded chip really doesn't have anything to do with Apple's desktop variant, nor does the Apollo timetable really figure in, since it will be used to keep the iMac, iBook, and TiBook competitive for the foreseeable future, while the PowerMac line will migrate to the G5.
I think Apple has really learned from its mistakes. Steve Jobs is a smart man, and he wants the fastest and the best. Do you not think speed is important to him? Do you remember what a mockery he made of the P3 when the G4 was first announced? Do you remember the look on his face? I do, and I think he's pulling what AMD did a long time ago: waiting out a bad design and devoting the PowerMac team resources to the G5 project to completely leapfrog Intel AND AMD in terms of performance.
AS I mentioned earlier, March could be the intro month, since we've got 10.2 to be said to be released then, in addition to the Screen Saver promo running out by March 31. The G5s could begin shipping by mid April, which leaves the current Dual GHz G4s the same kind of life span as the 450 MHz G3 had before the G4s were sprung on us.
OK lets say apple has been developing the G5 and the chip is almost or is ready. Apple has been waiting for the iMac to release it. The best time would be MWT would it not. Maybe even seybold.
Maybe we will get new LCDs as well.
Would they, do you think, release G5 powerbooks at the same time as Powermacs. I wonder if the G5 has the same heat issues as the G4. MAybe...
<strong>OK I don't want to turn this thread into another G5 existence debate (it is going well so far) but, seeing is believing. </strong><hr></blockquote>
These improvements seem particularly minor in comparison with the big things Mac insiders say Apple has in store for its professional systems. Multiple eyewitnesses insist that Apple is far along in the development of new towers that already offer clock speeds in the neighborhood of 1.6GHz and triple the performance of the current systems' 133MHz system bus. My sources say the tangible boosts in graphics performance and Mac OS X speeds are immense.
As far as when G5 will hit the streets, only those inside the (infinite) loop can say; There are so many factors that govern a major product release that prediction is practically impossible (we've proved that here, haven't we?
That said, it would surprise me if we saw the new PMs as early as March; I'd say that would be crowding the iMac unnecessarily. I think this summer is much more likely: more room for the iMac, and sales of these puppies could go through the roof once the proud owners thereof start posting their experiences. Seybold would be good, but MWNY is the natural venue for it; either way, it's a cinch that SJ will be on-stage for the roll-out (including one last Photoshop bake-off ),
Once it's out, I'd look for SJ to bump all other Mac lines to the gHz level, and maybe keep a mid-priced line of PMG4s.
Looking further down the road 3-5 years, we can expect to see the iMacs pickup G5 power when the PMs move on to G6.
From what I've read the G5 is very big, and hot, so I doubt we'll see it in the PowerBook until 2003. I think the 7445 Apollo at 667 and 800 MHz will hold over the pro portables until Seybold in August. (or Sept., I forgot)
<strong>Brilliant post, Capt. Obvious! I have said and believed for some time now that Apple is playing a big role in G5 development, moreso than a lot of people here think. I think this is why the 8540 embedded chip really doesn't have anything to do with Apple's desktop variant, nor does the Apollo timetable really figure in, since it will be used to keep the iMac, iBook, and TiBook competitive for the foreseeable future, while the PowerMac line will migrate to the G5.
</strong><hr></blockquote>
Ditto from me to Capt. Obvious. Well reasoned post!
Re: 8540
I think most people have this backwards. They assume the desktop G5 will be a derivative of the embedded 8540 processor. My suspicion is that the 8540 is actually derived from the desktop G5. At the very least, a joint development agreement for the general architecture, where Apple get the desktop version (which it keeps to itself) and Moto gets the embedded version (which it publicly announces). There is no reason, IMHO, to assume that the G5 must come out after the 8540.
I also agree with you about the Apollo. It has struck me from the start that its real home was in the iMac and the i/TiBooks. Its appearance in the PM is a stopgap measure until the big guns get rolled out - MWNY by my guess.
:: Thanks, I welcome your comments! (thanks to TJM & Tarbash, too!)
[quote]<strong>However let me correct you on the G4 being more of a G3+Altivec a bit.
First of all, the G3 and a G4 without altivec would still differ, for example in the FPU, being double precision in the G4, while only single prec in G3.
Also, the G3 wasn't a new "architecture" either, it's based on the 603e and ev chip's design, the G4 being the son of the 604e line of chips.
Of course they were most likely rebuilt up from scratch, but the ideas and tricks of the old G2 chips flowed in, thus it's nothing abnormal, that the G4 also was influenced by the G3, as it was released afterwards.</strong><hr></blockquote>
I didn't intend to suggest that G3 & G4 were identical in all ways except for Altivec; my impression (which may be faulty) is that the differences between those two generations were more minor than those separating them from their successor(s) and antecedents.
As you rightly point out, each chip design draws from and builds upon its predecessors; this is one of the reasons that designers have to look two generations ahead; for example, implementing a design will result in ideas that can't be incorporated in the design (due to cost factors and delivery schedules, among others) - but which can be make it into the next generation.
It's a lot like life, really: you see your project thru to the end, consider what you've learned, and use your experience to take a fresh look at your next challenge. It's what gives rise to the metaphor of generations.
[quote]<strong>One thing I like especially about your piece above is the fact that you alone seem to consider/know that development of a chip starts WAY before the public ever hears about it.</strong><hr></blockquote>
Not me alone, surely! I'm certain Programmer can say the same about software development. ANY major developmental project is a lengthy process, subject to snags, reversals, and outright failures; involves a steeper-than-expected learning curve; builds on previous efforts; and could be revised and reworked endlessly were it not for those pesky considerations...
<strong>Before the brain trust is totally dismantled, and we return to our regular diet of "predictions", counter-"predictions", and character assassination...here's my contribution.
*snip!*
</strong><hr></blockquote>
Wow, true or not, speculation or not (it is.) this is a great read. I found myself smiling and trying to supress a maniacal giggle. This is AAPL as it should be. I can totally buy the part about making the G5 a crash priority, though. Gods, I hope so!
Thanks for the wonderful post!
tsukurite
edit: Just a second thought. I really wish financial reports could be written with this much drama. Maybe APPL (I like the idea of using the stock market abbreviation), could contract Tom Clancy or someone. How many financials are also best sellers? Just thought. - tsukurite
I must contribute additional praise for the insightful, entertaining post by Capt. This is a really great thread, so let's keep it alive!
Many individuals are hoping for a G5 surprise at MWT or MWNY. I just don't see it happening. I still believe we will receive an Apollo Power Mac bump at MWNY, hopefully bringing us to 1.3GHz. The Motorola reps have said Apollo has legs, but on the other hand, if it can scale so well, why did we only get 133MHz after six long months? 1.3GHz seems like a lot to hope for. If Apollo doesn't scale as well as hoped, then that would likely prompt Apple to push the G5 release date forward.
I don?t have confidence in a G5 introduction pre-MWSF; hopefully the extra time Apple has will allow for greater MHz attainment. I don't know about anyone else, but a G5 that tops out at 1.6GHz is not a panacea, since by MWNY Intel maybe pushing 2.4GHz or beyond. I understand the GHz Myth, so I don't need any instruction on the subject. However, I want the G5s introduction to not be tainted in anyway, and that means, at least to me, that the line should be damn close to GHz parity, if not GHz dominance.
@ Capt. Obvious: Nice to see we share a common view of things.
@ Eugene: No the 9500 started as the first PCI PowerMac at 120MHz 604 chip. Very shortly after, though, it was introed at 132MHz along with the the 8500 at 120MHz and then also the 7500 at 100MHz 601 chip.
I don't know for sure, but it could be that the 9500/120 was a European only model, yet I doubt it.
@ everyone else: This is the best thread in years here.
Comments
G-News
In thought of all that is going on, I think the G5s are baited for a MW Tokyo announcement,, then released a month later.
Hopefully we'll see
1.4Ghz - 1.8Ghz G5s
333Mhz DDR
400Mhz effective bus
plus other 'goodies'.
Of course, this would be very expensive, but I think the new G5s will be aimed to combat SGI workstations. Do you think Mr Jobs would love to see animators using G5s than SGI worksations?
As for the optimistic clock speeds, the G5 is a much better chip than the G4, far better. It is supposed to have three times the speeds at the same clock speed in terms of GigaFlops.
[ 01-30-2002: Message edited by: mattyj ]</p>
What follows is a speculative scenario; I do not claim it is true; I do claim that it is possible and reasonable, and based in (some kind of) reality. I ask you to hold your fire until you've read the whole thing.
First, a timeline:
SJ returns to Apple, determined to vindicate his vision & hard work by reclaiming Apple & the Mac & returning them to 'respectability'; in doing so, he kills the clones. Motorola, hurt financially by the loss of their clone investment, + sales to other clonemeisters, says, "Apple is just another customer now.
I'll spare us an analysis of the AIM partnership (FTM), except to observe that this was a watershed event: Motorola's turning-away. From this point on MOT has been perceptibly cool on the subject of general-purpose CPUs. Not only did the clone-death cut into their revenue projections, but the endless death-dance of Amiga reduced MOT potential desktop penetration to the point of invisibility (remember, this was before the iMac & the post-Amelio rennaissance). MOT no longer counts on the desktop as a keystone for the future.
Knowing that only the fastest machines - much faster than those then-existing - will be able to support his plans for the future, SJ keeps the G4 under wraps & begins to map out G5 & G6. The G4 is released to much wonder, but trouble surfaces: the design doesn't scale, meaning, the usual tricks for squeezing speed increases out of a CPU design bring no speed increases; not only that, but yield problems lead to a clock downgrade shortly after release. The G4 begins to stagnate, and there are reports that Apple is not merely sinking more money into MOTs processor design unit, but that they're becoming 'more actively involved' in the design process.
As one of the AIM partners, Apple has ALWAYS been heavily involved in PPC design - for damn' good reason: Apple's machines are the only ones that DEPEND on the PPC (of any generation) being a solid, fast, scalable and extensible design.
There are 2 significant pieces in this section of the timeline: the paralysis of the G4, and the news of AAPLs increasing involvement in the design team. In some ways, "G4" is a misnomer: essentially a "G3+AltiVec", the chip might be more accurately called G3.5, rather than a generational leap (as the G3 was over the 6xx, and as G5 is expected to be). I speculate that the failure of G4 is due to fundamental design flaws related to the Altivec bolt-on tactic, and that these proved 'unfixable' in the short term - and this led to the real megaHertz gap. Satisfying as it may be to blame MOT's "stupidity", all it really took was a serious flaw and a lack of motivation...and their motivation was already waning. MOT's eggs were increasingly placed in the telephony/embedded-systems basket, and they lacked enthusiasm for the emergency design surgery AAPL demanded, even with the influx of extra AAPL cash.
In this scenario (here's where my speculation heads for the Outer Limits), faced w/ a public-relations disaster, a seriously black eye among tech types, sales resistance, and an increasingly distracted design partner, AAPL reacts by taking over G5 development. My reasoning is that AAPL must have some contractual leverage to protect them from losing the PPC; despite all the gains expected by each partner, the fact is that AAPL is the most vulnerable of the 3: it needs the PowerPC; not true of the others. So, AAPL being full of pretty smart characters, and some of them being lawyers, it stands to reason that the arrangement between AAPL & MOT would lock in Apple's intellectual property rights, etc., to the PPC, in the event that "changing fortunes" threatened MOT's (or IBM's) abiity or willingness to perform as needed in PPC development...in which case, MOT would be partner no nore, but just a contract chip fabricator (and AAPL "...just another customer".
This being the case, I'd expect that AAPL would have hired certain key people from MOT (perhaps after one of their "layoffs" - you know AAPL loves to travel under the radar) and elsewhere, &push HARD on G5 (w/ an eye toward G6, 'cause that's how these things go), undistracted by cell-phones and satellites. It's been 2 years (?) since then, and what have we heard? Vague rumours about mystery boxes that make Athlons look like dead snails, and pretty much nothing more.
RECAP
AAPL moves from triumph to triumph: G3, SJ's homecoming, iMac, B&W PMG3, PMG4;
Disaster strikes: tho a perfectly acceptable CPU, design flaws cripple its speed and its scalability;
Dismayed by MOT's preoccupation w/ its other products, and faced by destruction in the marketplace, AAPL takes steps: assumes control of PPC development; assigns MOT the task of getting whatever legs they can on the G4; assigns G5 crash priority; plays the situation close to the vest, in order to prevent further humiliation while sorting things out.
With both the iMac & PowerMac lines stale and floundering, a share-price tumble that leaves the entire market gasping, and the Cube fiasco still fresh, light begins to gleam: iMac's re-re-design comes together, OSX matures, and G5 prototypes exceed expectations.
The computer industry desperately needs "a compelling reason" for sales to increase, and AAPL is poised to pull the trigger on the most compelling line of computers ever seen.
Faced w/ an abundance of riches, AAPL chooses to release the iMac first; this will give it an undiluted spotlight, and allow time for orders, production and enthusiasm to gear up; it will also allow time to put the finishing touches on the new PowerMac G5.
After teh G4 fiasco, AAPL is taking no chances: the PowerMac G5 is compelling: state of the art, from mobo to I/O - the latest & greatest buzzword compliance, drool factor 11; to make it so, the extra time will all go into the legendary AAPL fit & finish; but when to announce? After all new iMac models are shipping: once that's done, the decks are cleared for introduction of the PMG5. Beyond that: any major software releases on the horizon? How's 10.2 coming along? Any big events coming up?
The piece of resistance: as soon as possible after G5 introduction, move all non-G5 machines to 1+gHz. The fastest, most capable, most attractive comuters ever made: half a dozen compelling reasons to buy not just a new computer, but a new Macintosh.
Hardware & software developers, desperate for a selling platform to support (& thereby generate sales of their own), flock to the Mac: gadget and peripheral makers fall over each other to be part of the digital hub.
This is me speculating. Not predicting, not claiming, not reporting. Speculating.
<strong>It's been interesting watching those who say "There is no G5" claim some sort of superiority over those who insist the G5 is coming out tomorrow.
...
And the mere statement that the G5 will _not_ be here is based on as much (or little) fact as any other argument.</strong><hr></blockquote>
A basic principle of prediction is that the less evidence, the more conservative your prediction should be. Right now, it IS a better prediction to say that the G5 is not coming out for a long time, simply because there's a lack of evidence that it's coming out soon.
However let me correct you on the G4 being more of a G3+Altivec a bit.
First of all, the G3 and a G4 without altivec would still differ, for example in the FPU, being double precision in the G4, while only single prec in G3.
Also, the G3 wasn't a new "architecture" either, it's based on the 603e and ev chip's design, the G4 being the son of the 604e line of chips.
Of course they were most likely rebuilt up from scratch, but the ideas and tricks of the old G2 chips flowed in, thus it's nothing abnormal, that the G4 also was influenced by the G3, as it was released afterwards.
One thing I like especially about your piece above is the fact that you alone seem to consider/know that development of a chip starts WAY before the public ever hears about it. Or does anyone here think the P4 core was developed during the last 1.5 years? 3 is most likely closer to it.
Chip design is a very complicated thing, you gotta start early, to deliver in time. And in time you want to deliver, because the competition doesn't wait. Now I think even Moto knows that, even if you guys don't think so.
I'm confident we'll see the G5 sooner than many think.
G-News
Tell me philbot why are you so confident about these G5s?
<hr></blockquote>
OK I don't want to turn this thread into another G5 existence debate (it is going well so far) but, seeing is believing.
Let's all just have a good time discussing the future shall we?
The iMac is great publicity and will expand market share for Apple, which is top priority.
I think they can afford to wait a while longer to get the G5 just right. By the way, it might not be that expensive to put duals in the lineup. But they could go bananas over the graphics system etc.
Here's hoping!
[ 01-30-2002: Message edited by: philbot ]</p>
I think Apple has really learned from its mistakes. Steve Jobs is a smart man, and he wants the fastest and the best. Do you not think speed is important to him? Do you remember what a mockery he made of the P3 when the G4 was first announced? Do you remember the look on his face? I do, and I think he's pulling what AMD did a long time ago: waiting out a bad design and devoting the PowerMac team resources to the G5 project to completely leapfrog Intel AND AMD in terms of performance.
AS I mentioned earlier, March could be the intro month, since we've got 10.2 to be said to be released then, in addition to the Screen Saver promo running out by March 31. The G5s could begin shipping by mid April, which leaves the current Dual GHz G4s the same kind of life span as the 450 MHz G3 had before the G4s were sprung on us.
Maybe we will get new LCDs as well.
Would they, do you think, release G5 powerbooks at the same time as Powermacs. I wonder if the G5 has the same heat issues as the G4. MAybe...
<strong>OK I don't want to turn this thread into another G5 existence debate (it is going well so far) but, seeing is believing.
Just a little reminder for everyone:
[quote]Topic: G5 next week. Certain.
posted 01-19-2002
philbot <hr></blockquote>
<a href="http://www.eweek.com/article/0,3658,s%3D25201%26a%3D22082,00.asp" target="_blank">www.eweek.com/article/0,3658,s%253D25201%2526a%253D22082,00.asp</a>
specially this part :
These improvements seem particularly minor in comparison with the big things Mac insiders say Apple has in store for its professional systems. Multiple eyewitnesses insist that Apple is far along in the development of new towers that already offer clock speeds in the neighborhood of 1.6GHz and triple the performance of the current systems' 133MHz system bus. My sources say the tangible boosts in graphics performance and Mac OS X speeds are immense.
That said, it would surprise me if we saw the new PMs as early as March; I'd say that would be crowding the iMac unnecessarily. I think this summer is much more likely: more room for the iMac, and sales of these puppies could go through the roof once the proud owners thereof start posting their experiences. Seybold would be good, but MWNY is the natural venue for it; either way, it's a cinch that SJ will be on-stage for the roll-out (including one last Photoshop bake-off
Once it's out, I'd look for SJ to bump all other Mac lines to the gHz level, and maybe keep a mid-priced line of PMG4s.
Looking further down the road 3-5 years, we can expect to see the iMacs pickup G5 power when the PMs move on to G6.
Speedbump history (top models):
{604xx: 120-132-150-180-200-225-233-350 (8 bumps, 3 different CPUs)}
1.5 mainboard generations
G3: 266-300-333-400-450 (5 bumps, 3 beige, 2 blue)
2 mainboard generations
G4: 400-450-500-533-733-867-1000 (7 bumps, 1 yikes, 1 downgraded sawtooth, 1 sawtooth, 3 sawtooth digi audio, 1 apollo)
3.5 mainboard generations
Even leaving the Yikes and downgrade out of the game, we have reached the 5 bump-mark with this release..."sense a pattern?"<hr></blockquote>
I seem to remember the 604's debut at 132 MHz, for Apple anyway. The first one was the 9500/132, wasn't it?
Anyway:
132-->350 = ~165% increase
266-->450 = ~69% increase
450-->1000 = ~122% increase
Anyway, I'd like to think of the processors differently though.
The 745x/744x series is remarkably different from the 7400/7410.
745x/744x = 733-->1000
7410/7400 = 450-->533
750 = 266-->500
750cx = 466-->700
604/604e = 132-->350
et cetera. You can manipulate the data whichever way you want. The G4 still has legs, IMHO.
[ 01-30-2002: Message edited by: Eugene ]</p>
<strong>Brilliant post, Capt. Obvious! I have said and believed for some time now that Apple is playing a big role in G5 development, moreso than a lot of people here think. I think this is why the 8540 embedded chip really doesn't have anything to do with Apple's desktop variant, nor does the Apollo timetable really figure in, since it will be used to keep the iMac, iBook, and TiBook competitive for the foreseeable future, while the PowerMac line will migrate to the G5.
</strong><hr></blockquote>
Ditto from me to Capt. Obvious. Well reasoned post!
Re: 8540
I think most people have this backwards. They assume the desktop G5 will be a derivative of the embedded 8540 processor. My suspicion is that the 8540 is actually derived from the desktop G5. At the very least, a joint development agreement for the general architecture, where Apple get the desktop version (which it keeps to itself) and Moto gets the embedded version (which it publicly announces). There is no reason, IMHO, to assume that the G5 must come out after the 8540.
I also agree with you about the Apollo. It has struck me from the start that its real home was in the iMac and the i/TiBooks. Its appearance in the PM is a stopgap measure until the big guns get rolled out - MWNY by my guess.
<strong>Very nice
:
[quote]<strong>However let me correct you on the G4 being more of a G3+Altivec a bit.
First of all, the G3 and a G4 without altivec would still differ, for example in the FPU, being double precision in the G4, while only single prec in G3.
Also, the G3 wasn't a new "architecture" either, it's based on the 603e and ev chip's design, the G4 being the son of the 604e line of chips.
Of course they were most likely rebuilt up from scratch, but the ideas and tricks of the old G2 chips flowed in, thus it's nothing abnormal, that the G4 also was influenced by the G3, as it was released afterwards.</strong><hr></blockquote>
I didn't intend to suggest that G3 & G4 were identical in all ways except for Altivec; my impression (which may be faulty) is that the differences between those two generations were more minor than those separating them from their successor(s) and antecedents.
As you rightly point out, each chip design draws from and builds upon its predecessors; this is one of the reasons that designers have to look two generations ahead; for example, implementing a design will result in ideas that can't be incorporated in the design (due to cost factors and delivery schedules, among others) - but which can be make it into the next generation.
It's a lot like life, really: you see your project thru to the end, consider what you've learned, and use your experience to take a fresh look at your next challenge. It's what gives rise to the metaphor of generations.
[quote]<strong>One thing I like especially about your piece above is the fact that you alone seem to consider/know that development of a chip starts WAY before the public ever hears about it.</strong><hr></blockquote>
Not me alone, surely! I'm certain Programmer can say the same about software development. ANY major developmental project is a lengthy process, subject to snags, reversals, and outright failures; involves a steeper-than-expected learning curve; builds on previous efforts; and could be revised and reworked endlessly were it not for those pesky considerations...
[ 01-30-2002: Message edited by: Capt. Obvious ]</p>
whether accurate or not your post made these boards fun again, the truth mixed with style mixed with hope is a joy to read. appreciate your post.
<strong>Before the brain trust is totally dismantled, and we return to our regular diet of "predictions", counter-"predictions", and character assassination...here's my contribution.
*snip!*
</strong><hr></blockquote>
Wow, true or not, speculation or not (it is.) this is a great read. I found myself smiling and trying to supress a maniacal giggle. This is AAPL as it should be. I can totally buy the part about making the G5 a crash priority, though. Gods, I hope so!
Thanks for the wonderful post!
tsukurite
edit: Just a second thought. I really wish financial reports could be written with this much drama.
[ 01-31-2002: Message edited by: tsukurite ]</p>
Many individuals are hoping for a G5 surprise at MWT or MWNY. I just don't see it happening. I still believe we will receive an Apollo Power Mac bump at MWNY, hopefully bringing us to 1.3GHz. The Motorola reps have said Apollo has legs, but on the other hand, if it can scale so well, why did we only get 133MHz after six long months? 1.3GHz seems like a lot to hope for. If Apollo doesn't scale as well as hoped, then that would likely prompt Apple to push the G5 release date forward.
I don?t have confidence in a G5 introduction pre-MWSF; hopefully the extra time Apple has will allow for greater MHz attainment. I don't know about anyone else, but a G5 that tops out at 1.6GHz is not a panacea, since by MWNY Intel maybe pushing 2.4GHz or beyond. I understand the GHz Myth, so I don't need any instruction on the subject. However, I want the G5s introduction to not be tainted in anyway, and that means, at least to me, that the line should be damn close to GHz parity, if not GHz dominance.
[ 01-31-2002: Message edited by: Big Mac ]</p>
@ Eugene: No the 9500 started as the first PCI PowerMac at 120MHz 604 chip. Very shortly after, though, it was introed at 132MHz along with the the 8500 at 120MHz and then also the 7500 at 100MHz 601 chip.
I don't know for sure, but it could be that the 9500/120 was a European only model, yet I doubt it.
@ everyone else: This is the best thread in years here.
G-News
<a href="http://www.everymac.com/systems/apple/powermac/stats/powermac_9500_132.html"; target="_blank">http://www.everymac.com/systems/apple/powermac/stats/powermac_9500_132.html</a>;
<a href="http://www.everymac.com/systems/apple/powermac/stats/powermac_9500_120.html"; target="_blank">http://www.everymac.com/systems/apple/powermac/stats/powermac_9500_120.html</a>;